In mice, each organ seems to age with its own set of proteins

Organs don't all wear out at the same pace. A person can have a strong heart and failing kidneys, or the other way round. To put a number on this, researchers build aging clocks: formulas that combine many molecular measurements into one score that predicts age. Most clocks read chemical tags on DNA, and a newer kind uses the levels of hundreds of proteins.

In people, protein clocks are usually built from blood, because blood is easy to take. But blood only carries proteins that leak or are released from organs, so it gives an indirect view. In mice, scientists can measure proteins inside each organ directly. A new study, posted as a preprint that has not yet been peer-reviewed, used that to ask a simple question: does the whole body age along one shared protein pattern, or does each organ have its own?

What they did

The team did not run new experiments. It re-analysed datasets that other labs had already published.

• Building the clocks: the starting data were protein measurements from eight organs (brain, heart, lung, liver, kidney, spleen, leg muscle and testis) of 44 male laboratory mice, aged from 3 to 26 months. For each organ, a statistical model picked out a set of proteins whose combined levels predict age.
• Testing them: the clocks were then applied, unchanged, to protein data from two independent groups of mice measured with different methods, and to mice given rapamycin, a drug that extends lifespan in mice and blocks the growth signal mTOR.
• Proteins versus RNA: a protein is made from a gene's RNA message, but cells also control how fast proteins are built and broken down. So the team applied each protein clock to the RNA levels of the same genes, to see how much of the protein aging pattern already shows up in RNA.
• Where in the tissue: finally, the clocks were scored on spatial maps of gene activity in thin tissue slices from mice aged about 4, 17 and 26 months.

What they found

• Each clock tracked age in its own organ. Within the original mice, predicted and real age were closely linked in every organ, from a correlation of 0.72 in the brain to 0.91 in leg muscle (1 would be perfect). The scores also rose with age in the independent groups of mice.
• The organs used different proteins. Of 911 proteins used by the eight clocks, 848 (93%) appeared in only one organ's clock. Only three proteins turned up in more than two organs, two of them parts of antibodies.
• Clocks rarely worked on other organs. When a clock trained on one organ was used on another organ of the same mice, the typical correlation with age was only 0.15. The kidney clock was the main exception: it still tracked age fairly well in brain, spleen and testis.
• RNA carried much of the signal, but not everywhere. In kidney, liver and lung, the RNA version of the score followed age about as well as the protein version. In brain and leg muscle, the proteins tracked age better, which hints that more of the aging change there happens after the RNA stage, in how proteins are made or broken down.
• Aging happened inside existing tissue regions. In the tissue maps, scores rose with age within defined regions, such as a type of kidney tubule or the brain's white matter, mostly without those regions growing or shrinking.
• Rapamycin gave mixed results. Treated mice had lower scores in some cases (leg muscle in males, liver in females in one dataset, liver in males in another) but not in others, and two kinds of long-lived mutant mice gave inconsistent results.

What to keep in mind

• This is a preprint: other scientists have not yet checked it in peer review.
• It is a study in mice, with no new animals. The authors note that most of the data came from male mice of a single strain, so it is not known whether the results hold for females or other strains.
• Group sizes were small: 44 mice for building the clocks and between 6 and 52 per group in the test sets.
• The scores in the test datasets are relative rankings, not ages in months, so they can't be compared directly between studies.
• The tissue maps came from only three ages, and the authors say their resolution cannot show finer changes between cell types.
• Higher scores in one brain region do not prove that region is biologically older, the authors write, and the proteins a clock picks are predictors, not proven causes of aging.
• Whether these organ clocks say anything about later disease will need further testing.

Why it matters

If organs age in largely separate molecular ways, one clock for the whole body will miss a lot, and a treatment that slows aging in one organ might do little for another. The authors point to studies in people in which organ ages estimated from blood proteins were linked to later disease in the matching organ. This mouse study adds a direct look inside the organs themselves. (A similar idea, applied to the compartments inside cells, is in /post/2026-11-07-organelle-aging-blood-proteins/.) For now it is a map of how aging differs between organs in mice, not a test that can be used in people.