Mice Lost Half of Their Brain Synapses. They Still Kept Their Memories

A strange hibernation experiment is changing how researchers think about memory.

A memory, it turns out, may be harder to erase than we thought.

In a striking experiment, mice kept memories formed even when more than half of the synapses in their hippocampus disappeared. Neuronal activity also plunged by roughly 70%, while the memories remained. “It was astonishing,” one of the study authors noted.

That suggests the brain may rely not only on the strength of individual synapses, but also on a more resilient architecture connecting groups of memory cells.

When the brain loses its connections

For decades, neuroscientists have considered synapses (the tiny junctions through which neurons communicate) prime candidates for the physical storage of memory.

When we learn something, particular groups of neurons become active together. Connections between them can strengthen, neurotransmitter release can increase, and tiny protrusions on neurons called dendritic spines can grow larger. Together, these changes contribute to what neuroscientists call an engram, the physical trace associated with a memory.

There is an obvious problem, however. The brain never stops changing.

Synapses appear and disappear all the time, and dendritic spines sometimes turn over. Even the patterns of neural activity associated with familiar places can gradually drift. And yet some memories persist for decades.

Kazumasa Tanaka of the Okinawa Institute of Science and Technology and his colleagues wanted to know how. What remains stable when so much of the underlying biological machinery changes?

×

Get smarter every day...

Stay ahead with ZME Science and subscribe.

Daily Newsletter
The science you need to know, every weekday.

Weekly Newsletter
A week in science, all in one place. Sends every Sunday.
No spam, ever. Unsubscribe anytime. Review our Privacy Policy.

Thank you! One more thing...

Please check your inbox and confirm your subscription.

For an unusually dramatic test, they turned to hibernation.

How do you test a mouse’s memory?

Natural hibernators can wake after months of inactivity and still remember useful information, including locations and social partners. Mice don’t naturally hibernate, but researchers can induce a state known as Q-neuron-induced hypometabolism, or QIH, in which metabolism and body temperature drop dramatically.

Before inducing this state, the team gave mice two kinds of memory tests.

In one, the animals learned to associate a chamber carrying an alcohol odor with mild electric shocks to their feet. When returned the following day, they froze — a standard sign in mouse experiments that they remembered the frightening context. In another experiment, mice learned a spatial task in a plus-shaped maze.

The researchers then induced an artificial hibernation-like state lasting about 48 hours.

During it, hippocampal neuronal firing dropped by roughly 70%, while more than half of the synapses in the region disappeared. Large dendritic spines were no safer than small ones. If strong and stable synapses were individually indispensable for preserving a memory, the mice should have forgotten a great deal.

But they didn’t.

Memories are hard to break

Five days after emerging from artificial hibernation, the mice continued to freeze when returned to the fear-associated chamber. They also performed normally on the spatial memory task. Their brains also showed something similar.

Hippocampal neurons called place cells normally become active when an animal occupies particular locations. After the mice recovered, those spatial representations were still present, and researchers could decode the animals’ location from population brain activity about as accurately as before hibernation.

“It was astonishing,” said Yu-Ju Lin, the study’s first author, in a statement. “Logically, if all our engram synapses were essential to memory retention as traditionally thought, memory should have massively deteriorated.”

Instead, memory survived massive remodeling.

Steve Ramirez, a neuroscientist at Boston University who was not involved in the research, told New Scientist that the result suggests “memories may be harder to break than we thought.”

So what survived when thousands of connections did not?

The neuron remembers

The researchers used several imaging techniques, including correlative light and electron microscopy, to examine synapses linking neurons associated with the learned memories.

They discovered that synapse loss wasn’t entirely random.

Engram-related synapses sitting close together in clusters along dendrites were preferentially preserved during artificial hibernation. More isolated engram synapses were much more likely to disappear.

After the mice woke, many of those missing connections grew back. In dendrites tracked over several days, roughly 82% of the returning spines reappeared at the same locations they had occupied before hibernation.

That is an intriguing clue because it suggests that although individual connections can vanish, some underlying organization may help the brain reconstruct them.

Architecture may matter more than size

The “classic” idea is that result memories persist when synapses becoming larger and stronger. This new study suggests there’s more at play.

“Previously, synaptic strengthening was thought to be key to memory recall, and that stronger synapses with larger dendritic spines were fundamental to long-term memory retention,” Tanaka said. “Here, we show that not every synapse matters and demonstrate instead the vital importance of engram architecture.”

But what does this mean for people?

RelatedPosts

Infants as young as six months old can sense mother’s angry tone
Young people have HD memories
Overeager immune system cells may be to blame for multiple sclerosis
Stunning Neurons on Canvas Painted by a Neuroscientist

The work was done in mice, and artificial hibernation is very different from human sleep, amnesia, brain injury, or neurodegenerative diseases such as Alzheimer’s. This certainly doesn’t show that memories apparently lost in Alzheimer’s disease remain intact somewhere in the brain or could simply be recovered.

Still, the experiment reveals something unexpected about how robust a memory trace can be.

Even when more than half of the hippocampal synapses disappeared and neural activity collapsed, the memories survived.

“It gives hope that even in instances where information seems to be lost in the brain, whether it’s amnesia or Alzheimer’s disease, memory may nonetheless persist,” Ramirez noted.

The next task is to learn how these clusters survive—and whether medicine can one day protect them.

The study was published in the journal Science.