CSF proteomics reveals distinct choroid plexus-related protein signatures in AD, PD and DLB

BACKGROUND: The choroid plexus (ChP) plays a role in cerebrospinal fluid (CSF) production, protein transport and clearance, and central nervous system homeostasis. Emerging evidence suggests that ChP dysfunction is implicated in the pathogenesis of neurodegenerative diseases. We aim to investigate ChP involvement in the pathophysiology of Alzheimer's disease (AD), Parkinson's disease (PD), and dementia with Lewy bodies (DLB) using CSF proteomics. METHOD: We included individuals with AD (n = 150), PD (n = 75), DLB (n = 53), and controls (n = 47) from 6 centers of the European Platform for Neurodegenerative Diseases (EPND) project. AD was defined as abnormal CSF Aβ42/40 (Roche NeuroToolKit) without meeting clinical criteria of PD or DLB. Controls had normal cognition and CSF Aβ42/40. Using the Olink Explore 3072 assay, 2902 CSF proteins were quantified. Proteins were classified as highly expressed in the ChP using the Allen Brain Atlas. Pairwise comparisons of protein concentrations between disease and controls were conducted, adjusting for age and sex. Weighted Gene Co-expression Network Analysis (WGCNA) was performed on ChP-expressed proteins to identify co-expression modules, followed by Wilcoxon tests to compare module expression profiles across groups. Pathway enrichment analysis was conducted using Gene Ontology. RESULT: Of the 2902 proteins quantified, 799 were highly expressed by the ChP. In PD versus controls, 37% of the dysregulated proteins were highly expressed in the ChP, compared to 30% in AD versus controls and 28% in DLB versus controls (Figure 1). The dysregulated ChP-expressed proteins in AD, PD and DLB were predominantly distinct and associated with different underlying pathways (Figure 2). WGCNA identified four co-expression modules related to the proteins with high ChP expression. Module 1, related to innate immunity (complement), was decreased in AD versus all other groups. Module 2, linked to transcriptional and translational processes, was increased in PD versus controls and AD. Module 3, related to the extracellular matrix, was increased in DLB versus AD. No significant differences were observed for module 4 (Figure 3). CONCLUSION: Our findings highlight the distinct involvement of ChP-expressed proteins in the pathophysiology of AD, PD, and DLB, revealing disease-specific patterns. Further research is essential to elucidate mechanistic involvement of the ChP in disease progression across these disorders.