The research, led by Marloes Oosterhof and Louis Coussement of the University of Groningen together with colleagues in Belgium, Austria and the Netherlands, focused on the liver, the metabolic command center that manages the fuel switch from glucose to fat oxidation that hibernation demands. Because facultative hibernators like the Syrian hamster enter torpor only in response to environmental cues, shortening days, cold and food scarcity, their bodies must actively reprogram gene expression rather than simply follow a fixed seasonal clock. That flexibility makes epigenetic mechanisms, the chemical systems that switch genes on and off without altering the underlying sequence, prime suspects for orchestrating the transition.

To test this idea, the team induced hibernation in Syrian hamsters by shortening the light-dark cycle and lowering ambient temperature, then collected liver samples at carefully defined phases: summer euthermia, when animals were active and warm; late torpor, more than 72 hours after the animals had settled into a bout; and both early and late arousal, shortly after the animals rewarmed. Using infrared motion detectors, the researchers confirmed each animal’s state before euthanasia, ensuring that the molecular snapshots truly reflected the hibernation phase rather than guesswork about body temperature.

The first layer of analysis examined the expression of 94 genes encoding epigenetic writer and eraser enzymes, the proteins that add or remove chemical marks on histones and DNA. Of the 83 genes that could be reliably measured, 11 shifted significantly between phases. Strikingly, four histone acetyltransferases, KAT6A, KAT6B, KAT7 and KAT13D, better known as the circadian gene CLOCK, were differentially expressed, with most declining during torpor and rebounding during arousal. Three deacetylases, SIRT2, SIRT7 and HDAC11, also changed, along with two histone methyltransferases, one demethylase and the DNA demethylating enzyme TET2.

Notably, several of these enzymes, including PCAF, GCN5L2, HAT1 and CBP, had been flagged in previous work on the thirteen-lined ground squirrel, an obligatory hibernator, but showed no expression changes in the hamster. This divergence suggests that facultative hibernators may deploy a distinct epigenetic toolkit, one tailored to switching hibernation on and off in response to the environment rather than to an internal circannual rhythm. The downregulation of CLOCK in the hamster liver is particularly intriguing, since recent work indicates that the liver pauses its circadian rhythms during hibernation while the brain retains them.

Gene expression, however, is only an indirect readout. The team therefore measured histone acetylation directly, first by Western blot and then by quantitative mass spectrometry, a technique that can determine exactly how many acetyl groups sit on specific lysine residues of histones H3 and H4. The results were consistent and telling. Overall acetylation of histone H3 dropped during late torpor, with the acetylation of lysines 18 and 23 falling by roughly 5 and 14 percent respectively, and rising again during arousal. H3K27ac, a mark classically associated with active enhancers and open chromatin, was lowest in torpor and recovered as animals rewarmed, a pattern that fits the idea of chromatin condensing, and transcription shutting down, while the animal is cold and metabolically dormant.

Histone H4 told a subtler story. Unlike H3, whose acetylation largely rebounded during arousal, the H4 peptide spanning lysines 5, 8, 12 and 16 remained under-acetylated even after the animals had rewarmed. The single most heavily used site, H4K16, a modification with the unusual ability to both promote and repress transcription depending on context, was significantly less acetylated in late arousal than in summer-active animals. The researchers point to SIRT2 and SIRT7, both of which target H4K16 and both of which were upregulated during arousal, as plausible culprits, noting that these sirtuins can translocate into the nucleus during cell-cycle transitions and stress, both of which accompany the metabolic turbulence of rewarming.

On the enzyme-activity side, histone acetyltransferase activity stayed flat across all phases, but histone deacetylase activity peaked in late torpor. Inhibitor profiling revealed that a single enzyme, HDAC6, accounted for roughly 71 percent of total hepatic HDAC activity. Yet here the plot thickened: although HDAC6 activity was high and histone acetylation low during torpor, the acetylation of alpha-tubulin, HDAC6’s best-known cytosolic target, actually increased. The authors suggest that HDAC6 may relocate to the nucleus during torpor, deacetylating histones there while leaving tubulin untouched, and note that the enzyme is also tied to hepatic gluconeogenesis and to protective responses in ischemic liver injury, processes highly relevant to both hibernation and, potentially, human organ preservation.

Perhaps the most surprising negative result concerned DNA methylation. Using both the Luminometric Methylation Assay and whole-genome bisulfite sequencing, the team found no significant differences in global cytosine methylation at CpG or non-CpG sites across any hibernation phase, and promoter methylation of the differentially expressed genes was likewise unchanged. This stability mirrored the near-absence of changes in the expression of DNA methyltransferases and demethylating enzymes, and it contrasts with reports of methylation shifts in ground squirrels and horseshoe bats. In the Syrian hamster, at least, the epigenetic levers of hibernation appear to be pulled at the histone level rather than at the DNA level.

