The study, led by Nadia Lunardi, MD, PhD, chief of UVA Health’s neuroanesthesia division, and Hari Prasad Osuru, PhD, examined how anesthesia, surgery and the physiological strain associated with intensive care affect the brains of aged mice. Together, the researchers refer to these exposures as anesthesia, surgery and intensive-care stress, or ASI. The experimental model was designed to reproduce several features of the perioperative experience that can challenge an older brain, including anesthesia exposure, tissue injury, inflammation and the disruption of normal sleep and daily routines that often occurs in critical care.

The team found that ASI produced a broad molecular imprint in the brains of the older animals. Rather than simply causing temporary changes in neuronal activity, the combined stressors altered gene regulation through epigenetic mechanisms. Epigenetics describes a layer of biological control that determines how strongly genes are expressed without changing the underlying DNA sequence. Chemical modifications to DNA-associated proteins can loosen or tighten chromatin, the complex structure that packages genetic material, making particular genes more or less accessible to the cellular machinery that reads them.

In the aged mice, these epigenetic changes affected genes involved in memory formation, neuronal structure and the circadian system—the internal biological clock that helps coordinate sleep, wakefulness, hormone release and other daily functions. The disruption was associated with impaired learning and memory, altered sleep patterns and behavioral abnormalities resembling aspects of postoperative delirium. The findings suggest that perioperative stress may interfere with the brain’s ability to adapt to new information at precisely the time when older neurons are already less resilient.

Memory formation depends on tightly coordinated changes in neuronal connections. When an animal learns, networks of neurons modify the strength and structure of their synapses, allowing information to be encoded and later retrieved. The UVA researchers observed evidence that ASI compromised this structural and functional plasticity. At the same time, genes that help synchronize the circadian rhythm were disturbed. Because sleep supports memory consolidation, waste clearance and metabolic recovery in the brain, a breakdown in the biological clock could amplify cognitive injury rather than merely accompany it.

The investigators then tested vorinostat, an FDA-approved drug currently used to treat certain types of T-cell lymphoma. Vorinostat belongs to a class of compounds known as histone deacetylase inhibitors. Histone deacetylases remove chemical groups from histone proteins, which help package DNA, and can thereby influence whether genes are active or silent. By inhibiting these enzymes, vorinostat can reshape chromatin accessibility and restore expression of genes that have been suppressed. The drug is not approved as a treatment for delirium or dementia, and its effects in this study should not be interpreted as evidence that patients should take it around the time of surgery.

In the mouse experiments, animals given vorinostat before exposure to anesthesia, surgery and intensive-care-related stress showed healthier neuronal structure and performed better on cognitive tests than untreated animals exposed to the same stressors. The drug also restored activity in gene networks associated with memory and circadian regulation. Their sleep patterns improved, and they displayed fewer behavioral features interpreted by the researchers as delirium-like symptoms. These results indicate that at least some of the brain changes induced by perioperative stress are not permanent. Instead, they may represent reversible alterations in gene regulation that can be targeted pharmacologically.

The discovery offers a possible biological explanation for why delirium and later cognitive decline can be linked without proving that delirium directly causes dementia in every patient. An episode of delirium may reveal an underlying vulnerability in the aging brain, while the molecular disruptions produced by surgery and critical illness could further reduce cognitive reserve. Epigenetic changes affecting synaptic plasticity and circadian control provide a plausible bridge between a short-term disturbance in attention and longer-lasting impairment in memory. The study also supports the idea that preventing or promptly correcting delirium-related biological changes could reduce the risk of prolonged cognitive problems.

Postoperative delirium affects millions of older adults worldwide and is associated with longer hospital stays, more complications and increased healthcare costs. The American Delirium Society estimates that approximately 7 million hospitalized Americans experience the condition each year, with annual costs reaching as high as $152 billion. Yet treatment remains largely supportive, relying on measures such as reorientation, sleep protection, early movement, careful medication management and correction of pain or metabolic abnormalities. A therapy designed to modify the molecular consequences of perioperative stress would represent a fundamentally different approach, although substantial testing would be required before vorinostat or related drugs could be considered for human prevention.

Lunardi and Osuru emphasize that the work is an early-stage discovery from an aged-animal model, not a clinical trial. Vorinostat can produce serious side effects, including blood abnormalities, gastrointestinal problems and effects on the heart, making indiscriminate use inappropriate. The researchers are now applying single-cell technologies and spatial transcriptomics to determine which brain cell populations respond most strongly to perioperative stress. These methods can reveal whether neurons, astrocytes, microglia or other cells drive the epigenetic changes, and where in the brain the molecular disruptions are concentrated. Such information could guide safer, more precise treatments aimed at preserving cognition after surgery.

The findings, published in Alzheimer’s & Dementia, identify epigenetic regulation as a promising target in the effort to understand postoperative delirium and its possible relationship to dementia. If future studies confirm the mechanism and establish a safe therapeutic window, drugs that selectively normalize gene activity could one day complement non-drug strategies for protecting vulnerable patients. For now, the results provide an important shift in perspective: the confusion that follows surgery may not be only a transient reaction to anesthesia or hospitalization, but a measurable biological event that leaves the aging brain temporarily—and potentially treatably—reprogrammed.

Subject of Research: Postoperative delirium, perioperative stress, epigenetic regulation, cognitive decline and dementia risk in aging brains.

Web References: https://doi.org/10.1002/alz.71556

References: Lunardi N, Osuru HP and colleagues, findings published in Alzheimer’s & Dementia, DOI: 10.1002/alz.71556.

