Comparing and validating current methods for isolating and characterizing brain-derived extracellular vesicles as central nervous system biomarkers of environmentally linked neurodegenerative diseases

Background and aims: Many environmental exposures (e.g., metals, pesticides) accumulate in central nervous system (CNS) tissues and are associated with adverse neurological outcomes. There is a great need for peripheral biomarkers that could provide insight into the CNS burden of these exposures to confirm the etiologic role of neurotoxic exposures. Our research group has observed CNS cells extruding toxicants via extracellular vesicles (EVs) and is characterizing the viability of CNS-derived EVs isolated from peripheral blood draws as potential biomarkers of environmental exposures. Several protocols for isolating EVs originating from the CNS via immunoprecipitation (IP) can be found in the literature, and these protocols are often replicated with only superficial validation. Here, we thoroughly validated and contrasted current EV isolation methods, tested optimal protease inhibition conditions, and compared IP efficiency and normalization techniques. Methods: We compared a one-step direct IP protocol with the widely used two-step Exoquick-based IP protocol for isolating CNS-EVs from human plasma using antibodies against CNS cell surface markers: EAAT1 for astrocytic EVs and L1CAM for neuronal EVs. EVs were characterized according to the Minimal Information for Studies of Extracellular Vesicles guidelines, including transmission election microscopy, nanotracking particle analysis, and protein ELISAs. Results: Our results show that a direct IP protocol is more effective for isolating CNS-derived EVs than commonly used Exoquick-based two-step protocols. These results demonstrate that EAAT1 antibodies are more effective than L1CAM in isolating-CNS-EVs even in conditions that are expected prevent L1CAM cleavage by metalloproteases. Results confirmed CD81 to be the best normalization factor for CNS-EV IP efficiency. Conclusions: These data add to the ongoing controversy that L1CAM and Exoquick may not consistently isolate EVs of neuronal origin. This may be partly due to variability in immunoreagent quality. These data strongly demonstrate the importance of rigorous method validation before beginning cell-specific EV-based biomarker studies.