Human brain cell-type-specific aging clocks based on single-nuclei transcriptomics

Abstract

Aging is the primary risk factor for most neurodegenerative diseases, yet the cell-type-specific progression of brain aging remains poorly understood. Here, we developed human cell-type-specific transcriptomic aging clocks using high-quality single-nucleus RNA sequencing data from post mortem human prefrontal cortex tissue of 31 donors aged 18 - 94 years, encompassing 73,941 high-quality nuclei. We observed distinct transcriptomic changes across major cell types, including upregulation of inflammatory response genes in microglia from older samples. Aging clocks trained on each major cell type accurately predicted chronological age and remained robust in independent single-nucleus RNA-sequencing datasets, underscoring their broad applicability. These findings demonstrate the feasibility of cell-type-specific transcriptomic clocks to measure biological aging in the human brain and highlight potential mechanisms of selective vulnerability in neurodegenerative diseases. We anticipate these clocks will serve as a basis for further studies in other brain regions and more diverse populations, ultimately advancing our understanding of age-related neurodegenerative processes at the single-cell level.