For most of us, sleep is a time to rest and a chance to dream. For biochemist James Krueger, sleep offers a window into how the brain is organized, and how its trillions of cells coordinate their actions to create a perceiving, reflecting, inventing human mind.

Krueger is the founding member of a research group that has developed at Washington State University over the past 15 years. With established researchers Krueger, Lynn Churchill, and Gregory Belenky, and up-and-comers Hans Van Dongen and David Rector, the WSU sleep team is one of a few nationwide—including groups at Harvard, Stanford, and the University of Pennsylvania—that focus on fundamental questions: What makes us sleep? Why do we do it? What are the effects of not getting enough sleep? And the most basic of all: What exactly happens in the brain when we sleep?

A Sleep Switch?

Some researchers think there’s a control center somewhere in the brain that emits electrical signals or chemicals that cause the rest of the brain to switch into sleep mode.

This sort of “top-down” model, of sleep being imposed on us, goes back centuries. An ancient Roman poet described sleep as a winged god that “steals o’er the greedy cares of men, and stoops and beckons from the sky, shrouding a toilsome life once more in sweet oblivion.”

It’s a reasonable idea, but James Krueger doesn’t buy it.

Tall and genial, with the easy manner of a natural storyteller, Krueger says anyone trying to identify a single sleep control center sooner or later hits a logical snag. Say you do find a control center; then you have to ask, what controls that?

If sleep isn’t imposed by some control center, what else could cause it? About 20 years ago sleep scientists began to find hints. A Russian group reported that dolphins only sleep in one half of their brain at a time. Though surprising at first, this made sense. Dolphins need to swim to the surface in order to breathe. If a dolphin ever slept in both halves of its brain at once, it would drown.

Investigators then found that sleep isn’t always an all-or-nothing affair in our brains, either. Working one side of the brain harder, by exercising one side of the body, or one ear or one eye, makes the worked hemisphere undergo more and deeper sleep than the idle side.

In 1993 Krueger and visiting colleague Ferenc Obal, who died in 2004, proposed that sleep starts in individual groups of neurons called neuronal assemblies, or cortical columns. Each group contains between 10,000 and 100,000 neurons, or nerve cells, that are highly interconnected and that work together to perform a specific function. Krueger and Obal suggested that after a group is used a lot, it shifts into a sleep state. When enough groups in an animal’s brain are in sleep state, the animal itself goes to sleep.

“It’s a bottom-up process,” says Krueger, a process that doesn’t require central control. He compares it to the way ant colonies perform complex, coordinated tasks. “Colonies have behavioral properties that are whole-colony properties, and yet there’s no individual ant directing that. “It’s an emergent property of the whole colony.”

Likewise, he and Obal theorized that sleep emerges from the activity of neuronal groups. No control center, no switch, just a sleep-wake shift depending on what the individual groups are doing.

It took Krueger a while to reach that view. He started working on the chemicals involved in sleep in 1974. Another investigator had found that injecting fluid from the nervous system of a sleeping animal into another animal made the recipient sleepy. Krueger set out to purify the substance that was causing the sleepiness.

“We were trying to isolate the sleep-promoting ‘factor S,'” he says. “We thought it was a single compound. In our dreams we had ideas that we’d have the sleep hormone.”

Now he chuckles at that optimism. The current list of chemicals thought to be involved in sleep regulation numbers well over a hundred.

Most belong to a class of small proteins called cytokines, which function in the immune system as well as the brain. Krueger’s lab is investigating several of them, notably one called tumor necrosis factor (TNF). It ‘s unsettling to learn that a chemical named for its ability to kill things permeates my brain every night, but Krueger says the name is simply a holdover from its initial discovery by cancer researchers who found that it cleared up malignant skin lesions on mice overnight. “It’s a miracle drug, if you’re a mouse with skin cancer,” says Krueger. Unfortunately, it doesn’t work that way in humans—but Krueger has shown that it does appear in our blood and brain just prior to and during sleep.

“There’s no other substance anywhere, of any sort, where you can correlate circulating levels with sleepiness,” says Krueger. Not only that, but injecting TNF enhances sleepiness, and Hitoshi Yoshida, a visiting scientist from Japan, has found that injecting TNF into one hemisphere of a rat brain causes that hemisphere to go into a deeper sleep than the other hemisphere.

Still, Krueger has no illusions that TNF is the sleep factor. “I don’t even know if phrasing the question in the sense of one being most important, or more important than another, is even a correct way to look at it,” he says.

“For whatever reason, people like to have a single chemical responsible for things. Drug business drives that in part, but I think it’s also somehow inherent in our minds, that we want a single gene, a single enzyme, a single neurotransmitter responsible for disease A, B, C, or D. But it ain’t so.

“To think that we understand how this molecular network works is naïve. We have no idea how it works.”

No Snooze, You Lose

Shifting focus from molecules to people brings a different kind of answer to the question of why we sleep: we just don’t function well if we don’t.

The exact amount we need changes through our lifetime—babies sleep a lot more, and old folks tend to sleep less—but the standard advice to get eight hours of sleep a night is right for most adults. Some oft-heard advice about naps, though, isn’t right—for instance, that a short daytime nap will make up for a short night’s sleep.

“That’s baloney,” says Greg Belenky. “A five-minute snooze in a meeting takes the edge off the sleepiness, but it doesn’t improve your performance.” If you only slept six hours in the past 24, you’ll need two hours of nap time to top yourself off. So if you’re going to rely on naps to stay rested, he says, do it right. “Nap early, nap often.”

He should know. Belenky studied the effects of sleep deprivation for the Army for more than two decades. He came to WSU Spokane in 2004 from the Walter Reed Army Institute of Research, where he directed the Division of Neuroscience.

