Basic Science and Pathogenesis.

Area of Science:

• Neuroscience
• Computational Biology
• Genomics

Background:

• Alzheimer's disease (AD) pathology spreads across the brain, but the underlying mechanisms remain unclear.
• Spatial patterns of amyloid and tau pathology emergence are known, yet the molecular basis for this propagation is elusive.
• This study integrates transcriptomic data with molecular imaging to model AD pathology spread.

Purpose of the Study:

• To unveil the molecular basis of amyloid and tau propagation networks in the brain.
• To develop a graph network model integrating transcriptomic and imaging data for AD pathology.
• To identify genetic and pathway contributions to inter-regional AD pathology spread.

Main Methods:

• Developed a graph attention network model using DTI, PET scans (amyloid and tau), and microarray data.
• Trained the model on data from the Human Connectome Project (HCP) and Alzheimer's Disease Neuroimaging Initiative (ADNI).
• Utilized Allen brain atlas transcriptomic data and performed Over Representation Analysis (ORA) to identify key pathways.

Main Results:

• The model accurately predicted regional pathology based on neighboring regions (MSE <0.01 for amyloid, <0.1 for tau).
• Key genes like APOE, B2M, and HLA were identified as associated with AD pathology.
• Synaptic transmission and calcium channel pathways facilitated tau propagation, while innate immune pathways were protective against both tau and amyloid spread.

Conclusions:

• A novel framework was developed for integrating microscopic and macroscopic data in neurodegeneration research.
• Constitutive expression of innate immune molecules appears protective against Alzheimer's pathology propagation.
• Findings support a protective role for neuroinflammation in the early stages of Alzheimer's disease.