Cerebrum Article

Sleep’s Dark and Silent Gate

Published: April 15, 2022
Author: Kayt Sukel
sleeping figure surrounded by various bubbles to represent dreaming and memory

Illustration by Amy Lauren

Each night, as you transition into deep sleep from wakefulness, your body undergoes a remarkable transformation. Your muscles relax. Your breathing slows. Your temperature and blood pressure drop. Even your brain activity changes, decelerating into slow, coordinated waves. Despite these remarkable physiological changes, scientists are now learning that the brain is far from idle during sleep. Rather, it remains hard at work, facilitating memory and learning while uncoupled from the external world.

“For a long time, we believed that being awake all day depleted you and that sleep was what was required to restore and reinvigorate the whole body, including the brain,” says Robert Stickgold, a pioneering sleep researcher at Harvard Medical School. “It turns out that rest has very little to do with the function of sleep—rather, our brain is sorting and consolidating the information we learned during the day so we can better access it when it’s needed.”

Anyone who has ever pulled an all-nighter knows the effect that sleep deprivation can have on cognitive function, including one’s ability to learn and retain new information. Yet, over the last few decades, neuroscientists across the globe have learned that sleep plays an integral role in memory—and it is a role that is highly conserved across the animal kingdom. To better understand how sleep helps us remember, these researchers have been working to characterize not only the physiological changes observed during sleep, but also the neural mechanisms underlying them.

Sleep Stages and Memory Consolidation

Nearly every animal on earth, from fruit flies to non-human primates, experiences some form of sleep, a naturally recurring state of altered consciousness and inhibited sensory activity. And while the exact amount of time spent in slumber, and the patterns of neural activity, differ from animal to animal, humans are no different. We need sleep to thrive.

While it is too easy to associate sleep with some sort of biological “off button,” it is actually a dynamic process that, in human beings, repeatedly cycles through four distinct stages. You have likely heard of the two main types of sleep: rapid eye movement (REM) sleep, characterized by quick back-and-forth eye movements; and non-rapid eye movement (NREM) sleep, a deeper stage of sleep which contains three unique stages, each with its own pattern of brain activity. Neuroscientists are particularly interested in one of those NREM stages, deep or slow-wave sleep (SWS), where the neurons fire in a slow, high-amplitude wave of activity called a delta wave.

Jan Born, a sleep researcher at Germany’s University of Tübingen, was one of the first scientists to show that the spatio-temporal pattern of neural firing seen in SWS is critical to consolidating memories for long-term storage. The specific pattern of activity is a crucial aspect of helping us to remember.

“During NREM sleep, the brain reactivates and reorganizes memories that were encoded when you were awake. And in many cases—not all—when you test recall, it is better after sleep than if you test after wakefulness,” he explains. “My theory is that, during wakefulness, you experience an event in a context. [After sleeping], you have a more generalized, episodic representation of that event. We hypothesize the brain abstracts certain aspects of that memory and puts it into the long-term store. These changes in neural activity help with that.”

Dozens of studies have now demonstrated that most animal species show better recall on a variety of memory tasks after sleep. In humans, this is the case whether it’s a task involving learning nonsense syllables (numerous letter combinations without meaning, used in learning experiments) or a new motor skill. And researchers in different labs are tirelessly working to better understand what neural processes in sleep facilitate this kind of learning.

Finding Balance

Of course, our brains are not primed to consolidate and store everything we have learned or experienced during waking hours. The energetic cost of doing so would likely be too much to bear. To help explain how the brain can find this crucial energy balance, where the brain is not always in a state of learning overdrive, Giulio Tononi and Chiara Cirelli, researchers at the University of Wisconsin-Madison’s Center for Sleep and Consciousness, proposed the synaptic homeostasis hypothesis of sleep nearly 20 years ago.

“The brain mainly learns from strengthening of synapses,” said Cirelli. “That means there’s a need for synaptic re-normalization because you can’t just keep strengthening more and more synapses. Synaptic activity is just too expensive in terms of the brain’s energy budget. So, to keep things in check, the brain weakens some of those connections when we sleep. We couldn’t learn and remember things if it didn’t.”

She and Tononi have shown, using a variety of different molecular and genetic approaches, that synaptic strength, on the whole, is depressed during sleep. This synaptic depression helps to remove unnecessary information—remove the noise, so to speak—as well as prepare the brain for learning the next day. Yet, their work has shown some synapses appear to be protected from this weakening process. Were those preserved synapses strengthened by sleep as part of the consolidation process? In a recent Nature Communications paper, the researchers used two-photon imaging to track synapses in the mouse cortex in sleep after a motor learning task to find out.

“We found that sleep doesn’t provide any special treatment to the synapses involved with the learned information. They stay at the same level of strength,” she says. “But it does bring down the strength of all the other synapses around them. And we see this balancing of synaptic strength is directly correlated with how well animals performed on the later memory task.”

Sleep Facilitates Many Types of Memory

Many of the earliest studies of sleep-dependent memory have focused on motor learning. But research over the past two decades has shown that all manners of memory are influenced by sleep—even the immune system’s ability to recognize a pathogen it encountered during waking hours. But beyond helping to consolidate and store the gist or essence of an event, sleep is instrumental in encoding emotions associated with it.

