One of the most remarkable features of QIH is that it can be switched on and off at will. In Tanaka’s experiments, mice remained in a hibernation-like state for 48 hours before waking naturally.
At the synaptic level, those two days appeared to function almost like a Thanos Snap for the brain—eliminating a large number of neural connections without destroying stored memories.
QIH hibernation causes a temporary synapse purge
To measure how many synapses disappeared during QIH, Tanaka’s team implanted tetrodes—bundles of thin electrodes—into the hippocampi of freely moving mice. These devices allowed researchers to record the activity of individual neurons. When hibernation began, neural activity fell by approximately 70 percent.
The researchers also used sequential block-face scanning electron microscopy to image brain tissue before hibernation, during the hibernation-like state, and several days after the mice regained consciousness. The images revealed that QIH temporarily eliminated more than half of the animals’ synapses.
In theory, losing so many synaptic connections should erase a significant portion of memory. “If we accept that memory traces reside in the efficacy of individual synapses, then we would naturally expect subsequent memory impairment if we lost more than half of our synaptic connections,” Tanaka says. Yet the mice showed no meaningful loss of memory after waking.
Before entering QIH, the mice were trained on two standard hippocampus-dependent memory tests. In the first, contextual fear conditioning, the animals learned to associate a specific enclosure with a mild electric shock. In the second, a plus-maze task, the mice learned how to navigate toward a reward.
The researchers confirmed that both memories depended on the hippocampus by creating lesions in that brain region after training. The lesions caused the learned memories to disappear.
However, when mice emerged from QIH, they performed both tasks just as well as mice that had not entered the hibernation-like state. “What we found in these two different behavioral paradigms is that memory is completely intact,” Tanaka claims.
Brain-recording experiments provided additional evidence that memories survive QIH-related synapse loss. Place cells—hippocampal neurons that activate when an animal occupies a particular location—continued firing in the same locations after the mice awoke. Decoders analyzing patterns of population activity could also reconstruct where the animals had been with the same accuracy as before hibernation.
These findings suggest that QIH may temporarily remove or remodel synapses while preserving the neural activity patterns that support long-term memory. The study could help researchers understand how the brain protects memories during extreme changes in neural activity and connectivity.
Source: arstechnica.com