Scientists create a tool to ‘edit’ brain functions and improve memory

The brain is often described as a dense forest of connections, and for years scientists have searched for ways to trim that forest with precision. They have learned how to turn neural activity up or down. Reworking the physical wiring itself has been much harder.

Now a team in South Korea says it has built a tool that can do just that.

Researchers at the Institute for Basic Science and the Korea Brain Research Institute developed a system called SynTrogo, short for Synthetic Trogocytosis, that lets astrocytes, the star-shaped support cells wrapped around synapses, selectively reduce synaptic connections in a chosen brain circuit. In mice, the method cut excitatory synapse numbers in a memory-related hippocampal pathway by about 27 percent. Yet instead of weakening the circuit, the remaining connections became stronger, more plastic, and better at supporting memory.

“This is the first demonstration that brain circuits can be directly edited by engineering physical interactions between neurons and astrocytes, independent of neuronal activity,” said Dr. LEE Sangkyu of the IBS Center for Memory and Glioscience. “It opens the possibility of ‘connectome editing’ and provides a new platform for studying and reshaping the physical architecture of neural circuits.”

Synapses are the contact points where one neuron passes signals to another. Together they form the connectome, the brain’s wiring diagram. Those links are not fixed. Throughout life, some are strengthened, some are weakened, and some are removed in a pruning process that helps circuits stay functional. When pruning goes wrong, either becoming too aggressive or too limited, it has been linked to disorders including schizophrenia, autism spectrum disorder, and Alzheimer’s disease.

Astrocytes already play a role in that natural cleanup. Their branches sit close to synapses, making them well positioned to help reshape circuits. The challenge has been finding a way to direct that kind of structural remodeling in a specific place without also changing the electrical activity of the neurons involved.

SynTrogo was built to do that.

The method uses a lock-and-key style design. In the experiments, neurons in a target circuit were engineered to display a molecular tag on their surface, while nearby astrocytes were engineered to carry a matching binding partner. Once those cells touched, the astrocytes were prompted to take up part of the neuronal membrane and nearby material in a trogocytosis-like process, a kind of cellular nibbling seen in other biological systems.

In cultured cells, the system worked across different cell types, including astrocytes and neurons. The transfer moved in one direction, from ligand-bearing cells to receptor-bearing cells, and it depended on strong binding between the engineered proteins. The work also suggested the process was self-limiting. As ligand-receptor pairs were internalized, the available surface molecules declined, which may help explain why the system did not appear to trigger widespread cell damage.

That self-restraint mattered once the team moved into the mouse brain.

They targeted the hippocampal CA3-CA1 pathway, one of the best-studied circuits involved in learning and memory. The ligand was expressed in CA3 excitatory neurons and the matching receptor in CA1 astrocytes. Three weeks later, excitatory synapse density in the targeted region was significantly lower.

The reduction was selective. Inhibitory synapses did not fall in the same way, and neighboring axons that lacked the engineered ligand did not show the same loss.

At first glance, the result looks counterintuitive. Cut synapses in a memory circuit, and you might expect memory to suffer.

That is not what happened.

Detailed imaging and electrophysiology showed that the remaining synapses did not simply survive. They changed. Presynaptic boutons became larger. They held more synaptic vesicles, including more docked vesicles ready for release. Mitochondrial volume increased in boutons, suggesting added support for the energy demands of stronger transmission. On the postsynaptic side, dendritic spines and postsynaptic density areas also grew, and the proportion of spines containing spine apparatus structures increased.

Electrical recordings pointed in the same direction. Miniature excitatory postsynaptic current frequency fell, consistent with fewer synapses overall. But spontaneous excitatory signaling rose, and the readily releasable pool of synaptic vesicles increased. Long-term potentiation, the cellular strengthening process widely tied to learning and memory, was significantly enhanced.

The researchers also found signs that the remodeled synapses had entered a more learning-ready state. Before fear conditioning, AMPA receptor-mediated responses were reduced. After learning, they recovered to control-like levels. That pattern suggests the altered circuit may have been especially primed for experience-dependent strengthening.

“We found that the brain can adapt and strengthen its function even when the total number of synaptic connections is reduced,” said Dr. LEE Kea Joo of the Korea Brain Research Institute. “This gives us new insight into how neural circuits maintain performance and may offer clues for restoring cognitive function in brain disorders.”

To inspect the structural changes more closely, the team used correlative light and electron microscopy. Those images showed unusually tight interfaces between astrocytes and axons, localized membrane deformation, and partial enclosure of axonal regions by astrocytic processes.

Ligand-labeled axons in contact with receptor-bearing astrocytes had fewer synapses than controls. The structural changes were spatially restricted, and the experiments did not show evidence of broad axonal loss or neuron death even weeks later. Astrocyte and microglia reactivity also did not appear elevated under the reported conditions.

Still, the study stops short of claiming a complete real-time view of synapse removal in the living brain. The authors note that while cultured-cell imaging clearly showed full internalization of cell material, fixed brain tissue cannot reveal the same temporal sequence. They also say the reduction in synapse number could reflect more than one mechanism. Local nibbling may contribute, but impaired synapse formation caused by intensified astrocyte-axon contact could also play a role.

That uncertainty is one of the study’s major limits.

Another is that the effect was context-dependent. Memory improvements varied with the strength of the fear-conditioning stimulus, suggesting the consequences of synapse reduction may not be uniformly beneficial across situations or circuits.

Behaviorally, the mice gave the most striking result.

In contextual fear-conditioning tests, animals with SynTrogo-modified hippocampal circuits showed stronger memory than controls. With mild conditioning, they froze more during both recent and remote memory tests. With stronger conditioning, they maintained their memory over time while control animals showed more decay.

Even so, the mice were still able to extinguish those fear memories when conditions changed. Their freezing responses dropped during extinction training, and the work found no significant changes in locomotion, anxiety, or working memory in the tested conditions.

Director C. Justin LEE of the IBS Center for Memory and Glioscience said the work shows that “selectively reducing a subset of synapses can paradoxically enhance circuit function by promoting adaptive remodeling of the remaining connections.”

That paradox may be the most important part of the study. The adult brain may not always perform best by preserving the largest possible number of synapses. Under some circumstances, carefully removing selected connections seems to push the network toward a leaner and more efficient state.

SynTrogo is still a research tool, not a treatment. But it offers neuroscientists a new way to study how brain circuits reorganize when synapse numbers change, without directly manipulating neuronal firing.

That could matter in disorders where abnormal synapse numbers or disrupted circuit organization are thought to play a role, including autism spectrum disorder, schizophrenia, Alzheimer’s disease, and brain injury. The platform may help researchers test a long-standing question: when synapses are lost or overproduced, how much of the problem comes from the number of connections itself, and how much comes from how the remaining circuit adapts?

The study also points toward a broader idea, connectome editing, in which researchers do not just modulate activity but deliberately reshape the brain’s physical wiring. Much more work is needed, especially real-time studies in living brains and experiments across different circuits. Still, the findings suggest that strategic subtraction can sometimes make a neural network work better, not worse.

Research findings are available online in the journal Nature Communications.

The original story "Scientists create a tool to 'edit' brain functions and improve memory" is published in The Brighter Side of News.

Like these kind of feel good stories? Get The Brighter Side of News' newsletter.