Keywords: Postoperative delirium, dementia, cognitive decline, aging brain, anesthesia, surgery, intensive care, epigenetics, vorinostat, histone deacetylase inhibitors, memory, circadian rhythm, sleep disruption, neuroanesthesiology.

Aging is accompanied by a cascade of molecular changes, notably including chronic inflammation, mitochondrial dysfunction, genome instability, and profound epigenetic shifts. Epigenetics refers to chemical modifications superimposed on the genome that regulate gene expression without altering the underlying DNA sequence. Among these modifications, DNA methylation has emerged as a key modulator of neuronal function and aging. The Salk team harnessed state-of-the-art single-cell technologies to chart the epigenetic landscape of eight brain regions, profiling over 200,000 cells through assays mapping DNA methylation and chromatin conformation—a measure of the genome’s 3D folding structure. Complementing this, nearly 900,000 cells were spatially profiled via transcriptomics, providing unprecedented resolution into gene expression patterns while preserving tissue architecture.

This newly published atlas in the journal Cell meticulously catalogs epigenetic changes that distinguish younger from aged mouse brains in a cell type-specific manner. It captures 36 distinct brain cell types, ranging from neuronal subtypes to various glial populations. Intriguingly, the data reveal that age-related DNA methylation changes are considerably more pronounced in non-neuronal cells, such as microglia and astrocytes, lymphocytes of the brain’s immune system. This discovery aligns with emerging evidence pointing to glial cells’ pivotal roles in brain aging and neuroinflammation, broadening therapeutic interest beyond neurons alone.

The atlas also identifies transposable elements—commonly known as jumping genes—as major epigenetic hotspots during aging. These repetitive DNA sequences, which comprise roughly half of mammalian genomes, normally remain epigenetically silenced to maintain genomic stability. However, the research reveals significant demethylation of these elements in aged cells, implicating their reactivation as a contributor to cellular dysfunction. This insight lays vital groundwork for considering how loss of transposable element suppression might drive age-associated neurodegenerative pathology.

Beyond methylation, the study integrates chromatin conformation data, spotlighting the spatial organization of the genome. The architecture is partitioned into topologically associating domains (TADs)—subregions within chromosomes that regulate gene expression through physical proximity in 3D space. Remarkably, the Salk team discovered increased TAD boundary strength and enhanced accessibility of CTCF binding sites—CTCF being a critical architectural protein—as biomarkers of brain aging. These changes point to a rewiring of genomic topology that may destabilize transcriptional programs and accelerate age-related decline.

Capturing nearly one million spatially resolved transcriptomes was a tour de force in scale and resolution, enabling the researchers to dissect regional heterogeneity in aging trajectories. An unexpected revelation is the differential aging rates of identical cell types based on their brain location. For instance, glial cells in posterior brain regions exhibited heightened inflammatory gene expression compared to their anterior counterparts. This spatial variance underscores the intricate interplay of local microenvironments with cellular aging, emphasizing that neurodegeneration does not unfold uniformly across the brain.

The methodological innovation of combining multi-omics with spatial transcriptomics represents a significant leap forward in aging research. The data’s spatial dimension preserves the anatomical context, allowing scientists to link molecular alterations to specific neural circuits and brain functions affected in disease. Moreover, by publicly releasing this extensive dataset through Amazon Web Services and the Gene Expression Omnibus, the researchers have democratized access, enabling a global community to probe these aging-related molecular changes without prohibitive computational infrastructure.

The implications for translational science are profound. Using the atlas, the research team developed deep-learning models that predict the trajectory of gene expression changes based on epigenetic signatures. These virtual models could ultimately simulate brain aging, providing a powerful platform for testing interventions before clinical trials. By elucidating precise molecular targets associated with aging and neurodegeneration, this resource opens new avenues for therapeutic development aiming to halt or even reverse brain aging processes.

Senior investigators Joseph Ecker and Margarita Behrens articulate that this cell type-specific epigenetic map delivers a crucial framework for unraveling how aging reshapes neural circuits at the molecular level. As aging profoundly impairs cognition, memory, motor functions, and emotional regulation, deep molecular insight is required to counteract these effects. This atlas paves the way to decipher the mechanisms underlying age-associated neuronal vulnerability and resilience, potentially transforming our approach to widespread neurodegenerative diseases.

In summary, the Salk Institute’s epigenetic atlas integrates methylation, chromatin architecture, and spatial transcriptomics to produce an unparalleled multi-dimensional portrait of the aging brain. By capturing the complexity at single-cell resolution and mapping longitudinal trajectories across regions and cell types, this work heralds a new era of precision neuroscience. Its open-access availability ensures that these rich datasets will catalyze rapid discovery and innovation, accelerating the global quest to understand and ultimately treat neurodegenerative disorders.

Subject of Research: Aging-related epigenetic changes in the mouse brain and their implications for neurodegenerative diseases

Article Title: Cell-type-specific transposon demethylation and TAD remodeling in aging mouse brain

News Publication Date: 11-Mar-2026

Web References:

• Atlas Data on AWS: https://registry.opendata.aws/salk-aging-mouse-brain-epigeneti/
• Journal Article DOI: http://dx.doi.org/10.1016/j.cell.2026.02.015

Image Credits: Salk Institute

Keywords: Neurodegeneration, aging brain, epigenetics, DNA methylation, chromatin conformation, single-cell atlas, transposable elements, spatial transcriptomics, TAD remodeling, CTCF, mouse model, deep learning