Summary: Aging is the main risk factor for neurodegenerative diseases such as Alzheimer’s and Parkinson’s, but the precise molecular changes that accumulate over time have been difficult to define. Researchers have now produced the most extensive single-cell atlas of the aging brain to date, revealing cell-type and region-specific epigenetic shifts that help explain how aging reshapes brain function.
By profiling more than one million cells from the mouse brain, the team mapped epigenetic marks—chemical tags that regulate gene activity—across 36 distinct cell types and eight brain regions. The atlas shows that aging proceeds unevenly: different cell types and brain areas age at different rates. It also highlights the reactivation of transposable elements (so-called “jumping genes”) that lose DNA methylation with age, a change that could drive cellular dysfunction and inflammation.
Key Facts
- Massive scale: The resource includes nearly 900,000 spatial transcriptomic profiles and over 200,000 single-cell methylation and chromatin-conformation measurements across eight brain regions.
- Non-neuronal vulnerability: Age-related epigenetic alterations were surprisingly stronger in non-neuronal cells—including glial populations—than in neurons.
- Reactivation of “jumping genes”: The study found that many transposable elements lose methylation with age and may become transcriptionally active, a potential source of genomic instability and inflammation.
- Spatial variability: Identical cell types show different aging signatures depending on their location—for example, non-neuronal cells in posterior brain regions display more inflammatory markers than those in anterior regions.
- New aging biomarker: Strengthening of TAD boundaries—changes in 3D genome organization—emerged as a precise biomarker of brain aging.
Source: Salk Institute
Neurodegenerative disorders affect more than 57 million people worldwide, and their global incidence is projected to approximately double every 20 years. While aging is known to be a dominant risk factor, the molecular pathways by which aging increases vulnerability to diseases such as Alzheimer’s, Parkinson’s, and ALS have not been fully resolved.
One major driver of aging biology is epigenetic change: the gradual alteration of chemical marks on DNA and chromatin that control gene expression without changing the underlying genetic code. Researchers at the Salk Institute assembled the most comprehensive single-cell atlas to date of epigenetic changes in the aging mouse brain, documenting shifts in DNA methylation, 3D genome structure, and gene expression across brain regions and cell types.
The atlas covers eight brain regions and 36 major cell types, combining over 200,000 single-cell methylomes and chromatin-conformation profiles with nearly 900,000 spatially resolved transcriptomes. These data reveal clear epigenetic differences across age groups and enabled the development of deep-learning models that predict age-related changes in gene expression.
Published in the journal Cell on March 11, 2026, the atlas is publicly hosted on cloud platforms and gene-expression repositories, providing a reference framework to interpret human brain datasets and to support large-scale initiatives such as the NIH BRAIN Initiative.
“Age-related molecular changes in brain regions responsible for attention, memory, emotion, and movement profoundly affect quality of life,” says co-corresponding author Joseph Ecker, PhD, professor at Salk and an HHMI investigator. “Mapping how the epigenome shifts across individual brain cell types gives us a molecular framework to understand how aging reshapes neural circuits and to identify mechanisms that may contribute to neurodegeneration.”
What do we know about the aging brain?
Aging is accompanied by several molecular hallmarks, including chronic inflammation, mitochondrial decline, genome instability, and epigenetic drift. Evidence increasingly points to epigenetic change—particularly DNA methylation—as a central contributor to physiological aging. Methylation changes have been linked to neuronal function, behavior, and disease, but connecting specific methylation patterns to cellular dysfunction requires high-resolution, cell-type-resolved data.
Because the brain contains many specialized regions and cell types, general bulk measurements obscure critical differences. Single-cell, multi-omic approaches are necessary to capture the diversity of aging trajectories across cells and anatomical areas.
“The brain is highly interconnected and heterogeneous; different regions control distinct functions and age at different rates even within the same cell type,” says co-corresponding author M. Margarita Behrens, PhD, research professor at Salk. “Cell type–specific maps of aging will expand opportunities for targeted therapies.”
What is an epigenetic atlas?
Bulk tissue analyses average signals across many cells and lose critical specificity. To overcome this, the team generated a multi-omic, single-nucleus atlas that integrates DNA methylation, chromatin conformation (3D genome organization), and spatial transcriptomics to preserve tissue context. The spatial dimension is especially valuable for identifying microenvironments and brain regions that are selectively vulnerable to aging.
“Spatial resolution lets us see which regions and local niches are most affected, how cell-type composition shifts over time, and how neighbors influence each other’s aging,” explains first author Qiurui Zeng. “The spatial dataset—nearly 900,000 cells—gives unprecedented statistical power for a longitudinal aging study.”
In mice, the researchers measured methylation in 132,551 single nuclei and collected joint methylation–chromatin-conformation profiles from 72,666 nuclei, representing 36 major cell types. The full dataset was released in December 2025 on public cloud and genomic repositories to ensure broad accessibility without heavy local computing requirements.
Cloud hosting places this atlas alongside other major brain resources, creating an interconnected ecosystem of openly available data that researchers in aging, neurodegeneration, and spatial genomics can use immediately to accelerate discovery.
What does the atlas reveal?
Methylation analyses showed that age-related changes were generally more pronounced in non-neuronal cells. A striking pattern was the demethylation of transposable elements in certain cell types, indicating that sequences normally kept silent can become derepressed with age. Because transposable elements make up a large portion of mammalian genomes and their activation is linked to genomic instability and inflammation, this change may contribute to age-associated decline.
