Integrative transcriptomic analysis identifies meningeal-hippocampal immune communication in Alzheimer's disease

Alzheimer's disease (AD) exhibits spatial heterogeneity, yet the mechanisms by which immune-active meninges communicate with vulnerable brain regions during disease progression remain unclear. This study aimed to investigate meninges-hippocampus crosstalk and map the communication patterns during AD pathology development.

We employed 5xFAD transgenic and PS19 tau-pathology mouse models and integrated bulk RNA sequencing, proteomics, and spatial transcriptomics across time-course analyses to distinguish AD pathology from normal aging processes. Transcription factor activity inference was coupled with ligand-receptor dynamics analysis to examine regulatory modules in homeostatic and disease-associated microglia.

We identified structure-specific disease genes, including a robust 19-gene hippocampal signature that was reproducible across an independent 5xFAD dataset and PS19 tau-pathology dataset. Spatial analysis defined a disease progression axis that transferred to the independent spatial dataset, where meningeal spots were enriched in advanced disease progression layers and disease modules showed consistent gradients. Ligand-receptor and microglial-state LRaxis analyses further supported candidate communication axes associated with pathology gradients and state-specific regulatory modules in both homeostatic and disease-associated microglia.

Our spatial framework supports the meninges as a candidate modulatory interface associated with hippocampal pathology in AD mouse models. These mouse-derived, computationally inferred ligand-receptor and regulatory modules should be considered hypothesis-generating candidates that require validation in human AD tissues and functional perturbation studies before translational relevance can be inferred.