Last spring, scientists in South Korea wanted to find out what would happen if they persuaded the brain to trim some of its own wiring. Could removing connections between brain cells change how well mice learned and remembered?

The researchers put some of the brain’s own cleanup crew to work, getting cells called astrocytes to remove connections between nerve cells. They injected modified viruses carrying instructions that made selected nerve cells grow a molecular tag and nearby astrocytes grow a protein that could grab it. The technique encouraged the astrocytes to remove material from the nerve cells’ connections. 

They tried it in the brain’s memory center in mice and published their findings in “Nature Communications” in April 2026.

Then came a frightening lesson. Researchers gave the mice a mild electric shock inside a chamber and brought them back the next day. The animals froze in place, a defensive response that suggested they remembered where they’d been hurt.

Three weeks later, the mice with fewer brain connections were still spending more time frozen when researchers returned them to the chamber. They seemed to remember the shock better than the other mice did. The surprise was how they got there, with roughly 30% of the connections removed from the brain circuit researchers had targeted.

The South Korean researchers are part of a growing effort among scientists around the world to rewire living brains with cutting-edge technology and ingenuity. This year’s Nobel Prize in medicine went to Stanford’s Karl Deisseroth and two researchers based in Germany, Georg Nagel and Peter Hegemann. Their work helped establish optogenetics, which lets scientists use light to turn selected nerve cells on or off.

Researchers hope that physically changing connections within the brain could eventually help treat depression, PTSD and other mental health issues. But the prospect of altering how people respond to their own memories with sci-fi-esque techniques also raises questions about how far treatment should go — particularly if an unwanted change can’t be undone.

“We are beginning to develop tools to change how specific neurons communicate, rather than simply increasing or decreasing the activity of large regions of the brain,” Daniel Colón-Ramos, a professor of neuroscience and cell biology at Yale School of Medicine, told The Post. 

At Duke University, psychiatrist Dr. Kafui Dzirasa and his collaborators are developing a technique called LinCx. It involves injecting modified viruses into selected brain areas to deliver genetic instructions. The cells then make engineered proteins that fit a matching partner, creating an extra route for signals to travel. It’s like installing a new communication line between two computers.

The hope is that doctors could someday use new connections to work around a faulty brain circuit, much as heart surgeons do with a blocked artery.

“We envision it being implemented the way [coronary artery bypass surgery] is,” Dzirasa told The Post. “We aim to permanently restore circuit function so that repeated administration isn’t required.”

But, there’s a catch. The proteins could pair up with ones already in the brain, creating connections that scientists hadn’t intended. To help prevent that, the team designed two proteins more likely to attach to each other.

Thus far, Dzirasa’s team has used the new connections to change how mice handle stress and behave around other mice. In one experiment, they treated young mice prone to grooming themselves excessively, and the behavior never developed. 

Now his team is trying it in mice that have already begun grooming themselves excessively. They want to see whether the new connections can help once the behavior has taken hold.

Dzirasa hopes doctors could someday help people withstand stress before it leads to depression, but that’s still a ways off.

“There is an enormous amount of work between demonstrating a technology in mice and using it safely in people,” he said.

Colón-Ramos’ lab has had success rewiring brain circuits in worms, which have a simple brain centered on a ring of nerve connections near the head. 

For a study published in 2018, his researchers trained worms to prefer cooler temperatures. When allowed to choose where to go, the worms sought out a cooler spot, but adding an electrical connection between two kinds of nerve cells changed their direction. The worms now crawled toward warmth, despite what they had learned. That one change in their wiring had overridden their earlier preference.

“We didn’t create a new memory of a warmer environment,” Colón-Ramos said. “Instead, we changed how the nervous system used its previous experience to make a decision.”

Others use young zebrafish for brain experimentation. Their nearly transparent bodies give researchers a window into the nervous system and let light reach the cells they want to control.

Scientists can also disconnect brain cells, and the results can be just as surprising. Sangkyu Lee, a neuroscientist at South Korea’s Institute for Basic Science, helped develop SynTrogo, the tool used in the mice that remembered the shock chamber more strongly. 

Researchers injected modified viruses into the part of the brain involved in memory, delivering instructions for two proteins. Selected nerve cells made one protein on their surface, and nearby support cells called astrocytes made the other, which could latch onto it. In lab-grown cells, the team watched astrocytes nibble pieces off the tagged nerve cells, taking some of the material at their connections too.

In mice, the intervention reduced connections in the targeted circuit. Those that remained grew larger and became better able to release chemical signals. They could also strengthen more readily, suggesting the circuit had become more prepared to learn.

The mice showed no significant change in working memory under the conditions tested. Their stronger recall of the shock chamber was a specific result, and Lee cautions that removing too many connections could damage brain function. Nobody should take this as permission to improve their memory by losing brain wiring.

For Don Arnold, a professor of biological sciences at the University of Southern California, disconnecting cells raises an enticing possibility for treating PTSD. He wonders if cutting certain connections could allow someone to recall a traumatic event without reliving the fear. The memory would remain, but minus the anxiety and terror. Think “Eternal Sunshine of the Spotless Mind,” only you’d still remember your ex.

Arnold believes this is possible “in principle,” though his team has yet to test whether it’s practical.

The ethical debate predates these tools. In a 2011 interview with PBS’s NOVA, bioethicist Arthur Caplan discussed the prospect of treatments that alter troubling memories. “Our personal identity is deeply tied into our memory,” he said.

Caplan also argued that relieving disabling memories could justify changing them. The difficult question, he suggested, was where to draw the line between treating serious suffering and removing experiences simply because they were unpleasant.

Arnold’s tools are proteins designed in the lab that researchers get selected brain cells to make. They break down other proteins that keep certain connections working, stopping messages from getting through those connections. Researchers can choose which brain cells get the treatment, but they have to shut down a whole type of connection on those cells. They can’t yet pick out just the connections tied to a traumatic memory.

Fear helps us recognize danger, so doctors would have to be careful about which memories they tampered with, Arnold said. There’s also the problem of getting a memory back if something went wrong. He knows of no way to do that once it’s been erased.

The other researchers have their own control problems to solve. Lee’s team has yet to demonstrate an undo function, and Dzirasa says reversibility remains a work in progress. Before human testing, scientists would also need to establish long-term safety and figure out which connections need changing in each patient.

“In psychiatry, two people with the same diagnosis may have very different underlying circuit dysfunction,” Dzirasa said. A successful treatment would require understanding what has gone wrong in that particular brain.

So far, scientists have changed brain connections in laboratory animals and watched their behavior change. Getting this to work safely in people will take much more research. Dzirasa hopes the payoff will be a treatment that fixes what keeps going wrong.

“Our long-term goal is to develop approaches that restore healthier circuit function, rather than simply suppressing individual symptoms,” he said.

Share.