The human genome is not stored like a straight strand of text. Inside each cell nucleus, DNA folds into loops, compartments and neighborhoods that help determine which genes can be read. Two studies of postmortem human brain tissue now add spatial organization to the list of molecular systems that change with age and Alzheimer’s disease.
The work, published in Science in July and brought together in an August 25 research summary from the National Institutes of Health, does not show that disordered DNA folding causes memory loss. It does show that gene regulation in an aging brain cannot be understood only by counting mutations or measuring which genes are active. The physical arrangement of the genome is another layer of the problem.
Two maps of a changing nucleus
One team examined prefrontal-cortex tissue from 20 people aged 75 or older: 10 who had Alzheimer’s disease and 10 age-matched people who did not. Using a single-cell method that jointly measures gene activity and chromatin contacts, the researchers found that brain cells from the Alzheimer’s group had more interactions between distant parts of the genome and fewer interactions between nearby regions.
Normally, relatively active and inactive chromatin tend to occupy distinct compartments. In the Alzheimer’s samples, those compartments mixed more extensively. Greater mixing was associated with lower activity in several neuronal programs, while some genes involved in energy production and RNA processing became more active. A computational model also found that 3D-genome features improved predictions of disease-related gene-expression changes.
The second team studied hippocampal tissue from 40 neurologically typical donors distributed evenly across four age groups from 20 to 100 years. The investigators profiled gene expression, chromatin accessibility, DNA methylation and 3D contacts at single-nucleus resolution. Across most cell types, they observed weaker topologically associating domains—local chromosomal neighborhoods—and more contacts between different chromosomes in older tissue.
The hippocampus data also pointed to cell-specific changes. Microglia shifted toward a primed inflammatory state, while the researchers reported age-linked losses of astrocytes, oligodendrocyte precursor cells and endothelial cells. Reduced accessibility at binding sites for CTCF, a protein that helps maintain chromatin boundaries, tracked with the decay of those genomic neighborhoods in several cell types.
What the comparison adds
Read together, the two studies form a useful comparison rather than a single causal chain: normal aging appears to erode parts of the genome’s neighborhood map, while Alzheimer’s tissue shows a related but disease-specific reshuffling of contacts and gene activity.
That overlap matters because age is the largest risk factor for Alzheimer’s disease. It raises a sharper research question: does disease exploit an architectural vulnerability that is already emerging during ordinary aging, or do Alzheimer’s processes reorganize the nucleus by a separate route? The studies cannot answer that from postmortem snapshots, but they narrow the places where experiments can look.
The crucial distinction is between an address book and an engine. A 3D genome map can identify where regulation differs, but it does not yet show whether altered folding drives degeneration, follows cellular stress, or reflects shifts in the mix and state of cells that survive.
The Alzheimer’s comparison was small, and all donors were at least 75. The normal-aging study covered a broader adult lifespan, but it also relied on tissue collected after death. Neither design follows the same person through time. Associations between chromatin contacts and gene activity can suggest mechanisms, yet intervention studies are needed to establish which changes are causal, reversible or functionally important.
A research platform, not a treatment result
No human therapy was tested, and the findings do not validate a diagnostic scan, a biological-age score or a way to prevent dementia. Even a strong molecular association in brain tissue can sit several steps away from cognition, clinical symptoms or healthspan.
That makes the immediate value infrastructural: the studies nominate cell types, genomic neighborhoods and regulatory contacts for targeted experiments, while setting a higher bar for any future claim that reorganizing chromatin could preserve cognition.
Researchers can now perturb candidate boundary proteins or regulatory contacts in laboratory models, then test whether gene activity and cell function change in the predicted direction. Independent cohorts will also be needed to determine which 3D-genome patterns reproduce across brain regions, ancestries, sexes and disease stages.
The deeper lesson is that aging biology is layered. DNA sequence, chemical marks, accessibility, folding and gene activity interact, but they are not interchangeable measures. Mapping those layers together may reveal why some cells lose regulatory stability before others. For now, the new work is a detailed atlas of association—valuable precisely because it shows where evidence is strong and where causality remains unproven.
TENS Magazine conceptual illustration