The broader implications reach well beyond rodent physiology. Because facultative hibernators can trigger hypometabolic, cytoprotective states essentially on command, understanding their epigenetic switches could inform strategies to protect organs during transplantation, to limit ischemic damage, and even to slow aging, given recent evidence that epigenetic aging is effectively stalled during hibernation in yellow-bellied marmots. The authors caution that their bulk liver measurements largely reflect hepatocytes and that single-cell follow-up work will be needed to resolve contributions from rarer cell types, and that the expression changes, though modest in fold change, match the magnitude of the measured histone modifications. Still, the message is clear: when a hamster decides to hibernate, it does not rewrite its DNA. It retunes the packaging, loosening and tightening the spools of chromatin to silence the genome through the cold and swing it back open for the warmth of arousal.

Subject of Research: Epigenetic regulation of gene expression, specifically histone acetylation and DNA methylation, in the liver of facultatively hibernating Syrian hamsters

Article Title: Changes in histone lysine acetylation, but not DNA methylation during facultative hibernation in Syrian hamster liver

Article References: Oosterhof, M. M., Coussement, L., van Pijkeren, A., Kwiatkowski, M., Zwinderman, M. R. H., Dekker, F. J., de Meyer, T., Reitsema, V. A., Bischoff, R., Guryev, V., Bouma, H. R., Henning, R. H., & Rots, M. G. (2024). Changes in histone lysine acetylation, but not DNA methylation during facultative hibernation in Syrian hamster liver. Epigenetics Communications, 4(1), Article 1. https://doi.org/10.1186/s43682-023-00024-2

Image Credits: AI Generated

DOI: 10.1186/s43682-023-00024-2

Keywords: hibernation, Syrian hamster, epigenetics, histone acetylation, DNA methylation, HDAC6, sirtuins, liver metabolism, torpor, chromatin, KAT6A, epigenetics communications

The experimental design was deliberately rigorous. The team infected 7-to-10-week-old and 12-month-old female and male Syrian golden hamsters intranasally with the BavPat1/2020 SARS-CoV-2 isolate, then followed them for five days under biosafety level 3 conditions. Crucially, the researchers paired classic disease readouts with a comprehensive lipidomic analysis: they quantified viral RNA in oropharyngeal swabs, bronchoalveolar lavage fluid, lung, duodenum, and colon by real-time reverse transcription PCR, scored pulmonary lesions histopathologically, and measured cholesterol and six distinct oxysterol species by high-performance liquid chromatography in intestinal tissues, lung, and serum. The lipid measurements were performed after methanol-mediated viral inactivation, validated with spiking experiments, so that samples could safely leave the high-containment laboratory for analysis.

Clinically, the model reproduced what is now well established for aged hamsters. From day three post infection onward, the older animals deteriorated markedly: by day five their mean clinical score reached 6.5 points, compared with 3.7 points in young animals whose condition had stabilized, and aged hamsters lost the most weight relative to their starting body mass. Hypersalivation, particularly pronounced in aged females, was an additional distinctive sign. Yet when the pathologists examined the lungs, the picture was surprisingly uniform. Total lesion scores, comprising alveolar, airway, and vascular changes, rose significantly in all infected groups but did not differ between age groups, and mild rhinitis and tracheitis appeared across the board. The aged animals were sicker without having visibly worse lungs.

The intestine told a different story. Although microscopic examination of the duodenum and colon revealed no overt tissue damage or SARS-CoV-2 nucleoprotein immunoreactivity in either age group, molecular testing showed that aged hamsters carried significantly higher SARS-CoV-2 RNA loads in both intestinal segments than their young counterparts. Respiratory viral RNA in swabs and lavage fluid was also slightly higher in the aged groups. This combination, a morphologically silent but virologically active gut infection concentrated in older hosts, positioned the intestine as a key site where age-dependent differences in viral burden and metabolic response could be dissected without the confounding noise of severe tissue pathology.

The lipidomic data revealed a stark divergence. Upon infection, young hamsters showed a striking upregulation of the side-chain oxysterols 25-hydroxycholesterol and 27-hydroxycholesterol in duodenum and colon, and this pattern was mirrored in serum, where both molecules rose significantly. These compounds are not passive metabolic byproducts. 25-hydroxycholesterol in particular is recognized as a central component of innate antiviral defense, capable of blocking viral fusion and replication, suppressing inflammasome activation through liver X receptor–dependent inhibition of sterol regulatory element-binding protein signaling, and participating in interferon-mediated immune regulation. In aged hamsters, this protective response was essentially absent: intestinal 27-hydroxycholesterol was mostly undetectable even after infection, and the modest rise in 25-hydroxycholesterol never reached statistical significance in any tissue.

