Exercise's benefits likely lie in its ability to trigger a coordinated, whole-body response.
For generations, the wisdom that movement preserves the mind has been accepted without a full accounting of why. A study published in Cell now offers a partial answer: the liver, during exercise, releases a protein called GPLD1 that repairs the aging brain's most critical gateway — the blood-brain barrier — reversing memory loss in mice and reducing the hallmarks of Alzheimer's pathology. The discovery does not promise a pill that replaces the walk, but it illuminates one thread in the vast, coordinated conversation the body holds with itself when it moves.
- Memory loss in aging is not simply inevitable decay — it is partly driven by a leaky blood-brain barrier, corrupted by an accumulating protein called TNAP that inflames and weakens the brain's protective gateway.
- Researchers at UCSF identified GPLD1, a liver-released enzyme triggered by exercise, as the agent that strips TNAP away and restores barrier integrity — reversing cognitive deficits in aging mice with striking consistency.
- In mice engineered to model Alzheimer's disease, GPLD1 injection reduced amyloid plaques, tightened blood vessels, and recovered cognitive performance to levels comparable to those achieved by months of voluntary running.
- A parallel discovery from South Korean researchers found a separate muscle-secreted exerkine that grows new neurons in the hippocampus — suggesting exercise speaks to the brain through many voices, not one.
- Scientists urge measured optimism: mouse models cannot replicate the decades-long complexity of human Alzheimer's, and no single molecule is likely to capture what the whole body achieves through sustained movement.
The brain ages in ways that feel personal and inevitable — names dissolve, curiosity dims, memory falters. For decades, researchers knew exercise slowed this decline without fully understanding how. A new study published in Cell has identified a specific protein, GPLD1, released by the liver during physical activity, that appears to reverse memory loss in aging mice and those modeling Alzheimer's disease.
When the body moves, the liver and muscles release chemical messengers called exerkines into the bloodstream. Gregor Bieri and colleagues at the University of California, San Francisco, focused on GPLD1. Older mice given access to running wheels showed elevated GPLD1 levels; those kept sedentary did not. When researchers injected GPLD1 directly into aging mice, cognitive deficits reversed — the animals regained memory and curiosity that had faded with age.
The mechanism centers on the blood-brain barrier, a tightly controlled network that governs what enters the brain from the blood. With age, a protein called TNAP accumulates in this barrier — particularly around the hippocampus — making it leaky and inflamed. GPLD1, it turns out, is an enzyme that cleaves TNAP away, restoring the barrier's integrity. The researchers confirmed this causal chain through three interlocking experiments: aged mice with high TNAP had leaky vessels; young mice artificially given high TNAP developed the same cognitive deficits as old ones; and when GPLD1 was injected alongside elevated TNAP, its protective effect disappeared entirely.
The team then applied this logic to Alzheimer's. Genetically modified mice engineered to accumulate amyloid — the toxic protein central to Alzheimer's pathology — were either allowed to run or given direct GPLD1 injections. Both interventions reduced TNAP in the blood-brain barrier, decreased amyloid plaque burden, and restored cognitive function. The results are promising, though researchers caution that mouse models cannot replicate the full complexity of human Alzheimer's disease, which unfolds over decades through overlapping biological, genetic, and lifestyle factors.
GPLD1 is not the only exerkine reshaping this field. Researchers in South Korea recently identified a separate molecule, secreted by skeletal muscle, that improves cognition by stimulating new neuron growth in the hippocampus — a different pathway entirely. This multiplicity suggests that exercise's benefits to the brain are not channeled through a single molecule but through many coordinated systems working in concert. As one researcher put it plainly: we are not yet close to replacing exercise with a pill. The body's response to movement remains too intricate, too whole, for any single intervention to replicate.
The brain ages. Names slip away. Thoughts scatter. Habits calcify. For decades, researchers have known that exercise slows this decline, but the mechanism remained opaque—a black box of cardiovascular benefit and metabolic improvement. A new study published in Cell has cracked open that box, identifying a specific protein released by the liver during physical activity that appears to reverse memory loss in aging mice, and potentially in those modeling Alzheimer's disease.
When we exercise, our bodies release small molecules called exerkines into the bloodstream. These chemical messengers—peptides, lipids, hormones, nucleic acids—travel throughout the body, mediating the broad benefits we associate with physical activity: better cardiovascular fitness, improved glucose metabolism, enhanced insulin sensitivity, stronger immune function. But exerkines do more than that. They speak directly to the brain. Atefe Tari, a researcher at the Norwegian University of Science and Technology studying exercise and brain health, explains that exercise "affects the brain more directly by many different processes, including changes in signaling molecules."
