Midlife proteome‐wide analysis identifies plasma biomarkers for 25‐year dementia risk linked to diverse pathophysiology

Abstract

AbstractBackgroundA diverse set of biological processes have been implicated in the pathophysiology of Alzheimer’s disease (AD). However, there is limited understanding of the biological pathways mechanistically relevant in the earliest phase of disease and a lack of protein biomarkers available to capture the biological systems involved.MethodTo discover early biomarkers and further understand on a molecular level the systemic factors that may promote neurodegeneration, we used a large‐scale proteomic platform to relate 4,877 plasma proteins measured in middle‐aged adults to dementia risk over a 25‐year period (n=10,981; 1,874 incident dementia cases) in the Atherosclerosis Risk in Communities (ARIC) study (Figure 1). Associated proteins were related to 20‐year cognitive decline in the Whitehall II study, and to AD risk and CSF biomarkers in the European Medical Information Framework‐AD (EMIF‐AD) study.ResultProteome‐wide associations in ARIC discovered 32 dementia‐associated plasma proteins involved in proteostasis, immunity, synaptic function, and extracellular matrix organization (Figure 2). We replicated the relationship between 15 candidate proteins and neurocognitive outcomes in Whitehall or EMIF‐AD, and demonstrated that dementia‐associated plasma proteins are associated with cerebrospinal fluid markers of Aß42, p‐tau181, neurodegeneration, and neuroinflammation (Figure 3). We found many of the protein biomarkers were abnormally expressed in AD brain tissue, though some of the strongest midlife dementia‐associated proteins (e.g., GDF‐15) were not detected in brain. Using network analyses, we found an evolving plasma protein signature for dementia risk suggesting a dysregulation of specific immune pathways and disrupted proteostasis/autophagy and extracellular matrix organization in midlife ∼20 years before dementia onset, and abnormal coagulation/complement signaling ∼10 years before dementia onset (Figure 4). We used two‐sample Mendelian randomization to identify which midlife plasma proteins and protein networks may play a causal role in Alzheimer’s disease and demonstrated that many of the dementia‐associated proteins may play a causal role in systemic (e.g., vascular, inflammatory) diseases associated with increased AD/dementia risk. We used the totality of evidence to prioritize midlife plasma proteins as AD/dementia biomarkers linked to discrete pathophysiological processes.ConclusionThe current study identified multiple pathway‐specific biomarkers and potential therapeutic targets likely relevant in the earliest phase of the AD and related dementia.