Omics-derived biological modules reflect metabolic brain changes in Alzheimer's disease.

Introduction: Brain glucose hypometabolism, indexed by the fluorodeoxyglucose positron emission tomography ([F]FDG-PET) imaging, is a metabolic signature of Alzheimer's disease (AD). However, the underlying biological pathways involved in these metabolic changes remain elusive.

Methods: Here, we integrated [F]FDG-PET images with blood and hippocampal transcriptomic data from cognitively unimpaired (CU, n = 445) and cognitively impaired (CI, n = 749) individuals using modular dimension reduction techniques and voxel-wise linear regression analysis.

Results: Our results showed that multiple transcriptomic modules are associated with brain [F]FDG-PET metabolism, with the top hits being a protein serine/threonine kinase activity gene cluster (peak-t = 4.86, P value < 0.001) and zinc-finger-related regulatory units (peak-t = 3.90, P value < 0.001).

Discussion: By integrating transcriptomics with PET imaging data, we identified that serine/threonine kinase activity-associated genes and zinc-finger-related regulatory units are highly associated with brain metabolic changes in AD.

Highlights: We conducted an integrated analysis of system-based transcriptomics and fluorodeoxyglucose positron emission tomography ([F]FDG-PET) at the voxel level in Alzheimer's disease (AD). The biological process of serine/threonine kinase activity was the most associated with [F]FDG-PET in the AD brain. Serine/threonine kinase activity alterations are associated with brain vulnerable regions in AD [F]FDG-PET. Zinc-finger transcription factor targets were associated with AD brain [F]FDG-PET metabolism.