The focus of the investigation was the lipid droplet, a dynamic intracellular organelle consisting of a core of neutral lipids such as triglycerides and cholesteryl esters wrapped in a phospholipid monolayer. Lipid droplets bud off from the endoplasmic reticulum with the help of proteins like seipin and FIT2, and their size and turnover are managed by an array of peripheral and transmembrane proteins. Although lipid droplets are best known as energy reserves in fat cells, they also serve as protective buffers, sequestering potentially toxic lipids and shielding cells from oxidative stress. In the brain, their role has become a hot topic, particularly because lipid droplet accumulation in microglia, the resident immune cells of the central nervous system, has been linked to aging and to neurodegenerative diseases such as Alzheimer’s and Parkinson’s.

Previous work had left the field divided. One influential study introduced the term lipid droplet accumulating microglia, or LDAM, and showed that these fat-laden immune cells produce reactive oxygen species and perform phagocytosis poorly, suggesting they drive neurodegeneration. Other research pointed in the opposite direction, proposing that transferring lipid droplets from neurons to glial cells is a protective response that spares neurons from lipotoxicity. The new study adds crucial comparative evidence to this debate by examining a species that ages remarkably slowly. The researchers used coronal brain sections from naked mole rats aged about four and thirteen years, corresponding to young and middle-aged mice of twenty-six and fifty-two to fifty-four weeks, and stained the tissue with the neutral lipid probe BODIPY 493/503 together with markers for microglia and astrocytes. To confirm that the fluorescent signal truly came from lipids, they also performed Stimulated Raman Scattering imaging, a label-free technique that targets the methylene vibrational band of lipid acyl chains, and found near-complete overlap between the two signals.

The results in the pyramidal cell layers of the hippocampus were revealing. In mice, the lipid droplets inside neurons of the CA3 region grew significantly larger with age, expanding by sixty-eight percent from an average cross-sectional area of 0.19 to 0.32 square micrometers. In naked mole rats, by contrast, neuronal lipid droplet size was unchanged between the young and middle-aged groups, remaining at roughly 0.15 square micrometers in both. The density of droplets per unit area of the pyramidal layer was similar between the species and unaffected by aging, meaning the key difference lay in droplet size and in where the fat was being stored. In the CA1 region, the naked mole rat droplets were actually smaller than those of mice at both ages, and neither species showed an age-related change there.

The most dramatic contrast emerged when the researchers examined microglia. Naked mole rat microglia were packed with lipid droplets that clustered extensively, making individual droplets impossible to distinguish. Measured as the percentage of microglial area covered by lipid droplets, the naked mole rat values dwarfed those of mice: in the CA3 region, young naked mole rats showed 12.07 percent coverage versus 2.41 percent in young mice, and middle-aged naked mole rats reached 20.06 percent versus just 1.79 percent in mice. In the CA1 region, the coverage was generally fifty to sixty-three percent lower than in CA3, but the pattern held, with middle-aged naked mole rat microglia carrying nine times the droplet load of young mice and fourteen times that of middle-aged mice. Critically, the number of microglia was similar across groups, so the difference reflected genuine fat storage rather than a change in cell counts.

When the researchers calculated the ratio of lipid droplet content in microglia to that in neurons, the species difference became stark. Naked mole rats showed six- to eightfold higher droplet coverage in microglia than in neurons in CA3, and sixteen- to twentyfold higher in CA1, while mice distributed their droplets more or less equally between the two cell types. The authors propose that this pattern represents a shuttling mechanism: the naked mole rat moves lipids out of neurons and into microglia, preventing the accumulation of toxic fats inside the cells that cannot be replaced easily. In mice, this transfer appears far less efficient, leaving neuronal droplets to enlarge with age. Because apolipoproteins, including the Alzheimer’s risk factor APOE4, govern lipid transport between neurons and glia, the finding raises the possibility that the efficiency of this neuron-to-glia lipid shuttle is a determinant of how vulnerable a brain is to lipotoxicity and degeneration.

Importantly, the study challenges the assumption that fat-laden microglia are inherently harmful. The naked mole rat, a species with exceptional resistance to brain aging, carried the heaviest microglial lipid load, yet its microglia showed no age-related loss of ramification, no increase in cell body area, and no rise in cell density. In mice, by contrast, microglial area increased with age in CA3 and microglial density tended to rise, hinting at early neuroinflammation. Correlation analyses showed a weak inverse relationship between droplet load and branching in both species, suggesting mildly reactive microglia hold more fat, but the causality remains unresolved. The authors note that lipid droplets may buffer peroxidized lipids, sequester fatty acids released during autophagy, and even detoxify misfolded proteins, supporting the view that droplet accumulation is protective at physiological levels and becomes damaging only when excessive.

