Confocal microscopy analysis reveals that only a small proportion of extracellular vesicles are successfully labelled with commonly utilised staining methods

Assessing genuine extracellular vesicle (EV) uptake is crucial for understanding the functional roles of EVs. This study measured the bona fide labelling of EVs utilising two commonly used fluorescent dyes, PKH26 and C5-maleimide-Alexa633. MCF7 EVs tagged with mEmerald-CD81 were isolated from conditioned media by size exclusion chromatography (SEC) and characterised using Nanoparticle Tracking Analysis (NTA), Transmission Electron Microscopy (TEM), MACsPlex immunocapture assay and immunoblots. These fluorescently tagged EVs were subsequently stained with C5-maleimide-Alexa633 or PKH26, according to published protocols. Colocalisation of dual-labelled EVs was assessed by confocal microscopy and quantified using the Rank-Weighted Colocalisation (RWC) algorithm. We observed strikingly poor colocalisation between mEmerald-CD81-tagged EVs and C5-Maleimide-Alexa633 (5.4%+/- 1.8) or PKH26 (4.6%+/- 1.6), that remained low even when serum was removed from preparations. Our data confirms previous work showing that some dyes form contaminating aggregates. Furthermore, uptake studies showed that maleimide and mEmerald-CD81-tagged EVs can be often located into non-overlapping subcellular locations. By using common methods to isolate and stain EVs we observed that most EVs remained unstained and most dye signal does not appear to be EV associated. Our work shows that there is an urgent need for optimisation and standardisation in how EV researchers use these tools to assess genuine EV signals.

Automated summary

Assessing the genuine uptake of extracellular vesicles (EVs) is crucial for understanding their functional roles. This study used two commonly used fluorescent dyes, PKH26 and C5-maleimide-Alexa633, to measure the bona fide labeling of EVs. The results showed poor colocalization between the labeled EVs and the dyes, indicating that the dye signal is not associated with EVs. These findings highlight the need for optimization and standardization in how researchers assess genuine EV signals using these tools.