Summary: A large-scale single-cell study reveals that Alzheimer’s disease involves widespread erosion of epigenomic control — the mechanisms brain cells use to keep their gene expression stable. By building a multimodal atlas of 3.5 million cells from multiple brain regions, researchers show that vulnerable neurons and supporting cells in memory-critical areas such as the hippocampus and entorhinal cortex lose nuclear compartmentalization and key epigenomic information as disease progresses. This breakdown in genomic regulation is linked directly to increased activity of disease-associated genes and to cognitive decline, suggesting new directions for therapies that preserve epigenomic stability.
The study reframes Alzheimer’s beyond plaques and tangles: it identifies fundamental failures in how the genome is organized and regulated inside individual brain cells. These failures, the researchers argue, may underlie the loss of cell identity and function that leads to dementia.
Key Facts
- Massive single-cell atlas: Combined transcriptomic and epigenomic data from 3.5 million cells across six brain regions, collected from 384 postmortem samples spanning 111 donors.
- Epigenomic erosion: Cognitive decline correlated with loss of nuclear compartmentalization and a decline in cell-specific epigenomic information rather than solely with plaque burden.
- Therapeutic implications: Stabilizing epigenomic programs or the factors that protect chromatin structure may help preserve neuronal function and cognitive resilience.
Source: Picower Institute at MIT
Most public discussion of Alzheimer’s focuses on symptoms such as memory loss and on pathological markers like amyloid plaques and tau tangles. However, the new study, published in Cell, emphasizes a different but complementary view: Alzheimer’s is a disease of genome regulation inside the nucleus of brain cells. To map these changes at high resolution, the team combined single-cell RNA sequencing (to read which genes are active) with ATAC-seq profiling (to show which regions of chromatin are accessible), producing a multimodal atlas that links gene expression programs to the underlying epigenomic regulatory landscape.

The researchers produced a comprehensive map of more than one million candidate cis-regulatory elements and organized these into regulatory modules across 67 cell subtypes, including many classes of excitatory and inhibitory neurons as well as glial cells. By comparing samples from individuals without pathology to those with early and late-stage Alzheimer’s, and by relating molecular patterns to each donor’s clinical history, the team identified two dominant epigenomic trends that track with disease progression:
- Breakdown of nuclear compartments: Regions of the genome that should remain tightly repressed became more accessible, while some normally open regions became repressed, reflecting a loss of the compartmental systems that segregate active and inactive chromatin.
- Loss of epigenomic information: Cells lost the distinct pattern of regulatory marks that define their identity and functional programs, making them more likely to adopt disease-associated expression profiles.
These changes were not uniform: they were strongest in regions affected earliest in Alzheimer’s, notably the entorhinal cortex and hippocampus, and they varied by cell type. Microglia (the brain’s immune cells), oligodendrocytes (myelin-producing cells), and specific excitatory neuron subtypes showed pronounced vulnerability. Where cells preserved compartmental order and retained epigenomic information, donors tended to retain cognitive function — a molecular signature of resilience.
To quantify epigenomic integrity, the team developed an epigenomic information score for individual cells. On average this score declined with pathology, but resilient individuals and resilient cell populations maintained higher scores, correlating with preserved cognition. The atlas also traced how epigenomic remodeling corresponded to cell-state transitions, changes in cell-type proportions, and coordinated dysregulation of gene networks linked to inflammation, oxidative stress, and synaptic connectivity.
Risk genes and chromatin ‘guardians’
The study sheds light on how well-known genetic risk factors interact with epigenomic stability. For example, microglia from brains carrying the APOE e4 variant initially increased epigenomic information — possibly a compensatory response — but later displayed a sharp decline as disease progressed, especially in individuals with two APOE e4 copies. This pattern suggests APOE4 may predispose microglia to genomic destabilization and functional burnout.
Neurons that express RELN (Reelin) in the entorhinal cortex and hippocampus were confirmed as pivotal: these cells show early and severe loss of epigenomic information in progressing disease but remain protective when their epigenomic programs are preserved. The authors also identified sets of genes they term “chromatin guardians” whose expression helps maintain proper chromatin states; loss of these factors coincides with increased accessibility of regions normally silenced by Polycomb repression and other silencers, and with the upregulation of inflammatory and stress-response pathways.
“Alzheimer’s is not only about plaques and tangles, but about the erosion of nuclear order itself,” said Manolis Kellis, senior author. The findings point to a model in which cognitive decline emerges when chromatin guardians and epigenomic memory marks are lost, allowing cells to shift from resilience toward vulnerability at the deepest level of genome regulation.
Lead authors of the study are Zunpeng Liu and Shanshan Zhang. Additional contributors include Benjamin T. James, Kyriaki Galani, Riley J. Mangan, Stuart Benjamin Fass, Chuqian Liang, Manoj M. Wagle, Carles A. Boix, Yosuke Tanigawa, and many others. Funding came from the National Institutes of Health, the National Science Foundation, the Cure Alzheimer’s Fund, and multiple foundations and donors.
About this genetics and Alzheimer’s disease research news
Author: David Orenstein
Source: Picower Institute at MIT
Contact: David Orenstein – Picower Institute at MIT
Image: Credit to Neuroscience News
Original Research: Open access. “Single-cell multiregion epigenomic rewiring in Alzheimer’s disease progression and cognitive resilience” by Manolis Kellis et al., Cell. DOI: 10.1016/j.cell.2025.06.031
Abstract
Single-cell multiregion epigenomic rewiring in Alzheimer’s disease progression and cognitive resilience
Alzheimer’s disease (AD) causes progressive cognitive decline, but its epigenetic drivers have been incompletely characterized. This study integrates single-cell epigenomic and transcriptomic profiles for 3.5 million cells from 384 postmortem samples across six brain regions in 111 individuals with and without AD. The authors identify over one million candidate cis-regulatory elements and organize them into regulatory modules across 67 cell subtypes. They define large-scale epigenomic compartments and a single-cell epigenomic information metric, revealing widespread epigenome relaxation and region- and cell-type-specific erosion signatures during AD progression. These epigenomic stability dynamics associate closely with changes in cell-type proportions, glial-state transitions, and coordinated dysregulation tied to AD pathology, cognitive impairment, and resilience. The atlas provides a comprehensive resource to advance understanding of AD mechanisms and potential therapeutic strategies focused on preserving epigenomic control.