A molecular matchmaker sent T cells after aging cells in mice and old monkeys

With age, more cells stop dividing but refuse to die. These senescent cells leak inflammatory signals into the surrounding tissue, and in mice, removing them eases many problems of old age. The drugs that do this are called senolytics. Most of those in clinical trials are small molecules, whose results, the authors of this paper note, have "generally fallen short of predictions", partly because such molecules are not good at telling senescent cells from healthy ones.

A different idea is to let the immune system do the killing. Senescent cells carry a protein called uPAR on their surface, which normally helps cells remodel their surroundings during wound healing. Earlier work aimed CAR-T cells at it — a patient's own T cells, genetically rebuilt to hunt a chosen target — but those multiply inside the body, so the dose is hard to control. This team tried something more adjustable: a T-cell engager, one small protein with two grabbing ends, one for uPAR and one for CD3, a handle found on T cells. Holding the two cells together makes a passing T cell attack whatever it is tethered to, and because the protein washes out of the blood, dosing can be turned down or stopped. For the uPAR end the researchers used not an antibody but a 46-amino-acid piece of uPAR's own natural partner protein, which binds less tightly and so should mostly notice cells carrying a lot of uPAR.

What they did

In dishes, human lung cells pushed into senescence and mixed with human T cells were killed at engager concentrations as low as 0.015 nanomolar, while young cells, and a version of the protein missing its uPAR end, came to no visible harm. A mouse version was built for the animal work.

In mice aged 22 months, the team then tested five doses. The highest, 1 mg per kg of body weight, killed the animals within 24 hours. At 0.3 mg/kg the mice looked ill — ruffled fur, low spirits, loss of appetite — and two liver enzymes in their blood, AST and ALT, rose 20- to 30-fold. These leak out of damaged liver tissue and are measured in ordinary blood tests. Examination showed liver damage but no obvious damage elsewhere.

The reason was the target itself. Liver blood-vessel lining cells are among the few non-immune cell types carrying uPAR, and in treated mice the engager piled up in the liver, sitting where uPAR and a vessel-lining marker overlapped; in mice genetically lacking uPAR, the same dose caused neither liver injury nor any enzyme rise. So the team dropped to 0.1 mg/kg and dosed naturally aged mice once a week for three weeks.

What they found

• In aged mice, three weekly doses lowered uPAR and the senescence markers P16 and P21 in several tissues, reduced staining for senescent cells in fat, liver and lung, and left liver gene activity showing less inflammatory signalling. Two inflammatory messengers roughly doubled during treatment, well short of a dangerous "cytokine storm", then returned to normal. Physical function and muscle fibre density had not improved yet, which the team puts down to the short treatment.
• With longer treatment (weekly for two weeks, then monthly, from 20 months of age), mice showed less scarring in liver, lung and kidney, lower inflammatory messengers in blood, and after six months better grip strength, endurance, running speed, muscle fibre density, bone volume and blood-sugar handling. Liver enzymes stayed normal.
• In three rhesus macaques aged 24 years — roughly a 70-year-old person — a first dose test in one animal found that 40 micrograms/kg caused a dulled mental state, loss of appetite and a 50–100% enzyme rise, still inside the normal range. At 20 µg/kg every three days, one macaque had no signs; another went off its food, with enzyme rises of 100–400% that briefly passed the normal range and cleared within three days. White blood cell counts and blood pressure dipped for a day. Senescent cells and senescence markers in fat, muscle and skin fell by day 42 — and were back up by day 90.

What to keep in mind

• An experimental protein, tested only in cells and animals. No person has received it.
• Liver injury is the main known risk, and it comes from the target itself: uPAR sits on liver vessel-lining cells, which carry more of it with age. The authors' point is that the risk is predictable and visible in a blood test, not that it is absent.
• Very small numbers. Mouse groups held three to five animals; the monkey work used three.
• The effect is temporary: in the monkeys, senescent cells returned within about three months.
• It needs the body's own T cells, which weaken with age; the authors note that older patients might need higher or more frequent doses.
• For prevention in healthy people, the authors write, "the risk–benefit ratio requires careful evaluation"; they suggest starting instead with conditions such as chronic kidney disease, Alzheimer's disease or liver fibrosis. They declare no conflicts of interest.

Why it matters

Senolytics have struggled with aim: hit senescent cells hard enough and you tend to hit healthy ones too. This study points a tool from cancer medicine at a marker of senescent cells and finds a working dose plus a cheap way to watch for the main harm — and shows the catch: the same marker sits on a vital part of the liver.