Integration of more than 11,000 transcriptomes across 4 mammals reveals universal transcriptomic signatures of mammalian ageing and mortality with a modular architecture encompassing inflammation, interferon signalling, mitochondrial function, chromatin modification and extracellular matrix organization, providing a framework for quantifying and targeting ageing of cellular subsystems across species and tissues.
Key Findings
Methods
Accurate transcriptomic biomarkers of chronological age and expected mortality were developed from integration of over 11,000 transcriptomes across more than 25 tissues in four mammalian species.
The dataset encompassed transcriptomes from mouse, rat, macaque, and human
Both rodent-specific and multi-species biomarkers were developed
The biomarkers predicted lifespan-modulating interventions, time to death, chronic diseases, and rejuvenation
The study developed interpretable (not just predictive) biomarkers, allowing mechanistic insight
Results
Ageing-related transcriptomic changes were conserved across species and cell types, revealing universal signatures of mammalian ageing and mortality.
Conservation was observed across mouse, rat, macaque, and human
Conserved changes were identified across more than 25 tissue types
Universal signatures included specific genes such as CDKN1A and LGALS3
Conserved signatures spanned both chronological ageing and mortality-associated changes
Results
CDKN1A and LGALS3 protein levels were associated with mortality and multimorbidity in the UK Biobank cohort.
These two genes were identified as part of universal transcriptomic signatures of mammalian ageing and mortality
Associations were validated at the protein level, not just the transcript level
UK Biobank is a large-scale human biomedical database and research resource
CDKN1A encodes p21, a cyclin-dependent kinase inhibitor associated with senescence; LGALS3 encodes galectin-3, associated with inflammation and fibrosis
Results
Mortality-associated transcriptomic features were recapitulated across multiple in vivo and in vitro damage-accumulation models.
Models included inflammation, replicative senescence, metabolic inhibition, and γ-irradiation
Both in vivo and in vitro systems showed convergent mortality-associated signatures
These features were attenuated or reversed by cell immortalization, reprogramming, heterochronic parabiosis, and early embryogenesis
This convergence across diverse damage models supports a common underlying mechanism of ageing-associated transcriptomic change
Results
Network analysis revealed a modular architecture of ageing- and mortality-associated hallmarks encompassing five major biological processes.
The five modules identified were: inflammation, interferon signalling, mitochondrial function, chromatin modification, and extracellular matrix organization
Module-specific clocks were developed to quantify ageing of individual cellular components
The modular framework allowed pathway-specific effects of interventions to be distinguished
Caloric restriction and Klotho (Kl) deficiency targeted mitochondrial and metabolic modules
Results
Transcriptomic and DNA methylation clocks showed correlated age acceleration in human blood, with the strongest correlation observed for the chromatin-associated module clock.
Correlation between transcriptomic and epigenetic (DNA methylation) clocks was assessed in human blood samples
The chromatin-associated module clock showed the strongest correlation between the two ageing modalities
This finding highlights mechanistic links between transcriptomic and epigenetic ageing
The result suggests chromatin modification may be a central convergence point for multiple molecular ageing processes
Results
Module-specific transcriptomic clocks revealed distinct pathway-specific effects of ageing interventions.
Chronic diseases primarily accelerated ageing of the inflammatory module
Caloric restriction targeted mitochondrial and metabolic modules
Klotho deficiency also targeted mitochondrial and metabolic modules
This pathway-level resolution was not achievable with single composite ageing clocks
The approach provides a framework for targeting ageing of specific cellular subsystems
Results
Mortality-associated transcriptomic signatures were attenuated or reversed by rejuvenation-associated interventions including heterochronic parabiosis and early embryogenesis.
Heterochronic parabiosis (surgically joining young and old animals to share circulation) reduced mortality-associated transcriptomic features
Early embryogenesis was also associated with attenuation of mortality signatures
Cell immortalization and cellular reprogramming similarly reversed these features in vitro
These findings support the biological reversibility of at least some ageing-associated transcriptomic changes
What This Means
This research suggests that ageing leaves consistent molecular fingerprints — patterns of gene activity — that are shared across mammals including mice, rats, monkeys, and humans, and that these patterns are predictive not just of age but of how long an individual is likely to live and whether they have chronic diseases. By analyzing over 11,000 gene expression profiles from more than 25 different tissue types, the researchers built biological clocks that can estimate a person's or animal's biological age from tissue samples, and identified specific genes (including CDKN1A and LGALS3) whose levels in the blood are linked to death risk and having multiple diseases simultaneously in a large human cohort (UK Biobank).
This research suggests that the biology of ageing is not a single unified process but rather a collection of semi-independent modules — including inflammation, energy production in cells (mitochondrial function), chromosome packaging (chromatin modification), immune signaling, and structural tissue changes (extracellular matrix) — each of which ages somewhat independently and responds differently to interventions. For example, caloric restriction appears to primarily slow ageing of the mitochondrial and metabolic modules, while chronic diseases mostly accelerate the inflammatory module. This modular view means that different anti-ageing strategies may work through different cellular pathways, and module-specific clocks can detect these pathway-specific effects that a single overall ageing clock would miss.
This research also suggests that ageing-associated molecular changes are not irreversible: procedures like connecting young and old animals' circulatory systems (heterochronic parabiosis), cellular reprogramming, and even normal early embryo development can partially reverse or reset these ageing signatures. The convergence of the transcriptomic (gene activity) and epigenetic (DNA methylation) clocks — particularly in chromatin-related processes — points to chromatin regulation as a possible hub linking different types of molecular ageing. Together, these findings provide both new tools to measure biological ageing more precisely and new targets for interventions aimed at slowing or reversing specific aspects of the ageing process.
Tyshkovskiy A, Kholdina D, Davitadze M, Molière A, Moldakozhayev A, Tongu Y, et al.. (2026). Universal transcriptomic hallmarks of mammalian ageing and mortality.. Nature. https://doi.org/10.1038/s41586-026-10542-3