Chromatin-conformation data uncovered a separate aging signature: increased strength of topologically associating domain (TAD) boundaries and greater accessibility at nearby CTCF binding sites. TADs partition the genome into regulatory neighborhoods, and changes in their boundary strength reflect alterations in 3D genome organization that could influence gene regulation during aging.
Spatial transcriptomics tied these molecular changes to anatomical context, showing that identical cell types can follow different aging paths depending on location—for example, posterior non-neuronal cells showed higher inflammatory signatures than anterior counterparts. This spatial heterogeneity highlights the need for region- and cell-level specificity when studying brain aging.
How will this atlas help science and medicine?
The dataset has already enabled the creation of machine-learning models that predict gene expression from multi-omic epigenetic features, laying the groundwork for virtual models of brain aging. Public access to the atlas allows researchers worldwide to test hypotheses, validate findings, and explore new therapeutic strategies aimed at restoring healthy epigenetic states—such as re-silencing transposable elements or correcting 3D genome organization—to protect neural circuits before irreversible damage occurs.
By providing a comprehensive, cell-resolved reference of the aging brain, this resource aims to accelerate research into mechanisms of neurodegeneration and to support the development of targeted interventions.
Other authors and funding
Other authors include Wei Tian, Anna Bartlett, Joseph Nery, Rosa Castanon, Julia Osteen, Nicholas Johnson, Wenliang Wang, Wubin Ding, Huaming Chen, Jordan Altshul, Mia Kenworthy, Cynthia Valadon, William Owens, Cindy Tatiana Báez-Becerra, Silvia Cho, Chumo Chen, Jackson Willier, Stella Cao, Jonathan Rink, Jasper Lee, Ariana Barcoma, Jessica Arzavala, and Nora Emerson of Salk; Qiurui Zeng and Amit Klein of Salk and UC San Diego; Hanqing Liu of Salk and Harvard; Zhanghao Wu of UC Berkeley; Maria Luisa Amaral, Yuru Song, and Nathan Zemke of UC San Diego; and Yuancheng Ryan Lu of Whitehead Institute.
Funding: The work was supported by the National Institutes of Health (5R01AG066018-05, RRID: SCR_014839, CCSG P30 CA014195, S10-OD023689, S10 OD034268) and the Howard Hughes Medical Institute.
Key Questions Answered:
A: No. Aging is highly localized. The atlas demonstrates that the same cell type can appear molecularly “younger” in one region and more aged in another. Aging in the brain is a mosaic rather than a uniform decline.
A: Transposable elements—often called jumping genes—form a large fraction of the genome. They are normally silenced by DNA methylation. With age, some of these elements lose methylation and can become active, promoting genomic instability and inflammation, which are implicated in neurodegeneration.
A: The atlas is an important step. By pinpointing specific epigenetic changes associated with aging, it enables experimental strategies to test whether restoring methylation patterns or 3D genome structure can rescue cellular function. While reversal of aging remains a long-term goal, this resource identifies concrete targets for intervention.
Editorial Notes:
- This article was edited by a Neuroscience News editor.
- The journal paper was reviewed in full.
- Additional context was added by the editorial staff.
About this aging and genetics research news
Author: Salk Comms
Source: Salk Institute
Contact: Salk Comms – Salk Institute
Image: The image is credited to Neuroscience News
Original Research: Open access.
“Cell-type-specific transposon demethylation and TAD remodeling in aging mouse brain” by Qiurui Zeng, Wenliang Wang, Wei Tian, Amit Klein, Anna Bartlett, Hanqing Liu, Joseph R. Nery, Rosa G. Castanon, Julia Osteen, Nicholas D. Johnson, Wubin Ding, Huaming Chen, Jordan Altshul, Mia Kenworthy, Cynthia Valadon, William Owens, Zhanghao Wu, Maria Luisa Amaral, Nathan R. Zemke, Yuru Song, Cindy Tatiana Báez-Becerra, Silvia Cho, Chumo Chen, Jackson Willier, Stella Cao, Jonathan Rink, Jasper Lee, Ariana Barcoma, Jessica Arzavala, Nora Emerson, Yuancheng Ryan Lu, Bing Ren, M. Margarita Behrens, and Joseph R. Ecker. Cell
DOI: 10.1016/j.cell.2026.02.015
Abstract
Cell-type-specific transposon demethylation and TAD remodeling in aging mouse brain
Aging is a major risk factor for neurodegenerative disease, but its epigenetic drivers are incompletely understood.
We generated a comprehensive single-nucleus atlas of brain aging across multiple regions, comprising 132,551 single-cell methylomes and 72,666 joint chromatin-conformation–methylome nuclei. Integration with companion transcriptomic and chromatin accessibility data produced a cross-modality taxonomy of 36 major cell types.
Transposable element methylation alone distinguished age groups, showing cell-type-specific, genome-wide demethylation patterns. Chromatin-conformation analysis revealed age-related increases in topologically associating domain (TAD) boundary strength with enhanced accessibility at CCCTC-binding factor (CTCF) sites.
Spatial transcriptomics across 895,296 cells exposed regional heterogeneity in aging within identical cell types. Deep-learning models built on these multi-modal features reliably predict age-related gene expression changes, providing mechanistic insight into gene regulation during aging.
Age comparisons used a 2-month baseline representing late-adolescent/early-young adult stages. This dataset advances understanding of brain aging and offers potential translational applications.