Human Brain Aging Mapped Cell by Cell from Infancy to Age 104
A landmark single-cell study reveals how gene expression and somatic mutations shift across every major brain cell type from birth to centenarian age.
Summary
Researchers at UMass Chan Medical School performed single-nucleus RNA sequencing, single-cell whole-genome sequencing, and spatial transcriptomics on prefrontal cortex tissue from 19 donors aged 0.4 to 104 years. They found that aging universally suppresses housekeeping genes involved in ribosomes, transport, and metabolism across all brain cell types, while neuron-specific genes remain largely stable. Infant brains contain unique clusters of immature neurons and astrocytes expressing developmental genes. Two age-linked mutational signatures accumulate in neurons, preferentially affecting short, highly expressed genes—the very genes whose transcription declines most with age—suggesting a feedback loop between genomic damage and transcriptional decline.
Detailed Summary
Understanding how the human brain changes across a full lifespan has been a central challenge in neuroscience and aging research. Most prior work relied on bulk tissue analysis, which obscures the distinct contributions of individual cell types. This study addressed that gap with unprecedented resolution, profiling prefrontal cortex from donors spanning infancy to 104 years of age using three complementary single-cell technologies.
Using droplet-based single-nucleus RNA sequencing, the team analyzed 367,317 nuclei across 19 donors, identifying 31 distinct clusters encompassing excitatory neurons from multiple cortical layers, four inhibitory neuron subtypes, microglia, oligodendrocytes, OPCs, astrocytes, and endothelial cells. Spatial transcriptomics via MERFISH validated these findings with single-molecule resolution in tissue sections, confirming correct laminar organization of neurons even in infant brains. Infant-specific clusters of neurons and astrocytes expressed neurodevelopmental genes—including SLIT3, ROBO1, HES5, and ID4—that disappear in post-infancy samples, marking a transition from developmental to mature cell states.
A striking pan-cell-type finding was the age-associated downregulation of housekeeping genes involved in ribosome function, intracellular transport, and metabolism. This decline was observed across all major cell types and correlated with increasing donor age. In contrast, neuron-identity genes remained remarkably stable throughout life, suggesting that neurons preserve their functional identity even as general cellular maintenance deteriorates. Transcriptional variability also increased in elderly inhibitory neurons (particularly IN-SST subtype), hinting at growing heterogeneity and potential vulnerability in old age. OPC abundance declined with age while mature oligodendrocytes increased, consistent with ongoing myelination dynamics across the lifespan.
On the genomic side, single-cell whole-genome sequencing of 100 sorted neuronal nuclei revealed two distinct age-associated mutational signatures. One signature correlated with active gene transcription and the other with gene repression. Critically, somatic mutation rates in neurons were gene-length- and expression-level-dependent: shorter, more highly expressed genes—which are enriched among the housekeeping genes showing transcriptional decline—accumulated mutations at higher rates. This creates a plausible mechanistic link: somatic mutations accumulate preferentially in the most active housekeeping genes, potentially impairing their transcription and contributing to the observed age-related decline in essential cellular maintenance functions.
These findings provide a detailed, cell-type-resolved atlas of human brain aging and development, identifying molecular signatures that bridge healthy aging and potential disease susceptibility. The convergence of transcriptomic decline in housekeeping genes with targeted genomic damage in those same genes is a particularly compelling observation with implications for understanding neurodegeneration and cognitive decline in later life.
Key Findings
- Housekeeping genes for ribosomes, transport, and metabolism decline universally across all brain cell types with age.
- Neuron-specific identity genes remain stable throughout the entire human lifespan from infancy to age 104.
- Infant brains contain unique clusters of immature neurons and astrocytes expressing developmental genes absent in older donors.
- Two somatic mutational signatures accumulate in aging neurons, targeting short, highly expressed genes preferentially.
- OPC abundance decreases with age while mature oligodendrocytes increase, reflecting ongoing myelination dynamics.
Methodology
The study used droplet-based snRNA-seq (367,317 nuclei), scWGS (100 sorted neuronal nuclei), and MERFISH spatial transcriptomics on fresh-frozen human prefrontal cortex from 19 neurotypical donors aged 0.4–104 years. Cell types were annotated against a published human PFC reference dataset, and differential expression and mutational analyses were performed across age groups.
Study Limitations
The study includes only 19 donors, limiting statistical power for some cell-type-specific analyses. scWGS was performed on only 100 sorted neurons total, and the causal direction between somatic mutations and transcriptional decline remains unproven. The cohort lacks demographic diversity and excludes donors with neuropathological diagnoses, which may limit generalizability.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.