ß-Amyloid in blood neuronal-derived extracellular vesicles is elevated in cognitively normal adults at risk of Alzheimer's disease and predicts cerebral amyloidosis.

ABSTRACT

BACKGROUND: Blood biomarkers that can be used for preclinical Alzheimer's disease (AD) diagnosis would enable trial enrollment at a time when the disease is potentially reversible. Here, we investigated plasma neuronal-derived extracellular vesicle (nEV) cargo in patients along the Alzheimer's continuum, focusing on cognitively normal controls (NCs) with high brain ß-amyloid (Aß) loads (Aß+). METHODS: The study was based on the Sino Longitudinal Study on Cognitive Decline project. We enrolled 246 participants, including 156 NCs, 45 amnestic mild cognitive impairment (aMCI) patients, and 45 AD dementia (ADD) patients. Brain Aß loads were determined using positron emission tomography. NCs were classified into 84 Aß- NCs and 72 Aß+ NCs. Baseline plasma nEVs were isolated by immunoprecipitation with an anti-CD171 antibody. After verification, their cargos, including Aß, tau phosphorylated at threonine 181, and neurofilament light, were quantified using a single-molecule array. Concentrations of these cargos were compared among the groups, and their receiver operating characteristic (ROC) curves were constructed. A subset of participants underwent follow-up cognitive assessment and magnetic resonance imaging. The relationships of nEV cargo levels with amyloid deposition, longitudinal changes in cognition, and brain regional volume were explored using correlation analysis. Additionally, 458 subjects in the project had previously undergone plasma Aß quantification. RESULTS: Only nEV Aß was included in the subsequent analysis. We focused on Aß42 in the current study. After normalization of nEVs, the levels of Aß42 were found to increase gradually across the cognitive continuum, with the lowest in the Aß- NC group, an increase in the Aß+ NC group, a further increase in the aMCI group, and the highest in the ADD group, contributing to their diagnoses (Aß- NCs vs. Aß+ NCs, area under the ROC curve values of 0.663; vs. aMCI, 0.857; vs. ADD, 0.957). Furthermore, nEV Aß42 was significantly correlated with amyloid deposition, as well as longitudinal changes in cognition and entorhinal volume. There were no differences in plasma Aß levels among NCs, aMCI, and ADD individuals. CONCLUSIONS: Our findings suggest the potential use of plasma nEV Aß42 levels in diagnosing AD-induced cognitive impairment and Aß+ NCs. This biomarker reflects cortical amyloid deposition and predicts cognitive decline and entorhinal atrophy.