The scientists implanted electrodes in the brains of rats, surveying the activity of up to 120 neurons (nerve cells) in the medial prefrontal cortex (a forebrain region responsible for goal-oriented executive functions such as organizing thoughts and actions) while the animals completed a navigational task, scampering between spots in sequence on a circular table top. The research team monitored the rats' brain activity daily for a few weeks as they scurried to complete the 50-minute running session and then napped for 20 minutes to an hour.
Using two different methods—comparing the activity between pairs of cells and surveying patterns over the entire population of monitored neurons—the team noted that neuronal activity sequences that occurred when the rats were running seemed to reappear during sleep.
"We looked at them and it just hit us in the face that there [were] striking similarities," says David Euston, an assistant research scientist at the U.A. College of Medicine's Division of Neural Systems, Memory and Aging.
Not only were the same patterns reactivated while snoozing, but the replay would take place six to seven times faster than when the rats performed the task. "During behavior when we're actually interacting with the world, the brain has to go at the same speed at which the body is going," Euston speculates. "During sleep, maybe the brain can go faster when it's not time-locked to behavior."
This replay phenomenon has previously been shown in the hippocampus, a forebrain structure involved in episodic memory, and in the visual cortex, where sensory information related to sight is processed. Euston believes the process may be related to plasticity, the strengthening and weakening of connections between of nerve cells that is thought to underlie learning.
"One way you could strengthen memories is by playing them multiple times," says Euston, referring to the unique sequence of neuronal activity that accompanies new learning events. By replaying them repeatedly, the brain may be strengthening communication between neurons, thereby consolidating the memories.
Mayank Mehta, an assistant professor of neuroscience at Brown University, says the new findings are interesting, but is skeptical about the assumption that they're based on. "Is this consolidation or erasure?" he asks, questioning whether the phantom activity is really the filing of a memory or whether it is akin to clearing a chalkboard so more learning can take place. "From the point of view of behavior, both [processes] can help," he says.
Euston says the team will now try to determine whether this replay of neuronal activity is associated with learning. If so, he says, the replay patterns should be strongest when the animals have just completed learning-related challenges.