Unraveling Aberrant Metabolic Patterns in Alzheimer's Disease Subtypes: From Perturbed Metabolic Pathways to Candidate Drug Targets.

Alzheimer's Disease (AD) is characterized by multiple metabolic abnormalities that differ between individuals. In this study, we aimed to elucidate metabolic dysregulations associated with five distinct molecular subtypes of AD, defined in a recent report, based on brain transcriptome data from two autopsy-derived datasets with approximately 550 AD patients. We mapped transcriptome data from each individual on a human genome-scale metabolic network to create personalized metabolic models. We identified aberrant metabolic patterns common across AD subtypes as well as those specific to each AD subtype. More than 20 metabolic pathways were found to be dysregulated in both datasets, with a majority of these pathways unidentified by a conventional subtype-agnostic approach. Among these, fatty acid metabolism was found to be dysregulated commonly in all subtypes while inositol phosphate metabolism and nucleotide metabolism were among the subtype-specific perturbed pathways. These results were independently validated through analyses of metabolome data available in one of the datasets. Moreover, upregulated secretion of pregnenolone and N-acetylneuraminate and therapeutic targeting of ALDH18A1 and SLC6A12 were among the subtype-specific candidate metabolite biomarkers and drug targets, respectively, predicted by our metabolic modeling approach, with available literature support. Subtype-specific metabolic abnormalities catalogued in our study may provide novel insights for understanding biological mechanisms implicated in AD pathogenesis as well as precision-guided treatments targeting dysregulated metabolism in AD.