The second major discovery concerned blood supply. Using collagen IV staining to visualize capillaries, the team quantified vascular density in the dentate hilus, the CA1 stratum radiatum, and the sensory and motor cortex. In every region and at every age, naked mole rats had dramatically sparser capillary networks than mice, with densities as low as one quarter to one half of mouse levels. In the dentate hilus, for example, young mice showed about 1.00 percent capillary coverage against 0.35 percent in naked mole rats, and middle-aged mice rose to 1.31 percent while the naked mole rat stayed flat at 0.34 percent. Mouse capillary density actually increased with age in the hilus and cortex, possibly a compensatory response to declining blood oxygen that later fails in old age, whereas the naked mole rat vasculature was completely stable across the lifespan.

Far from being a deficit, the authors argue, this sparse vasculature is likely part of a coordinated adaptation. Naked mole rats inhabit chronically hypoxic, carbon dioxide-rich burrows, have a markedly reduced basal metabolic rate, possess hemoglobin with high oxygen affinity, and can survive eighteen minutes of complete anoxia without brain damage, partly by switching to fructose-fueled anaerobic metabolism. Their brains appear to run on a minimal but sufficient capillary network optimized for stable function under low oxygen rather than maximal perfusion. The low capillary density may even connect back to the lipid droplet story: reduced perfusion can promote reactive oxygen species and lipid peroxidation, and microglial lipid droplets may be the naked mole rat’s intrinsic defense against exactly that damage, soaking up peroxidized fats before they can injure neurons.

Together, the findings sketch a coherent picture of a brain built for endurance. Where the aging mouse accumulates enlarging fat droplets in its neurons and shows early signs of microglial activation, the naked mole rat keeps its neurons lean and parks the fat in microglia, all while maintaining a stable, sparse blood supply that never changes with age. The work suggests that lipid droplet accumulating microglia are not intrinsically deleterious and that the traffic of lipids between brain cells, mediated by apolipoproteins, may be a decisive factor in neural resilience. If future research can confirm and harness this shuttling mechanism, it could open new avenues for protecting the human brain against Alzheimer’s disease, stroke, and the slow attrition of aging itself.

Subject of Research: Age-related lipid droplet distribution and capillary density in the hippocampus of naked mole rats compared with mice

Article Title: Age-dependent changes in lipid droplet distribution and vascularization in naked mole rat vs. mouse hippocampus

Article References: Krogstad, L. S., Teppen, M. A., Mjønes, H. S., Brewer, J. R., Park, T., Geiseler, S., & Morland, C. (2026). Age-dependent changes in lipid droplet distribution and vascularization in naked mole rat vs. mouse hippocampus. GeroScience. https://doi.org/10.1007/s11357-026-02548-6

Image Credits: AI Generated

DOI: 10.1007/s11357-026-02548-6

Keywords: naked mole rat, lipid droplets, microglia, hippocampus, aging, neurodegeneration, capillary density, vascularization, lipotoxicity, BODIPY, hypoxia tolerance, GeroScience

The research, led by Maëlle Daunesse, Eulalie Liorzou, Elise Parey, Diego Villar and Camille Berthelot, tackles a problem that has long frustrated evolutionary biologists: distinguishing adaptive changes in gene expression from the background noise of neutral evolution. Over millions of years, transcriptomes—the complete sets of RNA molecules produced by tissues—accumulated countless random shifts in expression level. Most of these shifts mean nothing for survival. A minority, however, were driven by natural selection because they improved some aspect of physiology. Separating the two categories with confidence has been notoriously difficult, blurring the adaptive signatures written into mammalian genomes.

To cut through that noise, the team built a phylogeny-aware framework for modelling quantitative trait evolution, integrating comparative multiomics data across four rodent species. They generated RNA-sequencing data from liver and heart tissue in two mole-rats—the naked mole-rat (Heterocephalus glaber) and the Damaraland mole-rat (Fukomys damarensis)—and compared these profiles with two rodent outgroups, the guinea pig (Cavia porcellus) and the mouse (Mus musculus). The outgroups matter because they anchor the comparison: by modelling how expression levels should drift along each branch of the evolutionary tree, the researchers could flag genes whose expression had shifted faster in mole-rat lineages than neutral drift alone could plausibly explain.