A recent study by Sara Aton and colleagues at the University of Michigan, published in Nature Communications, showed that sleep is vital to encode fear connected to a visual stimulus in a mouse model. The researchers paired a visual image with a mild foot shock to help the animals learn fear of the stimulus. After learning, the researchers observed, neurons in visual cortex associated with the event were more active during sleep, helping the animals retain the association between the stimulus and emotion. But when the team disassociated the sensory stimulus from the emotion, by selectively activating the neurons in visual cortex without showing the animals the image, they found those animals that slept still held on to their fear.

“What was amazing to me was that when we disrupted the visual cortex so the animal could no longer discriminate between the different stimuli, if the animals are sleeping, they retain that fear. They remembered that they should be afraid, but instead of it being attached to a specific stimulus, it’s just this generalized fear,” Aton says. “But if you do the same thing and sleep-deprive the animals, they don’t have any fear at all.”

These results, Aton maintains, demonstrate that sleep is also vital to remembering the emotional aspects of an experience. The hippocampus must coordinate with both the neocortex and the limbic system to consolidate memories into long-term storage. This more expansive coordination, she says, makes a lot of sense.

“We need both the memory and the emotions, good or bad, to function properly. [Consolidating both the emotional and sensory aspects of an experience] gives us a much fuller context of what we’ve learned—whether it’s a person or place or some other thing—and why it should matter to us so we know how to deal with it when we encounter it again.”

Moving Forward

While researchers have made great strides in understanding sleep’s role in memory consolidation, they largely agree that many questions still need to be answered.

“What does good sleep really mean? And if we figure that out and can learn how to improve sleep, can we improve memory? We know that poor sleep is associated with a lot of psychological and psychiatric problems—could we find a way to go in there and change what kind of memories people are reactivating and consolidating to help with those conditions?” asks Ken Paller, a sleep researcher at Northwestern University. “There are a lot of potential applications if we can figure this stuff out.”

But while the idea of such applications is appealing, Stickgold says scientists are still tirelessly working on more basic mechanistic questions about the relationship between sleep and memory.

“One of the biggest problems with understanding what sleep does for memory is that we don’t really know what a memory looks like. We don’t know how a memory gets encoded. We certainly don’t know how a memory gets found,” he says. “We can talk about the hippocampus and synapses and cell activity all day but if you ask me, ‘Bob, what’s your mother’s name?’ I still can’t tell you exactly what my brain is doing to get you the answer.”

But Stickgold is optimistic that neuroscientists will get there—eventually. Aton concurs.

“We’re getting new tools all the time where we can start looking at different questions, whether it’s looking at how these different stages of sleep are involved with memory consolidation, to understanding what might be happening when we dream and whether that’s involved with memory, too,” she says. “As we get these tools, put together the findings that are out there, and start to test new ideas, I think the coming decades will have quite an impact on our understanding of sleep and why it’s so important to memory as well as to our survival.”


Sidenote: The Dream Factor

Sigmund Freud, the 19th century Austrian neurologist who fathered the field of psychoanalysis, once said that the ability to interpret dreams is “the royal road to a knowledge of the unconscious activities of the mind.” Anyone who has read his seminal book The Interpretation of Dreams knows that Freud thought dreams essentially manifested repressed or unconscious desires. Thinking has changed in the century since Freud’s classic, however, and today, research connecting sleep with memory has inspired new theories regarding the function of dreams.

To wit, Robert Stickgold of Harvard University, and Antonio Zadra, of the Center for Advanced Research in Sleep Medicine at the University of Montreal, proposed in their book, When Brains Dream: Understanding the Science and Mystery of Our Dreaming Minds, that dreams play a critical role in memory consolidation. Their theory, Network Exploration to Understand Possibilities (NEXTUP), suggests that when we dream, our brains are looking to find connections between what we learned during the day and what is already present in our stored memory.

“We’ve known for a long time that if you wake someone from REM sleep after a learning task, they are much faster at identifying links between weakly-associated word pairs,” Stickgold says. “So, if you presented ‘wrong and right’ and ‘wrong and thief,’ they will be better at remembering ‘wrong and thief.’ We believe, based on the studies, that when you are in REM sleep, you are actually in a hyper-associative state where the brain is taking these looser associations, building them into a narrative structure of some sort, and letting those narratives play out to see if you have any emotional response to them. If you do have an emotional response, then the brain knows that association is potentially valuable.”

And what about that time just before you fall asleep, when your brain conjures up that item on your to-do list that must get done the next day—or, the perfect comeback to a comment made by your archnemesis earlier in the day? Stickgold contends that such activity during the period between consciousness and sleep is the brain sorting through the day’s learning to identify important items for consolidation.

“These aren’t dreams—they really are just thoughts and snapshots,” he says. “We believe what’s happening is that these thoughts are being tagged with some sort of label so, once you are in deeper sleep, the brain can ensure that important information is put into memory.”

Stickgold concedes that it has been hard to find the tools to test the NEXTUP theory at the molecular or cellular level. Yet, Aton says that while the model is theoretical, current research looking at sleep and memory consolidation supports these ideas.

“We still need to do a lot of digging to say, yes, this is what sleep looks like and how it’s different from wakefulness,” she says. “We also still need to do quite a bit of work to see how non-REM sleep is different from REM sleep and what the two offer to our overall cognition. But it is entirely possible that REM sleep is there to prepare your brain for consolidation. Maybe dreaming is a way for us to adapt, to come up with new answers to problems, and to sort things out in a way that just isn’t possible when you are awake. It’s going to be exciting to do the experimental work to test these new ideas in the future.”

magazine cover, with image of two children facing each other across the line of a brain's corpus callosum

This article first appeared in the Spring 2022 issue of our Cerebrum magazine. Click the cover for the full e-magazine.