The Nightmare Is Over: Researchers Discover the Switch for REM Sleep

Berkeley scientists managed to get 94 percent of their test mice to fall into REM sleep almost immediately when they flipped their "switch." Jason Stewart/Eye Em/Getty Images

University of California scientists have found a powerful on-off "switch" that sends dozing mice into a full-fledged, eye-twitching, cheese-chasing mousy dreamworld.

That "switch" seems to be a particular group of neurons in the brain that, when stimulated in sleeping mice, immediately pushes the rodents into REM sleep. REM is that deep state of snoozing characterized by rapid eye movement, muscle paralysis and a super-active, dreaming brain.

"It was already known for many decades, the neurons necessary for REM sleep. But we still had to search within the brain stem," says Franz Weber, one of the main researchers and a postdoctoral fellow at UC Berkeley. "I was explicitly searching for a brain region in the brain stem that [induced] REM sleep."

There are other regions in the brain that relate to REM sleep. In fact, scientists have long known that the region that the Berkeley scientists were probing — the underside of the medulla, at the base of the brain atop the spinal cord — was involved at least casually in one part of REM, the muscle paralysis part.

But what Weber, lead author Yang Dan and their team showed was that the region and the group of neurons they were exploring were not just "into" that part of REM. They were into it all out. The eye movement, the muscle paralysis, the activation of other parts of the brain. All of it.

When they flipped their "switch" — done by an optical device that triggered certain neurons — 94 percent of the test mice fell deeply into REM sleep almost immediately. That indicated this node of neurons was especially important, a veritable vein of gold in the mine of sleep research.

"The capability to induce REM sleep on command," the researchers wrote, "may offer a powerful tool for investigating its functions."

It was just mice, you say? "The medulla, as a whole, is really an evolutionary brain structure. It exists, also, in humans," says Weber. "In that respect, it's likely that the role it plays in a mouse is similar to humans. But, I say, 'likely.'"

The research — outlined in the Oct. 15 edition of Nature — didn't stop with sending mice into REM sleep. The scientists also stimulated these particular medulla neurons (called GABAergic neurons) while the mice were awake.

Normally, when awake, these neurons are mostly active when mice are grooming or eating — pleasurable activities. So when the researchers triggered the neurons when the mice were awake, they got all happy. They ate more.

Just as important, the scientists guess that other neurons that are mostly active in stressful situations were shut down when the targeted neurons fired.

All that wakeful activity brings us to something that might be happening when we sleep.

"It is very likely that these ventral medulla neurons inhibit these stress neurons. So the link I want to make now is, when we are eating, and we are kind of feeling good, it might be because stress neurons are inhibited," Weber says. "Similar, the idea is that having REM sleep also helps you overcome stress or emotionally stressful experiences. So perhaps there is a similar mechanism going on during dreaming."

In addition, scientific evidence exists that shows that people with mood disorders like depression and neurological diseases like Parkinson's often report a disturbance in REM sleep. If the Berkeley researchers are on the right path — that these neurons not only induce REM but also close off stress — that mechanism may not be working properly in people with those types of disorders. And if all of that is correct, it could provide further insight into how to deal with those issues.

Lots of research remains to be done. Lots of questions still need to be answered.

For Weber and the Berkeley team, the work begins deep in the brain.

"There are many ideas," Weber says. "The big hope for the future, as we learn more and more about the neural hardware, is we might also learn more and more about what happens during sleep."