Lost brain support cells were thought irreplaceable — adult brains are replacing them

When a patch of the brain is destroyed, the surviving support cells at the wound’s edge hold position. They divide — and then send only their daughter cells’ nuclei sliding through their own cellular extensions into the damaged area to rebuild what was lost. Researchers at the University of Zurich caught this happening in living mouse brains, over weeks, using a microscopy technique that can image deep tissue in real time. No one had reason to expect it, because the adult brain was not supposed to do this at all.

The cells in question are astrocytes, the star-shaped glial cells that make up roughly half the brain’s volume. They feed neurons, regulate blood flow, anchor the blood-brain barrier, and help maintain the chemical environment that keeps neurons firing. When astrocytes are destroyed — by a blow to the head, by stroke, by autoimmune attack — the damage was thought permanent. Unlike most peripheral tissues, the mature brain was believed to lack the machinery to replace its structural support.

That assumption has now been overturned.

Recommended ReadingScientists mapped a fruit fly’s 160,000 neurons — and its legs run themselves

How they found it

The team used two-photon microscopy, a technique that fires infrared laser pulses into tissue to generate high-resolution images at depths unreachable by conventional light microscopes without destroying the sample. This allowed Bruno Weber, Marina Herwerth, Matthias Wyss, and their colleagues to observe the same patch of brain repeatedly over several weeks — tracking individual cells and their components as events unfolded.

They produced two types of damage in mouse brains: focal traumatic lesions with a diameter of just under 0.5 millimeters, and autoimmune injury modeled on neuromyelitis optica spectrum disorder, a condition in which antibodies target and destroy astrocytes. In both cases, the response at the wound’s edge was the same.

Surviving astrocytes near the lesion formed a distinct subpopulation. They divided, but their daughter cells did not migrate as intact units into the damaged zone. Instead, the parent cell held its position, and the daughter cell’s nucleus — separated from most of the cell body — traveled through the parent’s existing cellular processes directly into the injury site. Bruno Weber described the process as the newly formed nuclei of their daughter cells gliding across long distances to repopulate the damaged area of the brain and knit the astrocyte network back together. The team’s paper calls this nuclear translocation.

What this upends

Astrocytes can, after injury, proliferate and form a glial scar — a dense mesh that walls off the lesion and limits further spread of damage. That reactive behavior was already known. What the new study identifies is qualitatively different: a subpopulation of astrocytes that reconstitutes the lost interior of a lesion, not merely its border.

The mechanism is also unusual by the standards of cell biology in general. Nuclei moving long distances through a parent cell’s extensions — nuclear translocation — occurs during development in some specialized cell types, but had not been documented in adult brain repair before this study.

The team mapped gene activation across the repair site, identifying what Weber describes as numerous genes and signaling pathways that are temporarily activated during repair. These represent potential starting points for therapies aimed at triggering or amplifying this process in contexts where the natural response is insufficient.

What this discovery doesn’t settle

The most important limitation is the one that applies to nearly every neuroscience study: the experiments were done in mice. The damage types tested model human conditions but are not those conditions themselves. The adult mouse brain and adult human brain differ in astrocyte density, overall scale, and the degree of tissue destruction that follows a significant human stroke or traumatic brain injury.

The lesions studied were deliberately small — under half a millimeter across. Whether nuclear translocation scales to larger areas of damage, or whether it fails to activate past a certain threshold of destruction, remains untested. The repair window — how long after injury these regenerative astrocytes remain active — is also unresolved. A narrow window would significantly constrain any eventual therapeutic use.

The specific genes and pathways activated have not yet been individually validated as targets. The path from a candidate gene list in a mouse experiment to a drug that safely manipulates the same process in a human patient involves years of preclinical work and substantial failure rates at each stage.

Common questions about brain self-repair

How do astrocytes support neurons, and why does losing them matter?

Astrocytes absorb excess neurotransmitters after synaptic firing, supply neurons with metabolic fuel, regulate ion concentrations in the synaptic space, and help form the blood-brain barrier. When a patch of astrocytes is lost, neurons in the same area lose their local life-support infrastructure. Long-term, this contributes to cognitive or motor deficits depending on where the damage falls.

What exactly is nuclear translocation?

Nuclear translocation is the movement of a cell’s nucleus from one position to another. In the repair sequence the Zurich team observed, it refers to the daughter cell’s nucleus detaching from most of its cell body and traveling through the parent astrocyte’s long cellular extensions into the damaged zone. This is distinct from cell migration, where the entire cell moves. In this case, the parent astrocyte stays in place; only the nucleus travels.

Could this mechanism eventually be used to treat stroke or traumatic brain injury?

That is the direction the research points, but the distance is substantial. The team identified genes and pathways activated during repair — these are candidate targets for drugs or gene therapies that could amplify the natural response. No intervention has yet been tested even in animals. Weber’s group describes the gene list as starting points for influencing post-disease and post-injury regeneration processes. Any human application remains years away.

Why had this never been seen before?

Two-photon microscopy only recently reached the spatial resolution and tissue depth needed to track individual cellular components in a living brain over weeks. Earlier methods either lacked resolution, required killing the animal to take each image, or could not follow the same region repeatedly. The University of Zurich team’s continuous real-time observation of the same lesion — from the moment of damage through weeks of repair — was what made the translocation visible at all.

Weber’s group has begun work on identifying which of the activated signaling pathways can be selectively manipulated to amplify the repair. The immediate goal is to find levers that make a process the brain initiates naturally fast enough, or extensive enough, to reduce the deficit left by injury or autoimmune disease.

Reference: Herwerth M et al., “Focal astrocyte loss reveals nuclear translocation during lesion repopulation,” Nature Neuroscience 29, 1826 (2026). DOI: 10.1038/s41593-026-02354-5

Tags: neuroscience, astrocyte, brain repair, Bruno Weber, Nature Neuroscience, nuclear translocation

More Like This

Under anesthesia, the brain kept decoding speech and predicting the next word

A speck of metal with 10,000 atoms was held in two places at once

Ultra-processed food is quietly dismantling your cellular architecture

Human Brain Project study offers insights into neurotransmitter receptor organisation

A planet nursery was caught spinning for the first time, and parts ignore the rules

NASA Gets Unusually Close Glimpse of Black Hole Snacking on Star