Dueling brain waves during sleep may decide whether rats remember or forget

A sleeping rat may look peaceful. But inside its furry, still head, a war is raging.

Two types of brain waves battle over whether the rat will remember new information, or forget it, researchers report October 3 in Cell. Details of this previously hidden clash may ultimately help explain how some memories get etched into the sleeping brain, while others are scrubbed clean.

By distinguishing between these dueling brain waves, the new study helps reconcile some seemingly contradictory ideas, including how memories can be strengthened (SN: 6/5/14) and weakened during the same stage of sleep (SN: 6/23/11). “It will help unite the field of sleep and learning, because everyone gets to be right,” says neuroscientist Gina Poe of the University of California, Los Angeles, who wasn’t involved in the study.

Researchers led by neuroscientist and neurologist Karunesh Ganguly of the University of California, San Francisco, have been teaching rats to control a mechanical water spout with nothing but their neural activity. The team soon realized that the rats’ success with these brain-computer interfaces depended heavily on something that came after the training: sleep.

To study how the new learning was strengthened during snoozing, Ganguly and his team monitored the brains of sleeping rats after they practiced moving the spout. The scientists focused on brain waves that wash over the motor cortex, the part of the brain that was controlling the external water spout, during non-REM sleep. That stage of sleep usually makes up more than half of an adult human’s night.

Two kinds of brain waves caught the team’s interest. The first, called slow oscillations, were already suspected of helping strengthen memories. And that’s what the researchers found. When laser light and genetic tricks stopped these slow oscillations just milliseconds after they began during sleep, the rats took longer to move their spout using their brains after they woke up. Without these slow oscillations washing across a snoozing rat’s motor cortex, the information didn’t set as well.

The second type, called delta waves, was three to four times as prevalent as slow oscillations, but these waves’ role was a mystery. “We thought that delta waves must be doing something important, because they’re so prevalent,” Ganguly says.

Stopping delta waves had the opposite effect as halting slow oscillations, the researchers found. Knocking out delta waves led the rats to perform better at their task after they awoke, a result that suggests delta waves speed forgetting. The distinction between slow oscillations and delta waves had been squishy, Poe says, until now. “This paper makes it entirely clear that not only are they different, they have diametrically opposed functions,” she says.

Occasional slow oscillations, along with a burst of activity called a sleep spindle, may be able to hold off the memory-scrubbing effects of the more prevalent delta waves, Ganguly says. “Slow oscillations are really important for protecting new learning,” he says. 

Still unknown is how the brain decides which memories to keep, and which to chuck. The presence of a reward — either external, like a lick of water for a rat, or internal, like the good feeling a person gets from a friendly conversation — might be key, Ganguly suspects.

The results might ultimately also have implications for treating people learning how to move again after strokes. Delta waves are more common than usual in these people, Ganguly says.