The analysis, built around the Expression Variance and Evolution approach, revealed hundreds of genes whose expression levels had undergone lineage-specific acceleration consistent with directional selection rather than random walk. After careful normalisation—using variance-stabilising transformation and transcripts-per-million strategies validated through sensitivity analyses—the team confirmed that these shifts were not artefacts of sequencing depth, gene length or transcript annotation. Orthology reannotation and genome assembly quality checks further ensured that the genes being compared across species were true one-to-one counterparts, eliminating a common source of false positives in cross-species transcriptomics.

But identifying shifted genes was only half of the puzzle. To connect expression evolution with its underlying regulatory mechanisms, the researchers layered a second data type onto the transcriptomic map: epigenomic profiles of cis-regulatory elements, the DNA sequences near genes that control when, where and how strongly they are transcribed. Cis-regulatory landscapes are known to turn over rapidly in mammalian evolution, and regulatory changes are widely hypothesised to be a major engine of phenotypic innovation. If a gene’s expression truly evolved under selection in mole-rats, the logic goes, the regulatory elements governing it should show signs of accelerated change as well.

That is precisely what the team found. Genes with lineage-specific expression shifts were surrounded by epigenomic landscapes that also displayed signals of accelerated evolution—a concordance that is difficult to explain by chance or by neutral genomic processes diverging at different rates. The researchers developed a weighted regulatory score, assigning regulatory elements to nearby genes within optimised genomic windows, and showed that genes exhibiting both transcriptomic and regulatory shifts cluster at specific loci. At these candidate loci, the two independent lines of evidence—expression and regulation—point in the same direction, strengthening the case that selection, not drift, drove the change.

The functional identities of these fast-evolving loci are as striking as the methodological rigour behind them. Enrichment analyses showed that genes with evidence of selection at both expression and regulatory levels cluster in pathways central to mole-rat physiology, including metabolism and stress responses. For animals that spend their entire lives underground in hypoxic, carbon dioxide-rich burrows, metabolic rewiring is not a luxury but a necessity. Cellular stress pathways, likewise, must cope with the oxidative and environmental challenges of the subterranean niche. The study’s molecular signatures suggest that selective pressures honed exactly those biological systems a burrowing mammal would need to tune.

Among the genes implicated by the analysis is Pygl, encoding liver glycogen phosphorylase, a key enzyme in glycogen breakdown. RNA-sequencing read coverage across the Pygl locus, presented in the study’s supplementary figures, illustrates how expression and regulatory profiles diverge across the four species in ways consistent with lineage-specific tuning. Genes of this kind offer testable hypotheses for experimental follow-up: if regulatory sequences around Pygl indeed evolved adaptively in mole-rats, they could contribute to the extraordinary metabolic flexibility these animals display when oxygen is scarce.

Beyond the mole-rat findings themselves, the study’s framework carries broader significance for evolutionary genomics. Comparative multiomics studies have often treated gene expression and gene regulation as separate domains, analysed with different toolkits and interpreted in isolation. By jointly modelling both within a single phylogenetic framework, the Berthelot and Villar teams demonstrate that concordant evidence across molecular layers is a far stronger indicator of adaptive evolution than either layer alone. The approach could be applied to any group of organisms with multiple genome assemblies and tissue-specific expression data, from cavefish to high-altitude mammals, wherever convergent adaptation to extreme environments raises similar questions.

The work also underscores how much biology remains hidden in species that were long considered curiosities. Mole-rats have already reshaped research on ageing, cancer suppression and pain insensitivity, and the loci flagged by this study provide a molecular shortlist for understanding traits that could eventually inform human medicine. As the authors note, their results illuminate how gene regulation and expression evolve in concert in this mammalian model, and they highlight genomic loci where multiple lines of evidence converge on a single conclusion: these are the places in the mole-rat genome where natural selection left its clearest fingerprints, shaping a lineage into one of the most unusual mammals on Earth.

Subject of Research: Comparative multiomics analysis of gene expression and regulatory evolution in African mole-rats

Article Title: Coordinated shifts in gene expression and regulation identify fast evolving loci linked to unique mole-rat traits

Article References: Daunesse, M., Liorzou, E., Parey, E., Villar, D., & Berthelot, C. (2026). Coordinated shifts in gene expression and regulation identify fast evolving loci linked to unique mole-rat traits. BMC Biology. https://doi.org/10.1186/s12915-026-02713-5

Image Credits: AI Generated

DOI: 10.1186/s12915-026-02713-5

Keywords: mole-rats, evolutionary genomics, gene expression, gene regulation, adaptive evolution, cis-regulatory elements, comparative multiomics, phylogenetic modelling, BMC Biology, naked mole-rat, Damaraland mole-rat, subterranean adaptation