Benchtop isolation and characterization of functional exosomes by sequential filtration

Early and minimally invasive detection of malignant events or other pathologies is of utmost importance in the pursuit of improved patient care and outcomes. Recent evidence indicates that exosomes and extracellular vesicles in serum and body fluids can contain nucleic acid, protein, and other biomarkers. Accordingly, there is great interest in applying these clinically as prognostic, predictive, pharmacodynamic, and early detection indicators. Nevertheless, existing exosome isolation methods can be time-consuming, require specialized equipment, and/or present other inefficiencies regarding purity, reproducibility and assay cost. We have developed a straightforward, three-step protocol for exosome isolation of cell culture supernatants or large volumes of biofluid based on sequential steps of dead-end pre-filtration, tangential flow filtration (TFF), and low-pressure track-etched membrane filtration that we introduce here. Our approach yields exosome preparations of high purity and defined size distribution and facilitates depletion of free protein and other low-molecular-weight species, extracellular vesicles larger than 100nm, and cell debris. Samples of exosomes prepared using the approach were verified morphologically by nanoparticle tracking analysis and electron microscopy, and mass spectrometry analyses confirmed the presence of previously reported exosome-associated proteins. In addition to being easy-to-implement, sequential filtration yields exosomes of high purity and, importantly, functional integrity as a result of the relatively low-magnitude manipulation forces employed during isolation. This answers an unmet need for preparation of minimally manipulated exosomes for investigations into exosome function and basic biology. Further, the strategy is amenable to translation for clinical exosome isolations because of its speed, automatability, scalability, and specificity for isolating exosomes from complex biological samples.