Cerebrospinal fluid proteomic profiling of cognitively unimpaired individuals with suspected non-Alzheimer's disease pathophysiology

Suspected non-Alzheimer's disease pathophysiology (SNAP) is a biomarker-based concept describing individuals with abnormal tau and/or neurodegeneration markers but normal amyloid levels. SNAP is common in individuals with normal cognition (NC), but its underlying pathophysiology is understudied, while being relevant for clinical trial design and treatment approaches. We aimed to investigate the pathophysiology of individuals with NC who are amyloid-negative and tau-positive (SNAP) through cerebrospinal fluid (CSF) proteomics. Two hundred and ninety-one individuals with NC were classified based on CSF amyloid β1-42 and phosphorylated tau 181, as amyloid-negative/tau-negative (controls), amyloid-negative/tau-positive (SNAP), amyloid-positive/tau-negative and amyloid-positive/tau-positive. We measured 3102 proteins in CSF using tandem mass tag proteomic analyses. We compared protein abundance between groups using analysis of covariance and identified enriched biological pathways using Gene Ontology. We also examined differences between groups in genetic risk for Alzheimer’s disease, estimated using polygenic risk scores based on genome-wide association study data. SNAP individuals with NC showed mostly increased protein levels (n = 360) compared with controls, mainly associated with neuroplasticity, angiogenesis, and protein modification and degradation. The proteomic profile of SNAP was similar to that of amyloid-positive/tau-positive individuals, while distinct from amyloid-positive/tau-negative individuals, who showed mainly decreased proteins associated with neuroplasticity. Higher levels of amyloid β1-40 and amyloid β1-42 were observed in SNAP compared with the three other groups. Polygenic risk scores analyses showed no significant differences between SNAP, amyloid-positive/tau-negative, and amyloid-positive/tau-positive individuals, while SNAP showed some genetic differences from controls, which were driven by APOE. Individuals with NC and SNAP or amyloid-positive/tau-positive status showed similar CSF proteomic profiles, while amyloid-positive/tau-negative individuals showed a distinct CSF proteomic profile. This suggests that tau, rather than amyloid, might be the main driver of the proteomic profiles in SNAP and other amyloid/tau subgroups. This may have implications for future proteomic studies and clinical trial design, as these findings highlight the importance of considering tau status in future studies.