Automated liquid handling for highly specific and reproducible density‐based separation of extracellular vesicles from human body fluids

Introduction: Extracellular vesicles (EV) in body fluids are extensively studied as potential biomarkers for numerous diseases. Major impediments of EV‐based biomarker discovery include the specificity and reproducibility of EV sample preparation. To tackle this, we present an automated liquid handling workstation for the density‐based separation of EV from human body fluids and compare its performance to manual handling by (in)experienced researchers. Methods: Variability in density‐based EV separation using manual versus automated liquid handling is first evaluated by spiking PBS with trackable recombinant extracellular vesicles (rEV) followed by the quantification of rEV recovery efficiency using fluorescent NTA and ELISA (Geeurickx et al., Nature Comm, 2019). Next, EV are separated from blood plasma and urine through the orthogonal implementation of size‐exclusion chromatography and manual or automated density‐gradient centrifugation as previously described (Dhondt et al., STAR Protoc, 2020; Tulkens et al., Nature Protoc, 2020). Variability, EV yield and purity are assessed by MS‐based proteomics and transmission electron microscopy. Results: Automated versus manual liquid handling significantly reduces variability in rEV recovery after density‐based separation (CVauto 11.6% vs CVman 26.4%), and significantly decreases variability in EV preparations obtained from blood plasma and urine. Indeed, the median CVauto for protein group quantification are between 17.2% (plasma) and 20.0% (urine), CVman,exp between 38.5% (plasma) and 22.6% (urine), and CVman,inexp between 43.2% (plasma) and 27.6% (urine). While retaining an equal EV yield compared to manual liquid handling, automation significantly diminishes the presence of contaminating abundant proteins in EV preparations, including lipoproteins in plasma (2‐fold decrease in ApoB) and uromodulin in urine (2.6‐fold decrease in THP). Summary/Conclusion: Automated liquid handling ensures EV separation from body fluids with high reproducibility and specificity.