A compilation of reported alterations in the cerebrospinal fluid proteome in Alzheimer's disease

Alzheimer's disease is a multifaceted neurodegenerative disorder, with diverse underlying pathophysiological processes extending beyond amyloid-β and tau accumulation. The heterogeneity of Alzheimer's disease necessitates the identification of a broad array of biomarkers that capture the diverse mechanisms contributing to disease onset and progression. In this study, we systematically compiled and analysed cerebrospinal fluid proteomics data from omics studies utilizing mass spectrometry, Olink, or SomaScan platforms. Systematic literature searches for each platform revealed a total of 264 studies. From this, a set of 18 studies were selected based on sample size, number of markers analysed, and open data availability. We found a total of 1,448 differentially expressed proteins between Alzheimer's disease and amyloid negative controls across these datasets, with 635 being found in more than one study. A 'top' set of 61 differentially expressed proteins were consistently reported in at least six studies. Clustering and functional enrichment analysis of the top differentially expressed proteins indicated involvement in metabolic regulation, glutathione metabolism and proteins of the 14-3-3 family, reflecting importance of reactive oxygen species (ROS) response. Synaptic signalling processes were found to generally be downregulated. We further integrated the top differentially expressed proteins with results from a study on familial Alzheimer's disease cerebrospinal fluid to assess at which stage of disease progression these proteins change, highlighting markers shared between sporadic and familial Alzheimer's disease datasets. Lastly, we examine the overlap of the top differentially expressed proteins between cerebrospinal fluid and brain tissue using a publicly available database. This analysis provides a comprehensive overview of the Alzheimer's disease cerebrospinal fluid proteomic landscape, indicating changes in key pathways and cellular processes associated with Alzheimer's disease pathology. By integrating data from different platforms, we highlight reproducible protein changes that may serve as promising candidates for further biomarker research aimed at improving patient stratification, tracking disease progression, and assessing therapeutic interventions.