Concordant peripheral lipidome signatures in two large clinical studies of Alzheimer's disease

• Published in: Nature communications Vol. 11; no. 1; pp. 5698 - 11
• Main Authors: Huynh, Kevin, Lim, Wei Ling Florence, Giles, Corey, Jayawardana, Kaushala S, Salim, Agus, Mellett, Natalie A, Smith, Adam Alexander T, Olshansky, Gavriel, Drew, Brian G, Chatterjee, Pratishtha, Martins, Ian, Laws, Simon M, Bush, Ashley I, Rowe, Christopher C, Villemagne, Victor L, Ames, David, Masters, Colin L, Arnold, Matthias, Nho, Kwangsik, Saykin, Andrew J, Baillie, Rebecca, Han, Xianlin, Kaddurah-Daouk, Rima, Martins, Ralph N, Meikle, Peter J
• Format: Journal Article
• Language: English
• Published: England Nature Publishing Group 10.11.2020; Nature Publishing Group UK; Nature Portfolio
• Subjects: Alzheimer Disease - metabolism; Alzheimer's disease; Biomarkers - blood; Cohort Studies; Computer Simulation; Gangliosides; Humans; Lipid Metabolism; Lipidomics; Lipids; Lipids - blood; Metabolism; Metabolomics; Neurodegenerative diseases; Pathology; Phosphatidylethanolamine; Signatures; Sphingolipids; Triglycerides
• ISSN: 2041-1723; 2041-1723
• DOI: 10.1038/s41467-020-19473-7

Changes to lipid metabolism are tightly associated with the onset and pathology of Alzheimer's disease (AD). Lipids are complex molecules comprising many isomeric and isobaric species, necessitating detailed analysis to enable interpretation of biological significance. Our expanded targeted lipidomics platform (569 species across 32 classes) allows for detailed lipid separation and characterisation. In this study we examined peripheral samples of two cohorts (AIBL, n = 1112 and ADNI, n = 800). We are able to identify concordant peripheral signatures associated with prevalent AD arising from lipid pathways including; ether lipids, sphingolipids (notably GM gangliosides) and lipid classes previously associated with cardiometabolic disease (phosphatidylethanolamine and triglycerides). We subsequently identified similar lipid signatures in both cohorts with future disease. Lastly, we developed multivariate lipid models that improved classification and prediction. Our results provide a holistic view between the lipidome and AD using a comprehensive approach, providing targets for further mechanistic investigation.