A stage of sleep – reflected in the size of our pupils – is important to committing recent memories to the brain, which could be manipulated to improve cognitive function and even identify issues with being able to recall newer experiences when awake.
Scientists at Cornell University identified this contracted-pupil sleep state and its connection to recent memories in a complex study that opens up new avenues in understanding cognitive function – and how we could one day optimize this state for better recall of newer information.
Essentially, sleep is divided into two stages – non-rapid eye movement (non-REM, which the researchers refer to as NREM) sleep and REM sleep – and it's widely understood both play distinct roles in memory and cognitive function. But there are also subsets of non-REM sleep (N1, N2 and N3) as we move from light to deep slumber. It's within the NREM stage that the scientists made their discovery.
“Non-REM sleep is when the actual memory consolidation happens, and these moments are very, very short periods of time undetectable by humans, like 100 milliseconds,” said assistant professor Azahara Olivia, one of the study's lead researchers. “How does the brain distribute these screenings of memory that are very fast and very short throughout the overall night? And how does that separate the new knowledge coming in, in a way that it doesn’t interfere with old knowledge that we already have in our minds?”
In the month-long study, mice were repeatedly tasked with finding rewards in two kinds of mazes – a T-maze for looking at immediate short-term spacial memory and decision-making and a cheeseboard task to examine longer-term memory consolidation. These distinct tests enabled the scientists to observe specific hippocampal activity during learning and then memory replay and consolidation in sleep.
Because of the way our pupils change in response to brain activity, signaling states such as arousal, there's been some evidence that, when asleep, the eyes may offer similar clues reflecting neural processes. Which, of course, is not an easy thing to observe and measure.
To do so, the team used custom head-mounted sets that contained a small camera and high-tech brain sensors. The infrared camera recorded pupil size while the mouse moved through the mazes or slept, with the data analyzed in real time to track changes. To measure brain activity, tiny silicon probes were implanted into the hippocampus that monitored the electrical signals from neurons as well as known patterns such as sharp-wave ripples (SWRs) – an important activity in memory replay. The two measures were synchronized to give a complete picture of how pupil changes during sleep reflected different brain functions for recent and long-term memory consolidation.
The researchers also used optogenetics – light therapy to control the activity of specific neurons to manipulate and disrupt the sleeping brain – in order to see if and how it impacted the animals' memory when awake.
Through this, the team was able to see that new memories were replayed and cemented in the brain during a fleeting moment in one of the non-REM sleep stage subsets, which corresponded with the animals' pupils appearing contracted. Then, when the pupils were dilated, it signaled another substage of non-REM sleep, which was linked to the same process but for older, established memories and learnings. Essentially, the two channels of activity prevented the brain from saving new 'files' over old ones – also known as "catastrophic interference" – while maintaining the long-term memories.
“It’s like new learning, old knowledge, new learning, old knowledge, and that is fluctuating slowly throughout the sleep,” Oliva said. “We are proposing that the brain has this intermediate timescale that separates the new learning from the old knowledge.”
How does this relate to the human brain and sleep? Well, more closely than previously thought, according to the study. The real-time recordings garnered by the researchers showed that sleeping mice have brain activity and intricate sleep stages much like us. When the mice had their sleep disturbed to prevent these different memory stages, maze tests that followed revealed a significant difference in the animals' ability to access learnings and complete tasks that either required short-term or long-term memory recall.
"NREM sleep has a stereotypic microstructure that regulates memory replay," the researchers wrote. "Pupil size oscillated between contracted and dilated substates in a minute timescale, signaling an alternation between different hippocampal network states. Although SWRs occurred throughout the whole NREM sleep, replay of recent experiences occurred in SWRs during small-pupil substates compared with large-pupil substates."
Disrupting SWRs in dilated-pupil substates had no real effect on memory performance for recent learning (short-term memory), showing that this neuronal activity plays a key role in the substage of sleep when the pupils are contracted.
"In summary, our study reveals the existence of distinct substates and mechanisms within NREM sleep that segregate the replay of recent and previous memories," the researchers noted. "This finding provides a potential solution for the long-standing problem in both biological and artificial neural networks of preventing catastrophic interference while also enabling memory integration. Given that pupillometry is widely used as a non-invasive technique to study human cognition, our results could help to refine non-invasive intervention experiments for improving human memory, such as targeted memory reactivation, which often lack the temporal resolution of specific brain processes and has variable effects on memory."
The fascinating study provides a deeper understanding of how sleep facilitates learning and memory, and could potentially pave the way for targeted interventions, such as enhancing specific sleep stages to optimize memory consolidation for newly learned information or to aid identified memory impairments.
The study was published in the journal Nature.
Source: Cornell University via EurekAlert!
