Extracellular Vesicles (EVs) have been intensively explored for therapeutic delivery
of proteins. However, methods to quantify cargo proteins loaded into engineered EVs
are lacking. Here, we describe a workflow for EV analysis at the single-vesicle and
single-molecule level to accurately quantify the efficiency of different EV-sorting
proteins in promoting cargo loading into EVs. Expi293F cells were engineered to express
EV-sorting proteins fused to green fluorescent protein (GFP). High levels of GFP loading
into secreted EVs was confirmed by Western blotting for specific EV-sorting domains,
but quantitative single-vesicle analysis by Nanoflow cytometry detected GFP in less
than half of the particles analysed, reflecting EV heterogeneity. Anti-tetraspanin
EV immunostaining in ExoView confirmed a heterogeneous GFP distribution in distinct
subpopulations of CD63(+), CD81(+), or CD9(+) EVs. Loading of GFP into individual
vesicles was quantified by Single-Molecule Localization Microscopy. The combined results
demonstrated TSPAN14, CD63 and CD63/CD81 fused to the PDGFR beta transmembrane domain
as the most efficient EV-sorting proteins, accumulating on average 50-170 single GFP
molecules per vesicle. In conclusion, we validated a set of complementary techniques
suitable for high-resolution analysis of EV preparations that reliably capture their
heterogeneity, and propose highly efficient EV-sorting proteins to be used in EV engineering
applications.
