Unbiased data-driven analysis of five amyloid-beta peptides for biomarker investigations in familial Alzheimer’s disease

Abstract

INTRODUCTION: Changes to the relative abundance of amyloid-beta (Abeta) peptides are hallmarks of Alzheimer's disease (AD). iPSC-derived neurons offer a physiological model of Abeta production. We employed unbiased, data-driven analyses to investigate combinations of Abeta peptides as AD biomarkers and the relative contribution of peptides to AD pathogenesis. METHODS: We measured Abeta;37, Abeta;38, Abeta;40, Abeta;42 and Abeta;43 in ten iPSC-neuronal cultures from PSEN1 mutation carriers. We combined these data with published cell model data and used linear weighted combinations to 1) distinguish AD from controls, and 2) predict age-at-onset for PSEN1 mutations. RESULTS: Data-driven approaches distinguished Abeta;42 and Abeta;43 from shorter peptides, providing unbiased evidence for their contribution to disease. Weighted linear combinations of Abeta peptides outperform Abeta;42/40 and provide insights into relative peptide contribution as biomarkers; the optimal ratio for all data is represented as 21 . Abeta37 + 10. Abeta38+ 69 . Abeta40/(94 . Abeta42 + 6 . Abeta43). DISCUSSION: The algorithm discovered herein can be further refined to improve biomarkers for AD.