Sleep is much more than just rest for the body; It is an essential process for the brain, where our memories are organized, consolidated and protected. Although we know that sleep plays a key role in memory, new discoveries have revealed a fascinating detail: the size of our pupils during sleep could be the key to understanding how we process and preserve new and old memories. This finding opens a window into the mysteries of the brain while we sleep, and promises to transform our understanding of how we learn and remember.
A recent study published in Nature carried out on mice by researchers at Cornell University, led by professors Azahara Oliva and Antonio Fernández-Ruiz, have discovered that the contraction and dilation of the pupil during a subphase of non-REM sleep is correlated with the consolidation of new and old memories, respectively.
During the experiment, the mice were fitted with brain electrodes and tiny cameras to track the dynamics of their pupils while they slept. The results showed that when the pupil constricts during non-REM sleep, the brain is consolidating recent memories. On the contrary, when the pupil dilates, the brain reviews older memories.
This finding suggests that the brain possesses a previously unknown sleep microstructure that allows it to separate these two substages of sleep, the consolidation of new and old memories, thus avoiding what is known as “catastrophic forgetting,” where the consolidation of a memory I could delete another one.
The research also highlights that the temporal structure of sleep in mice is more complex and similar to the stages of human sleep than previously thought. These findings could lead to a refinement of memory-enhancing techniques in humans and could help computer scientists train artificial neural networks to be more efficient.
Secrets hidden in the pupils of mice
Although the connection between sleep and memory is widely known, the neural mechanisms supporting this interaction are still not completely clear. Studies attempting to unravel this process typically follow two main approaches: improving memory retention in humans during sleep or conducting mechanistic cellular investigations in sleeping rodents. However, integrating these two areas of study has been a challenge due to the marked differences between both species. Traditionally, it has been assumed that rodent sleep is simple and consists of two main phases: REM sleep, associated with dreaming, and slow-wave non-REM sleep. In contrast, human slow-wave sleep is much more complex, made up of four stages, the deepest being that in which the brain is believed to select and organize memories.
In their joint laboratory, Oliva and Fernández-Ruiz combined their expertise in animal behavior, technological development, and computational analysis to explore what happens in a sleeping mouse’s pupils and neurons during sleep.
Over the course of a month, a group of mice were taught to perform various tasks, such as collecting water or cookies as a reward in a maze. They were then fitted with brain electrodes and small spy cameras that dangled in front of their eyes to track the dynamics of their pupils. One day, the mice learned a new task, and when they fell asleep, the electrodes captured their neural activity and the cameras recorded the changes in their pupils.
“Non-REM sleep is when memory consolidation occurs, and these moments are very, very short periods of time, undetectable to humans, like 100 milliseconds,” Oliva said. “How does the brain distribute these very quick, very short memory scans throughout the entire night? And how does it separate out the new knowledge that comes in, so that it doesn’t interfere with the old knowledge that we already have in our brains?” minds?”
The recordings showed that the temporal structure of sleeping mice is more varied and more similar to the phases of sleep in humans than previously believed. By interrupting the mice’s sleep at different times and then checking how well they remembered the learned tasks, the researchers were able to analyze the processes.
When a mouse enters a subphase of non-REM sleep, its pupil shrinks, and it is at this time that recently learned tasks (i.e., new memories) are reactivated and consolidated, while prior knowledge is not. On the contrary, older memories are reproduced and integrated when the pupil is dilated.
“It’s like new learning, old knowledge, new learning, old knowledge, and that slowly fluctuates throughout the dream,” Oliva said. “We are proposing that the brain has an intermediate time scale that separates new learning from old knowledge.”
The fact that tracking pupils during sleep is a non-invasive procedure opens up applications for future studies in humans, Oliva said, and could be particularly beneficial for people who have memory deficits associated with mental health problems, opening up new avenues to explore how sleep contributes to cognitive health and how specific interventions during sleep could improve memory retention and organization.
“The brain can remember many things with a relatively small number of neurons, and it is not understood how that happens. How can the brain achieve such a feat of memory and cognitive abilities with so few resources compared to ChatGPT, which consumes hundreds of thousands of times more energy to perform any task?” asks Fernández-Ruiz. “This way of dividing, over time, two key memory functions is what underlies the enormous scope of biological brains to have such amazing memory capabilities with relatively low resources. “This provides a new opportunity to train artificial neural networks to be more efficient, perhaps by being more similar to how real brains work.”
Research continues to determine how these patterns of pupillary and brain activity during sleep can be modulated to enhance memory consolidation and what implications they have for diverse fields, from neuroscience to artificial intelligence.
Source: Cornell University
