Hibernation erased more than half of mice's synapses. The memories stayed.
A Science study from Kazumasa Tanaka's lab at the Okinawa Institute of Science and Technology Graduate University has put the leading theory of memory storage under real pressure. Tanaka's team induced a hibernation-like state in mice, erasing more than half of their synaptic connections. The mice retained their memories anyway.
Key takeaways
- A Science study from Kazumasa Tanaka's lab at the Okinawa Institute of Science and Technology induced a hibernation-like state in mice that erased more than half of their synaptic connections, yet the mice retained their memories.
- The finding challenges the dominant model that memories are stored in stronger, physically enlarged connections between neurons.
- The hibernation-like state was triggered using a technique developed by Takeshi Sakurai in June 2020 that artificially activates Q neurons in the hypothalamus.
- The result does not explain how memories survive on shifting synaptic hardware, but makes that question harder to sidestep and suggests the physical-enlargement model needs revision or an unaccounted-for storage mechanism.
- The next confirmable step would be follow-on work identifying where the memory trace resides and the actual storage mechanism.
A Science study from Kazumasa Tanaka's lab at the Okinawa Institute of Science and Technology Graduate University has put the leading theory of memory storage under real pressure. Tanaka's team induced a hibernation-like state in mice, erasing more than half of their synaptic connections. The mice retained their memories anyway.
The dominant model holds that memories form when the connections between neurons involved in a learning event grow stronger and physically larger. The trouble is that those connections don't hold their configuration. Tanaka has described the arrangement of synaptic connections on day one as very, very different from the same arrangement on day four or five. The hardware shifts constantly, even as the memories sitting on it apparently do not.
How the team triggered hibernation in mice
Hibernation is a specialty of squirrels, hamsters, and bears. The neural circuit behind it, though, is conserved across mammals and is present even in species that never hibernate in the wild. Mice fall into that category. In June 2020, Takeshi Sakurai, a neuroscientist at the University of Tsukuba and a collaborator on the Tanaka study, developed a technique to activate that circuit artificially. The method works by targeting a population called Q neurons in a region of the hypothalamus.
Tanaka's team used Sakurai's technique to induce the hibernation-like state and then measured what happened to synaptic structure. More than half of the synapses were significantly altered. Coming out of that state, the mice appeared to have kept their memories.
What the result means for the field
The finding does not answer how memories survive on shifting hardware. It makes the question harder to sidestep. If erasing more than half the synaptic configuration leaves memory intact, the physical-enlargement model either needs revision or there is a storage mechanism the field has not accounted for.
What to watch is whether Tanaka's lab can locate where the memory trace resides when the synapses that were supposed to hold it are gone. The Science publication is the current milestone. Follow-on work identifying the actual mechanism would be the next confirmable step.
Filed by the newsroom of MarketPR on August 22, 2026. Source: arstechnica.com. Indicative figures are not investment advice.