Engineered immune cells cleared inflamed bone-marrow cells and improved memory in old mice

As the brain ages it becomes slightly inflamed, and that low-grade inflammation is thought to contribute to memory loss. Most research has looked at microglia, the brain’s own resident immune cells. This study looks at a smaller group: border-associated macrophages, immune cells that sit at the brain’s edges, in its outer membranes and along its blood vessels.

The tool is CAR T cells: T cells, a type of white blood cell, that are taken out and given an engineered receptor so they recognise and kill cells carrying a chosen surface protein. CAR T cell therapy is used today against some blood cancers. Here the target is uPAR, a surface protein that marks inflammatory, worn-out cells and has been targeted with CAR T cells before.

What they did

Working at Cold Spring Harbor Laboratory, the team gave mice a single dose of mouse uPAR CAR T cells; control mice got ordinary T cells. The cells were injected either into the fluid around the spinal cord (intrathecal), which reaches the brain, or into a vein (intravenous), from which they did not reach the brain in meaningful numbers. They treated old mice, young mice who were then left to age, and PS19 mice, which carry a human tau gene and develop tangles, memory problems and early death.

Memory was tested with three standard tasks: recognising a new object, exploring a new arm of a Y-shaped maze, and finding a hidden platform in a pool. The team also counted and profiled immune cells in the brain and bone marrow, and transplanted bone-marrow cells from treated old mice into other old mice. Finally, they looked for the human version of uPAR in public gene-activity data from human brains and bone marrow.

What they found

• Better memory after one dose. Old mice given CAR T cells into the spinal fluid outperformed old controls on all three tasks 2.5 months later, and still did six months later. Young mice treated at 2 months kept better memory when tested at 18 months, 16 months after the dose. The cells were still present in their brains.
• Tau mice did better and lived longer. In PS19 mice, the treatment lessened memory problems, reduced tau build-up and increased median survival by 37.8%.
• The target: inflamed border macrophages. In old mouse brains, most uPAR-carrying cells were border-associated macrophages with a strongly inflammatory profile. Treatment reduced them and lowered inflammatory signals in the brain.
• It worked from the bloodstream too, more slowly. CAR T cells given into a vein improved memory after six months instead of 2.5. They removed uPAR-carrying blood-forming cells in the bone marrow, which become more common with age, and the remaining ones looked less inflamed and more “young”.
• The bone marrow is the source. Old mice that received bone-marrow cells from CAR T-treated old donors had fewer inflamed macrophages in the brain and better memory, similar to old mice given cells from young donors. In the dish, the offspring of uPAR-carrying marrow cells were more inflammatory.
• Hints in people. In human datasets, the uPAR gene was more active in brain-border macrophages of old people and of people with Alzheimer’s, and in bone-marrow cells of older donors.

What to keep in mind

• Mice, and not yet peer-reviewed. None of this has been tested in people; the human data only show where the gene is active.
• Small and not blinded. Groups were mostly 5 to 18 mice, and the bone-marrow transplant experiment used 5 or 6 per group and was done once. The methods state that mice were not randomly assigned and that data collection and analysis were not blinded.
• The authors’ own caveats. They say it is unexplored how these macrophages interact with other brain cells to keep inflammation going, and that the spinal-fluid route may partly work by reaching bone marrow in the skull.
• A demanding kind of treatment. In people, CAR T therapy is a prescription-only hospital treatment with serious possible side effects. The lead researcher is named on patent applications for CAR T cells that target worn-out cells.

Why it matters

The study suggests an unexpected path from bone to brain: with age, some blood-forming cells in the bone marrow become inflamed and pass that state on to the macrophages they produce, which then settle at the brain’s borders. Removing those marrow cells, without touching the brain, improved memory in mice. That reframes part of brain aging as a problem of the blood system, a target that may be easier to reach than the brain itself. It builds on earlier work with uPAR-targeted immune therapies, but whether any of this is safe or useful in people is entirely open.