In work published in the Proceedings of the National Academy of Sciences, a research team led by a Brown University neuroscientist describes groundbreaking recordings of activity in two brain regions during deep sleep.
The "dialogue" they captured occurred between the hippocampus and the neocortex, areas of the brain where scientists believe memories are made and stored. The findings were startling.

Researchers found that electrical activity in the brain cells of sleeping mice wasn't completely random, the conclusion of past research. Instead, the team found that the slow and regular firing of excitatory cells in the neocortex was echoed a fraction of a second later in the hippocampus. The echo of cortical activity was different in the three parts of the hippocampus. Excitatory cells in the dentate gyrus showed a strong echo, while CA3 region cells showed a weak echo. Cells in the CA1 region responded in reverse, quieting down when the cortex was active.

"It's quite surprising," said Brown's Mayank Mehta, an assistant professor in the Department of Neuroscience, who led the research. "We've known for a century that the hippocampus and the neocortex are anatomically connected. But this is the first time we've seen the effect of this connectivity in the brains of live animals. The dialogue is quite unexpected and complex, suggesting that this 'simple' brain circuit is much more sophisticated than we imagined."

To conduct the experiments, researchers recorded electrical activity in the brains of mice. To mimic the deepest sleep state, when neuroscientists believe memory storage occurs, animals were anesthetized and fitted with electrodes. One electrode measured cellular activity in the neocortex. Others measured activity inside cells in the three regions of the hippocampus.

Bert Sakmann, the Nobel-winning physiologist who co-authored the paper, created this novel recording technique at the Max Planck Institute for Medical Research in Germany. Thomas Hahn, a graduate student in the Department of Cell Physiology at the Max Planck Institute and the lead author of the paper, conducted the experiments.

In November, the same team published related research results in Nature Neuroscience. In those experiments, they gathered the best evidence yet of dialogue between the hippocampus and neocortex during sleep. In that study - and in the new one - they found that the interaction occurred in a surprising way: Instead of the hippocampus uploading information to the neocortex, the opposite is true. The neocortex seems to drive the dialogue.

"Scientists have long believed that the hippocampus acts like a scratch pad - it's the place where new memories are quickly 'written' during the day. During sleep, the theory goes, these memories get copied to the neocortex, the official ledger where they are stored," Mehta said. "This theory would suggest that the hippocampus should control communication with the neocortex during sleep. Instead, our findings show that the neocortex controls the communication. An interesting next question is this: What are the implications of this reversed dialogue?"

Mehta's work, and the work of others, is challenging traditional ideas of how the brain processes memories during sleep. In a January Nature Neuroscience commentary, Mehta discussed new research showing that everyday experiences are replayed during sleep in both the hippocampus and the cortex. He then suggested provocative theories to explain these findings: Memories may not be stored during sleep at all, but while humans are awake. The role of sleep, he suggested, may be to erase memories in the hippocampus as a way of creating a fresh page for the brain's scratch pad.

"All of this research raises more questions than it answers," Mehta said. "But we do know this: How we make and store memories is a more complex process than we thought."

The Rhode Island Foundation, the Salomon Foundation, NARSAD: The Mental Health Research Association, The National Institutes of Health and the National Science Foundation supported the work.