Plasma biomarkers of neocortical amyloid burden: an in-depth plasma profile using LC-MS

Background: Neocortical amyloid burden (NAB) is an important risk factor for Alzheimer’s disease (AD) that precedes the onset of clinical symptoms. It has become critical to identify individuals at early stages of NAB deposition to recruit into clinical trials of disease-modifying therapeutics. Blood-based biomarkers predicting NAB would have great utility for enrichment of AD clinical trials, including large-scale prevention trials. Despite this, only a handful of studies have investigated plasma biomarkers of NAB (Thambisetty et al; Burnham et al; Kiddle et al), with one study utilising Mass Spectrometry (MS) (Ashton et al). MS has several advantages as a discovery tool over panel-based assays however its limited sensitivity and protein coverage have been a restricting factor. Here, we present an in-depth MS-based screen of cognitively normal (CN) participants with varying degree of NAB. The objective was to identify a peripheral signature of AD pathology which precedes clinical manifestation. Methods: A proteomic workflow combining immunodepletion, isobaric peptide labelling and Isoelectric focusing (IEF) was established as a sensitive and robust strategy for in-depth plasma exploration compared with other proteomic methodologies (unpublished). Protein identification and relative quantitation was performed by MS (Orbitrap Velos). This methodology was applied to n=297 CN participants from the AIBL and KARVIAH cohorts. All participants underwent PET amyloid imaging at time of sampling with n=187 having subsequent imaging at multiple time-points. Linear Models, Discriminant Analysis and Support Vector Machines were performed to establish a cross-sectional and longitudinal proteomic model of elevated NAB. Results: A refined proteomic strategy has demonstrated a reproducible workflow that can profile >1000 plasma proteins, accurately detecting low picogram levels (unpublished). In addition to increased dynamic range, proteins indicative of the central nervous system (CNS) such as myelin basic protein and neurogranin are routinely observed. This discovery methodology has allowed the generation of a novel prediction model for both baseline and prospective NAB. Conclusions: An in-depth MS-based workflow has confirmed CNS proteins are readily measureable in the periphery, which has enabled the most detailed proteomic screen of AD pathology to date. Further replication of these putative markers will be needed although this additional evidence illustrates the potential use of a blood-based strategy in AD.