Back

A Multi-omic Atlas of Human Choroid Plexus in Alzheimer's Disease

Philippe, T.; Avey, D. R.; Vasco, N. K.; Vyas, H.; Tissera, S. D.; Xu, J.; Ng, B.; Saunders, D. M.; Lagrimas, A. K.; Duong, D.; Seyfried, N. T.; Bennett, D. A.; Wang, Y.

2025-12-17 genomics
10.1101/2025.11.23.690014 bioRxiv
Show abstract

The choroid plexus (CP) regulates barrier integrity, cerebrospinal-fluid (CSF) dynamics, and immune surveillance, yet its role in Alzheimers disease (AD) remains poorly defined. We performed snRNA-seq on CP samples from 69 ROSMAP participants spanning normal cognition, mild cognitive impairment, and AD dementia, and integrated these data with spatial transcriptomics, snATAC-seq, and proteomics from CP tissue and CSF. We identified 17 CP cell states and uncovered widespread disease-associated transitions that converged into three major phenotypic axes. Along the inflammatory axis, epithelial cells and border-associated macrophages (BAMs) showed progressive immune activation, with BAMs shifting from inflammatory to stress-dominant states. In the barrier axis, epithelial cells, fibroblasts, and endothelial cells exhibited reduced junction-related gene expression and broad alterations in transport pathways. Epithelial cells also showed late-stage cilia loss and CSF-regulatory pathway impairment, indicating a breakdown in epithelial polarity and CSF sensing, consistent with abnormal CSF proteomic signatures. Along the remodeling axis, fibroblasts showed ECM alterations, while epithelial and stromal cells demonstrated aberrant cell-matrix adhesion pathways. Spatial neighborhood analysis revealed a multicellular signaling hub, with epithelial-rich niches showing the strongest dysregulation in AD. Together, these findings define a unified model of CP dysfunction in AD and position the CP as an active, multicellular contributor to AD pathophysiology.

Matching journals

The top 3 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.