Instead, the aged animals showed a selective accumulation of ring oxysterols, oxidized at the sterol ring rather than the side chain. Colonic 4β-hydroxycholesterol surged in infected aged hamsters and correlated strongly and positively with clinical score, age, and viral RNA loads in both bronchoalveolar lavage fluid and oral swabs, while showing negative correlations with serum cholesterol and the protective side-chain oxysterols. This finding echoes earlier reports that 4β-hydroxycholesterol rises in colitis in humans and mouse models, hinting at a role in colonic inflammation. Meanwhile 7β-hydroxycholesterol, another ring oxysterol, was significantly downregulated in the intestines of infected aged animals, whereas 7-ketocholesterol and 5β,6β-epoxycholesterol remained unchanged, indicating a selective remodeling of specific oxysterol pathways rather than a blanket increase in oxidative cholesterol products.

A plausible mechanism for this age-dependent shift lies in the biology of aging itself. Advancing age is associated with mitochondrial dysfunction and elevated production of reactive oxygen species, which drive nonenzymatic autooxidation of cholesterol into ring oxysterols, a hallmark of age-related oxidative stress frequently observed in tissues of elderly individuals with inflammatory comorbidities. Side-chain oxysterols, by contrast, are generated enzymatically, and their coordinated induction in young animals reflects a targeted metabolic defense. Notably, oxysterols were undetectable in lung tissue regardless of age or infection status, yet young hamsters showed a significant depletion of lung cholesterol after infection while aged animals showed only a nonsignificant trend, suggesting that the pulmonary response to SARS-CoV-2 operates through cholesterol depletion rather than local oxysterol production.

The correlation analyses tied these threads into a coherent, if still correlative, framework. Intestinal and serum side-chain oxysterols were positively correlated with each other, supporting the intestine as a major source of circulating oxysterols, and both showed significant negative correlations with viral RNA in bronchoalveolar lavage fluid and with clinical severity. Serum 25-hydroxycholesterol inversely correlated with age. Perhaps most strikingly, side-chain oxysterols from the duodenum and colon, and serum 25-hydroxycholesterol, all correlated negatively with total lung cholesterol, which itself was inversely associated with viral load and clinical score. Because SARS-CoV-2 entry depends on cholesterol-rich lipid rafts that anchor the ACE2 receptor, reduced cholesterol availability in pulmonary cells could plausibly restrict viral entry, offering a mechanistic rationale for how gut-derived oxysterols might protect the lung from a distance.

The study also uncovered systemic signatures with practical potential. Because oxysterol measurement is not routine in clinical settings, the team searched for surrogates among standard blood chemistry parameters. Serum 25-hydroxycholesterol correlated positively with total bilirubin and negatively with total protein and globulin, markers that were themselves linked to advanced age and more severe disease, while declining potassium, inorganic phosphate, and creatine phosphokinase tracked with better outcomes and younger age. The drop in potassium and phosphate in aged hamsters alongside higher intestinal viral loads hints at an infection-induced malabsorptive phenotype, possibly involving a shift of the epithelium toward goblet cell hyperplasia at the expense of absorptive enterocytes, a remodeling described in other coronaviruses through disruption of Notch signaling. The authors are appropriately cautious: elevated 25-hydroxycholesterol may reflect an effective host response rather than cause it, and distinguishing these possibilities will require targeted mechanistic studies. They also propose that antioxidant nutritherapeutics, such as alpha-tocopherol and polyphenols, which counteract ring oxysterol toxicity, could merit exploration as a strategy to improve outcomes in elderly patients, pending validation in larger cohorts and clinical trials.

Subject of Research: Age-dependent intestinal oxysterol remodeling during SARS-CoV-2 infection in Syrian hamsters

Article Title: Intestinal oxysterol remodeling underlies severe COVID-19 in an aged hamster model

Article References: Mergani, A., Wirz, K., Reineking, W., Richter, A., Lechler, V., Pilchová, V., Störk, T., Michaely, L. M., Zdora, I., Buettner, M., Volz, A., Bleich, A., Baumgärtner, W., Schulz, C., & von Köckritz-Blickwede, M. (2026). Intestinal oxysterol remodeling underlies severe COVID-19 in an aged hamster model. GeroScience. https://doi.org/10.1007/s11357-026-02531-1

Image Credits: AI Generated

DOI: 10.1007/s11357-026-02531-1

Keywords: SARS-CoV-2, COVID-19, oxysterols, 25-hydroxycholesterol, aging, cholesterol metabolism, intestine, lipidomics, Syrian hamster, gut-lung axis, innate immunity, GeroScience