Evidence That Memories Are Consolidated During Sleep

The researchers said their findings lay to rest previous doubts that sleep enables consolidation of newly acquired memories, and also establishes roles for both slow-wave sleep and rapid eye movement (REM) sleep in memory consolidation. Slow-wave sleep is a deep dreamless sleep, and REM sleep is associated with dreaming.

The researchers published their findings on Jan. 19, 2004, in the online Public Library of Science (http://www.plos.org). Senior author on the paper was Miguel Nicolelis, Ph.D., a professor of neurobiology and of biomedical engineering, who is also co-director of the Duke Center for Neuroengineering. Lead author was Sidarta Ribeiro, Ph.D., in Nicolelis's laboratory. Other co authors were neurobiologists Damien Gervasoni, Ph.D., Ernesto Soares, Yi Zhou, Shih-Chieh Lin, M.D., and Janaina Pantoja; and Michael Lavine, Ph.D., of the Duke Institute of Statistics and Decision Sciences. Their work was supported by the National Institutes of Health and the Pew Latin American Program.

In their study, the researchers placed about 100 infinitesimal recording electrodes in the brains of rats, in four regions involved in memory formation and sensory processing. Those brain areas included the hippocampus, which is widely believed to be involved in memory storage, and areas of the forebrain involved in rodent-specific behaviors. The scientists employed the same neural recording technology that Nicolelis and his colleagues used to enable monkeys to control a robot arm, an achievement announced in October 2003.

The researchers next exposed the rats to four kinds of novel objects in the dark, since largely nocturnal rodents depend on the sense of touch via their whiskers to investigate their environment. The four objects were a golf ball mounted on a spring, a fingernail brush, a stick of wood with pins attached and a tube that dispensed cereal treats.

The researchers recorded and analyzed brain signals from the rats before, during and after their exploration, for several days across natural sleep-wake cycles. Analyses of those signals revealed "reverberations" of distinctive brain wave patterns across all the areas being monitored for up to 48 hours after the novel experience.

According to Ribeiro, "We found that the activity of the brain when the animal is in a familiar environment does not 'stick' -- that is, the brain keeps moving from one state to another. In contrast, when the animal is exploring a novel environment, that novelty imposes a certain pattern of activity, which lingers in all the areas we studied. Also, we found that this pattern was much more prevalent in slow-wave sleep than in REM sleep."

Conversely, previous studies by Ribeiro and his colleagues demonstrated that the activation of genes able to effect memory consolidation occurs during REM sleep, not slow-wave sleep.

"Based on all these results, we're proposing that the two stages play separate and complementary roles in memory consolidation," he said. "Periods of slow-wave sleep are very long and produce a recall and probably amplification of memory traces. Ensuing episodes of REM sleep, which are very short, trigger the expression of genes to store what was processed during slow-wave sleep." In principle, this model explains studies such as those by Robert Stickgold and his colleagues at Harvard University, showing that both slow-wave and REM sleep have beneficial effects on memory consolidation, he said. According to Nicolelis, the new experiments remedy shortcomings of previous studies.

"I think that this is another demonstration of the power of the capability of looking at multiple areas of the brain simultaneously," he said. "Previously, investigators have reported the possibility that memories are consolidated during sleep by looking at reverberations, but they only looked in the hippocampus and cerebral cortex. And they only looked for an hour or so. They never looked at several regions of the brain simultaneously, and they never looked for longer periods of time. We've now demonstrated that these reverberations occur in a much more distributed manner over the forebrain, and for a very long time period. Importantly, emphasized Nicolelis, the latest findings provide further evidence that the brain behaves as an integrated whole in processing information.

"The brain cannot be seen as just a mosaic of structures, with one performing a particular function and others doing other unrelated functions," he said. "This model has to be discarded, and this paper is one of the first studies to show that the brain has to be considered as a whole. So, while different aspects of memory consolidation may be happening in different structures, the whole brain is participating in this process, and not just the hippocampus or cortex, which was the idea prior to this work," said Nicolelis.

Next, said Nicolelis, the researchers will perform experiments in which they record from more brain structures over longer time periods. They will also genetically manipulate the animals, switching off specific genes to attempt to affect neural circuitry involved in memory storage.