Mice trained to fear one room and feel safe in another did not keep the two memories apart in the hippocampus. The brain’s maps of the dangerous place and the harmless place grew more alike as training went on, according to a team led by Robert Rozeske and Mark Brandon. The researchers read the overlap as a trade of precision for speed.
AI, explained every day: The Deep View is the go-to daily newsletter for 600,000+ founders, knowledge workers and professionals who want to stay up to date with artificial intelligence.
We may earn a commission if you sign up.
The finding runs against a long-held assumption that fear sharpens the line between threat and safety. Brandon, a McGill University psychiatry professor, said the results suggest fear may blur that boundary rather than sharpen it, and that the work could change how scientists think about anxiety and post-traumatic stress disorder.
An augmented reality arena for teleporting mice
The experiment was built around a trick. Rather than carrying animals from one cage to another, the team created an augmented reality arena in which sights and sounds could be switched on the spot, so a mouse was effectively teleported between a context tied to danger and one tied to safety. The method is described in ScienceDaily’s write-up of the McGill release, which carries the study’s findings and quotes.
Each mouse learned to associate one environment with danger and a different one with safety. Because the cues were digital, the researchers could flip an animal between the two contexts within a single session, which is what gave them a live view of the switch rather than a comparison of separate recordings made on different days. Mice stayed in place physically while the arena around them changed.
Tiny head-mounted microscopes watched hundreds of neurons in the ventral hippocampus while the animals moved through the two contexts. That region handles emotionally charged experience, and the recordings let the team follow, in Rozeske’s words, a moment-by-moment view of the brain flipping between competing maps of dangerous and safe memories. Neither the ScienceDaily page nor the McGill text gives an exact neuron count or number of mice.
The paper, “Overlapping representations in the ventral hippocampus support rapid fear memory retrieval,” appeared in Nature Neuroscience and carries the DOI 10.1038/s41593-026-02435-5. Six authors are listed: Rozeske, Léonie Runtz, J. Quinn Lee, Alexandra T. Keinath, Aaron Sossin and Brandon.
Overlap as a better-safe-than-sorry shortcut
Over training, the neural patterns for the threatening room and the safe room drifted toward each other instead of pulling apart. The authors argue the similarity helps the brain pull up a threat memory quickly, because a pattern that partly matches many situations can fire fast without waiting for a careful comparison. Rozeske, an assistant professor in the University of Toronto’s psychology department, said the ventral hippocampus may belong to a system that prioritizes a rapid “better safe than sorry” response. The cost is obvious: a brain tuned that way will sometimes treat a harmless place as risky.
The practical logic is easy to follow. A memory map that resembles both a risky place and a safe one is cheap to reactivate, because a partial match is enough to trigger it, and speed is what an animal needs when a predator might be near. According to the McGill team, the overlap appeared to help the brain recall earlier threats faster and respond more quickly, which they treat as a possible survival strategy and not a design flaw. A system that separated every context perfectly would be slower and, on a bad day, deadlier.
That trade-off is the core claim, and it is an interpretation of recordings made in mice. The release does not report effect sizes or statistical detail, so readers who want the strength of the effect will need the full paper.
The PTSD and anxiety link remains a hypothesis
Brandon went one step further, describing responses seen in PTSD as possibly an overactive version of a normal survival mechanism gone into overdrive. If that is right, anxiety disorders could reflect a useful blurring pushed past its useful range, not a broken system.
The idea has an appeal for clinicians, since it offers a way to think about why people with anxiety disorders or trauma histories often react to cues that look only loosely like the original danger. Brandon told the release audience the work could reshape how scientists think about both conditions. Reshaping, though, needs evidence from people, and mouse hippocampal recordings from a virtual arena are a long way from a veteran in a crowded street or a survivor in a stairwell.
The authors are explicit that human research is needed to learn whether the same mechanism contributes to anxiety or trauma-related conditions. Nothing in the mouse experiments tested a clinical condition, and the PTSD connection is a proposal the team offers for others to test. The McGill release, distributed on October 15 through Education News Canada, names the Canadian Institutes of Health Research and the Natural Sciences and Engineering Research Council of Canada as funders. The open question for the next round of work is whether people show the same overlap in the ventral hippocampus when a safe place begins to look like a dangerous one.
This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.
AI, explained every day: Get The Deep View →
We may earn a commission if you sign up.
More from Morning Overview
- Hybrids have 15% fewer problems than gas cars, while EVs and plug-in hybrids have about 80% more, Consumer Reports finds
- Regulators cleared the first U.S. small modular reactor, 4 months early
- Lake Powell sank to a record 3,517 feet, nearing the level that stops Glen Canyon Dam’s turbines
- Seven of Earth’s nine planetary boundaries are breached, and all seven are worsening