Gregor Bieri and colleagues at the University of California, San Francisco, focused on one particular exerkine produced by the liver: a protein called GPLD1. They gave older mice access to running wheels and measured GPLD1 levels in their livers. Mice that exercised voluntarily showed elevated GPLD1. Mice whose wheels were locked showed lower levels. Then came the cognitive tests. Young mice navigated a water maze with ease, finding a hidden platform quickly and with few errors. Old mice stumbled. They showed no preference for novel objects or unfamiliar spaces—the kind of curiosity that marks a sharp mind. But when researchers injected GPLD1 directly into the livers of aging mice, something shifted. The cognitive deficits reversed. Memory improved.
The mechanism turned out to involve the blood-brain barrier, a tightly regulated network of blood vessels and specialized cells that controls what enters the brain from the bloodstream. With age, this barrier accumulates proteins that compromise its integrity, making it leaky. One such protein, called TNAP, accumulated heavily in the barriers of old mice, particularly around the hippocampus—the brain's learning and memory center. TNAP drives inflammation and impairs the transport of beneficial factors from blood into brain tissue. The researchers ran three experiments to establish the causal chain. First, they confirmed that old mice with high TNAP levels had leaky blood vessels. Second, they artificially increased TNAP in young mice and watched as those young animals developed leaky barriers and cognitive defects matching those of aged mice. Third, they injected GPLD1 into old mice while simultaneously boosting TNAP in their blood-brain barriers—and the protective effect of GPLD1 vanished. The picture crystallized: GPLD1, an enzyme that breaks down surface proteins, was cleaving TNAP from the blood-brain barrier, restoring its integrity and protecting cognition.
The team then tested whether this mechanism might apply to Alzheimer's disease. They used genetically modified mice engineered to accumulate amyloid protein, the toxic hallmark of Alzheimer's pathology. Some of these mice were allowed to run on wheels for three months. Their GPLD1 levels rose, and their cognitive performance improved markedly compared to sedentary Alzheimer's mice. In another cohort, direct GPLD1 injection reduced TNAP in the blood-brain barrier, decreased leakiness, lowered amyloid plaque burden, and restored cognitive function. The results are encouraging, though researchers urge caution. Mouse models, as Tari notes, do not capture the full complexity of Alzheimer's disease in humans—a condition that develops over decades and involves vascular dysfunction, inflammation, metabolic changes, genetic factors, and lifestyle influences.
The implications extend beyond GPLD1 itself. Recent research from Seoul National University and the Korea Brain Research Institute identified a different exerkine, secreted by skeletal muscles, that improves cognition through a separate pathway—by boosting new neuron formation in the hippocampus. This suggests that exercise's cognitive benefits do not flow through a single molecular channel but rather through multiple coordinated systems. Bieri sees opportunity in this multiplicity: as therapeutic strategies expand, discovering complementary mechanisms could enable personalized medicine approaches, with future treatments combining different molecules to target different aspects of aging and disease. For now, though, experts are clear on one point. "It does not mean we are close to replacing exercise with a pill yet," Tari said. The body's response to movement is too intricate, too whole, for any single intervention to replicate.
Citações Notáveis
Exercise affects the brain more directly by many different processes, including changes in signaling molecules.— Atefe Tari, Norwegian University of Science and Technology
It does not mean we are close to replacing exercise with a pill yet.— Atefe Tari
A Conversa do Hearth Outra perspectiva sobre a história
So GPLD1 is released by the liver during exercise. But why the liver? Why not the brain directly?
The liver is a metabolic hub. It responds to systemic signals during exercise and releases exerkines into the bloodstream, which then circulate throughout the body. The brain-liver connection is ancient in evolutionary terms. The liver senses the body's state and communicates back to the brain through chemical messengers.
The blood-brain barrier becomes leaky with age. That sounds dangerous. What exactly leaks through?
Inflammatory molecules, immune factors, and other blood-derived substances that normally stay out. When they enter the brain, they trigger chronic inflammation in nerve cells, which accelerates neurodegeneration. The brain also becomes less efficient at clearing toxic proteins like amyloid.
In the Alzheimer's mice, GPLD1 reduced amyloid plaques. Does that mean it's clearing the plaques, or preventing new ones from forming?
The study doesn't fully distinguish between those mechanisms. What we know is that when the blood-brain barrier is restored—when TNAP is broken down and the barrier becomes less leaky—the brain's internal environment stabilizes. That may help clear existing plaques or prevent new accumulation, or both.
You mentioned multiple exerkines working through different pathways. Could they work together?
That's the real frontier. Exercise triggers a coordinated whole-body response. One exerkine repairs the barrier. Another boosts neurogenesis. A third might improve metabolism. The strength likely lies in that combination, not in any single molecule.
Why are researchers cautious about moving this to human trials?
Mouse Alzheimer's is artificially induced and develops rapidly. Human Alzheimer's unfolds over decades, involves genetics, vascular disease, lifestyle, inflammation—layers of complexity a mouse model can't capture. What works in mice often fails in humans.