Chronic Traumatic Encephalopathy (CTE) is a tauopathy that affects individuals with a history of exposure to repetitive head impacts, including National Football League (NFL) players. Extracellular vesicles (EVs) are known to carry tau in Alzheimer's disease and other tauopathies. We examined protein profiles of EVs separated from the plasma of former NFL players at risk for CTE. EVs were separated from the plasma from former NFL players and age-matched controls using size-exclusion chromatography. Label-free quantitative proteomic analysis identified 675 proteins in plasma EVs, and 17 proteins were significantly differentially expressed between former NFL players and controls. Total tau (t-tau) and tau phosphorylated at threonie181 (p-tau181) in plasma-derived EVs were measured by ultrasensitive immunoassay. Level of t-tau and p-tau181 in EVs were significantly different, and the area under the receiver operating characteristic curve (AUC) of t-tau and p-tau181 showed 0.736 and 0.715, respectively. Machine learning analysis indicated that a combination of collagen type VI alpha 3 and 1 chain (COL6A3 and COL6A1) and reelin (RELN) can distinguish former NFL players from controls with 85% accuracy (AUC = 0.85). Based on the plasma EV proteomics, these data provide protein profiling of plasma EVs for CTE, and indicate combination of COL6A3, RELN and COL6A1 in plasma EVs may serve as the potential diagnostic biomarkers for CTE.
Keywords: chronic traumatic encephalopathy; extracellular vesicles; machine learning; plasma; proteome.
Copyright: © 2021 Muraoka et al.