Disentangling amyloid polymorphs in normal aging and Alzheimer's disease using dual-probe spectral imaging.

Variability in Alzheimer's disease (AD) clinical presentation complicates mechanistic studies and therapeutic outcome prediction. Brain protein aggregate load does not directly correlate with clinical symptoms; however, different subtypes of AD have been reported to exhibit structural variation (polymorphism) of aggregates. Little is known about the structural diversity of the deposits in cognitively normal aged brains. This study investigates the structural heterogeneity of amyloid aggregates in the hippocampus and their association with age- and disease-related pathology. Post-mortem hippocampal tissue from cognitively normal aged controls and AD patients was co-stained with the amyloid-sensitive dyes BSB and MCAAD-3 and imaged across various subregions using spectral fluorescence microscopy. Machine learning analysis of spectral data differentiated amyloid polymorphs between cognitively normal and Alzheimer's cases. Our analysis revealed distinct spectral features across the amyloid plaques, neurofibrillary tangles and the background tissue parenchyma associated with AD compared to those observed in cognitively normal aging, irrespective of overall aggregate load. This study underscores the importance of amyloid polymorphism in determining the clinical impact of protein pathology in AD. Our findings highlight that focusing on amyloid structure rather than total load can aid in advancing personalized approaches in the diagnosis and treatment of neurodegenerative diseases.