A new study in mice has turned the mystery of memory persistence on its head. The animals lost more than half of their synapses during a state resembling hibernation, yet their memories remained completely intact, researchers report.
Kazumasa Tanaka, a neuroscientist at the Okinawa Institute of Science and Technology Graduate University in Japan, said the arrangement of connections in the brain can change dramatically from day to day. To understand how a memory lasting years can sit on hardware that shifts every few days, his team made the shift more extreme. In a paper published in Science, they induced a hibernation-like state in mice that erased the state of more than half of their synapses, and the mice still remembered.
Hibernation is associated with squirrels, hamsters, and bears, but the neural circuit that triggers it is conserved across mammals and exists in species that never hibernate in the wild, including mice. In June 2020, Takeshi Sakurai, a neuroscientist at the University of Tsukuba and a collaborator on this study, developed a technique to artificially activate this circuit. The method turns on a population of neurons called Q neurons in a region of the hypothalamus.
The resulting state is called QIH, short for Q-neuron-induced hypothermia and hypometabolism. With the protocol, Tanaka explained, a mouse’s body temperature drops to around 20 degrees Celsius, and heart rate and breathing slow significantly. Whether that qualifies as true hibernation depends on the reference point. Bears lower their metabolic demand but keep their body temperature around 36 or 37 degrees Celsius. Some squirrel species enter a deeper hibernation with body temperatures close to freezing. Artificial hibernation falls somewhere in between.
The crucial feature of QIH is that it can be switched on and off at will. In Tanaka’s experiments, mice spent 48 hours in the hibernation-like state and then woke up. For their synapses, those 48 hours worked like a purge.
To measure synapse loss, the team implanted tetrodes, bundles of fine electrodes, into the hippocampus of freely moving mice. These allowed them to record individual neurons firing. Activity dropped by about 70 percent once hibernation set in.
Some mouse brains were imaged with serial block-face scanning electron microscopy before, during, and days after the state. The imaging showed that hibernation eradicated more than half of the synapses. In principle, that should erase most memories. If memory traces reside in the strength of individual synapses, losing more than half would be expected to impair memory afterwards, Tanaka said. But the team found no such impairment.
Before hibernation, the mice were trained on two standard memory tasks. One was contextual fear conditioning, where an animal learns to associate a particular box with a mild electric shock. The other was a plus-maze task, where the mouse learns to navigate to a reward. Both tasks depend on memories stored in the hippocampus, which the team confirmed by creating a lesion in that region after training, causing the memories to disappear.
When other mice that had been put into QIH were aroused, they performed on these tasks just as well as mice that had not hibernated. In these two different behavioral paradigms, the memory was completely intact, Tanaka said.
Brain activity recordings confirmed the survival of memories. Place cells, hippocampal neurons that fire when an animal is in a particular location, still fired in the same places after arousal. A decoder that read population activity could reconstruct the mouse’s position just as accurately as before.
Watching the same dendrites over eight days showed that synapses lost during hibernation reappeared after arousal, and 82 percent of them returned to the same spot on the same dendrite they had occupied before, far above what chance would predict.
The vanished synapses were not a random sample. Using a technique called eGRASP, which makes a connection glow green where two neurons tagged during learning are connected, the team looked specifically at engram synapses, the specialized connections between memory-storing neurons. Engram synapses sitting alone on a dendrite were eliminated by hibernation, while those arranged in tight spatial clusters were preserved. Why clustering offers protection is not yet understood. We don’t know the mechanism, Tanaka said. That is one of the ongoing projects in the laboratory.
The team could, however, examine the architecture of the surviving clusters. They found that a third of the clustered engram synapses were attached to a structure called a multisynaptic bouton. Usually, a single presynaptic terminal connects to a single postsynaptic spine, a one-to-one relationship. In multisynaptic boutons, one presynaptic terminal makes connections with multiple postsynaptic spines. This is a rare structure and difficult to find in the brain, Tanaka said.
In randomly chosen synapses from non-hibernating mice, only 3.3 percent sat on a multisynaptic bouton. The clustered engram synapses in the hibernating mice were, if anything, slightly smaller than their neighbors, not the enlarged, strengthened connections the classic model would predict.
To ensure the pattern tracked memory rather than a general pattern of synapse loss, the team ran a negative control. They placed mice under long-term anesthesia combined with cytochalasin D, a drug that blocks the enlargement and stabilization of synaptic connections. That combination, like QIH, suppressed neuronal firing and stripped away a comparable share of synapses, but the mice came out with impaired fear memory. In those animals, the clustered engram synapses were destroyed indiscriminately, just like every other synapse.
Tanaka cautioned against drawing strong conclusions from the comparison. That is a major limitation of this study, he said. There is currently no way to manipulate the clustering of engram synaptic connections without compromising other aspects of the synapses and the network. In other words, researchers cannot selectively turn off these clusters to see if memories disappear with them.
The study is associative rather than a test of causality, Tanaka said. But if the findings hold, they may indicate that a memory does not depend on any particular synapse to survive, as long as the broader neural architecture around it is preserved.
Tanaka’s lab is already considering a stranger and potentially more profound implication. If the brain rebuilds itself after hibernation, does it rebuild the state it was in beforehand, or some preferred configuration of its own, he asked. To explore this, the team induced brief artificial hibernation in mice engineered as an epilepsy model, just before seizures had developed. The development of epilepsy was completely suppressed after hibernation, even though no additional manipulations took place, Tanaka said. This suggests the brain returns to a kind of default network state, something like factory settings, rather than simply resuming where it left off.
The epilepsy findings are still unpublished, but Tanaka hopes to include them in a follow-up paper. Even if confirmed, clinical applications in humans are far off. There are many challenges remaining, including more rigorous safety measures and ethics, Tanaka said. All studies so far have been done in mice. We need to move on to rats or monkeys and see whether artificial hibernation actually affects brain function. We still have so many things to do.
The study, published in Science with the DOI 10.1126/science.aee7004, offers a tantalizing glimpse into how memories might be preserved even when the brain undergoes dramatic physical changes. While the mechanism remains unknown, the findings suggest that the architecture of clustered synaptic connections could hold the key to enduring memory, raising questions that reach far beyond hibernation itself.
