Alzheimer's Disease Cerebrospinal Fluid Proteome Dataset

The Alzheimer's Disease Cerebrospinal Fluid (AD-CSF) Proteome and Longitudinally Paired Mild Cognitive Impairment (MCI) Drug Trial Participants dataset maps the proteomic landscape of cerebrospinal fluid changes associated with amyloid-β (Aβ) and tau pathology in 300 individuals.  

Alzheimer’s disease is currently defined at the research level by the aggregation of Aβ and tau proteins in the brain. While biofluid biomarkers are available to measure Aβ and tau pathology, few biomarkers are available to measure the complex pathophysiology associated with these two neuropathologies. AD biofluid proteomics holds promise for the development of biomarkers that reflect diverse pathologies for use in clinical trials and precision medicine. 

Researchers at Emory's Goizueta Alzheimer's Disease Research Center and Emory Healthy Brain Study generated the dataset using two independent proteomic platforms, then harmonized the results into a single protein co-expression network. 

This dataset is suited to research on CSF biomarkers of AD pathophysiology beyond Aβ and tau, including modules linked to APOE ε4 genotype (oxidant detoxification, MAPK signaling, neddylation, and mitochondrial biology) and a glycolysis-linked module tied to cognitive function and treatment response. Because case-control status is biomarker-confirmed rather than clinically assigned, the dataset supports rigorous case-control proteomic comparisons as well as work on AD heterogeneity that a binary Aβ/tau framework doesn't resolve. The paired atomoxetine trial samples also make it a rare resource for connecting CSF proteomic subtypes to drug response in MCI populations.

Harmonization and integration of both data types allowed for generation of a robust protein co-expression network consisting of 34 modules derived from 5242 protein measurements, including disease-relevant modules associated with autophagy, ubiquitination, endocytosis, and glycolysis. 

Three modules strongly associated with the apolipoprotein E ε4 (APOE ε4) AD risk genotype mapped to oxidant detoxification, mitogen associated protein kinase (MAPK) signaling, neddylation, and mitochondrial biology, and overlapped with a previously described lipoprotein module in serum. Neddylation and oxidant detoxification/MAPK signaling modules had a negative association with APOE ε4 whereas the mitochondrion module had a positive association with APOE ε4. 

The directions of association were consistent between CSF and blood in two independent longitudinal cohorts, and altered levels of all three modules in blood were associated with dementia over 20 years prior to diagnosis. Dual-proteomic platform analysis of CSF samples from an AD phase 2 clinical trial of atomoxetine (ATX) demonstrated that abnormal elevations in the glycolysis CSF module—the network module most strongly correlated to cognitive function—were reduced by ATX treatment. 

Individuals who had more severe glycolytic changes at baseline responded better to ATX. Clustering of individuals based on their CSF proteomic network profiles revealed ten groups that did not cleanly stratify by Aβ and tau status, underscoring the heterogeneity of pathological changes not fully reflected by Aβ and tau.