Other than using military style for clock times (“See you at ten-hundred hours”), Belenky seems more like a cheerful academic than a career Army officer. He laces his conversation with wordplay and movie references. The Treasure of the Sierra Madre and Night Shift come up in the first 15 minutes of our talk.

Belenky says the Army’s interest in sleep science centers on one thing: performance. Will a soldier in the field make the right decision and be able to follow through on it when called on to do so?

His research showed that you don’t need to lose an entire night’s sleep to be impaired. Even modest sleep deprivation over a period of days—dropping from eight hours to seven hours of sleep per night—significantly hinders mental performance, reaction time, and judgment.

Belenky studies sleep and performance both in the field and in the lab. In field studies, volunteers wear a wristwatch-like device called an actigraph that records when the wearer is asleep (motionless) and when he’s awake. The actigraph allows the wearer, or a supervisor, to keep track of how much sleep the individual has had. In the lab, Belenky restricts the sleep of volunteers by different amounts for periods of up to tw
o weeks. In both kinds of tests, he measures the subjects’ performance on a psychomotor vigilance task test, or PVT.

He hands me a Palm Pilot to let me try his latest version of the PVT. A bull’s eye image appears on the screen. I push a response button. The screen shows my reaction time—0.27 second. This goes on for about a minute, the target image showing up at odd intervals. My average score at the end is 0.302 second, a bit slower than a well-rested person with PVT experience.

I note that three of my scores were much higher than the others—over 0.4 second—and say it felt like they came after a moment of inattention.

“Exactly!” Belenky says. He notes that Hans Van Dongen has shown that the number of “PVT lapses”—response times of half a second or longer—shoots up after just two nights of six hours’ time in bed.

Half a second doesn’t sound like much, but if you’re driving 75 miles per hour, in half a second you’ll travel 55 feet—far enough to cross two lanes of traffic and the median.

Add to that the finding that up to one-third of us are chronically sleep deprived, and you get an even scarier picture. It may not matter if you got a full quota of sleep this week. Odds are the driver behind you, the one beside you, or the one in the oncoming lane, didn’t.

“For years the Federal Motor Carrier Safety Administration said, ‘Oh, only about 5 percent of accidents are fatigue-related'” says Belenky, “—but another 35 percent are from inattention!”

In fact, sleep shortage played a role, often the primary role, in almost every accidental disaster in recent memory—the meltdown at Chernobyl and the wreck of the Exxon Valdez, for example.

Belenky says people who do well with substantially less than eight hours of sleep per night are rare. And a person’s own assessment of his or her fitness “is next to useless,” he says. “Some people perform poorly and think they’re doing great, while others perform fine and think they’re doing poorly.”

He says most tales of legendary figures who “never” sleep can’t be believed. General Maxwell Thurman, who helped rebuild the Army after the Vietnam era, “was rumored never to sleep,” says Belenky.

“He was unmarried, sort of a military monk. A real fire-eater. His staff all had to be there before he got there, and only could leave after he left, and he had a driver. So basically he could spend every waking minute on Army business and still get seven or eight hours of sack time. To his staff it looked quite different. That’s where the idea came from that the Old Man never slept. But that simply wasn’t true.”

Belenky and Van Dongen are getting ready to start studies of sleep and performance in the new “sleep suite” at the Spokane campus. Equipped with private sleeping rooms and a shared kitchen and living room, the sleep suite allows observers to track the brain activity, sleep cycles, and PVT performance of four live-in volunteers at a time. Also in the works are field studies with medical residents at Spokane-area hospitals and Spokane police officers, among others.

Piece by Piece

Another intriguing bit of information came out of Belenky’s sleep deprivation studies: as a PVT test goes on, performance drops. Well-rested people do a bit worse after three or four minutes of the test. Sleep-deprived people do a lot worse after just one minute. This is the “time-on-task” effect—the longer we do one thing, the more fatigued we get, and the more mistakes we make. Sleep deprivation accentuates it.

Belenky thinks the drop in performance happens because the brain cells responsible for doing the task enter a sleep-like state after they’ve been used a lot. That’s why a person can feel tired after working for a long time on one task, but perk up by switching to a different activity that uses different sets of cells.

That fit nicely with Krueger and Obal’s hypothesis. Still, direct evidence that that’s how sleep happens remained elusive. Electroencephalograms (EEGs) of humans couldn’t give fine enough resolution to see what specific neuronal groups were doing. Studies with animals offered more detail, but few ways of measuring performance. There was no PVT test for lab rats.

Amazingly, says David Rector, there wasn’t even any way of knowing whether neurons were “asleep” or “awake.”

“So far the sleep community has only had one definition of sleep, and that is if you’re lying down, or in some particular posture, and you’re not responding to external stimuli,” he says. “It’s not a very useful definition”—especially if you’re looking at brain cells rather than a whole animal.

In a warren of small rooms crowded with workbenches and electrical equipment, Rector combined EEGs and a standard behavioral test with new brain-mapping techniques to create a new way of asking the question.

“You should see how hard these experiments are,” says Krueger. “He has to train the animals for months.”

First, Rector trained lab rats to be comfortable hanging out in a hammock for a couple of hours, dozing intermittently while having a whisker twitched and brain activity recorded. The hammock was needed to keep the rats from curling up in a ball when they napped, which would have made the whisker work impossible.

Then, using a specially-designed electrode array, Rector mapped the area on the surface of the brain that contained the neurons connected to the whiskers. Once he found the column of cells “belonging” to a particular whisker, he twitched that whisker with a mechanical device. The size of the neuronal response told him whether the column was asleep or awake.

While the whisker was being twitched, EEGs showed when the whole animal was awake and asleep. Most of the time, the column’s state matched that of the whole animal; but sometimes, the rat was awake, but those particular cells had zoned out. The harder the cells had been required to work—the longer the whisker had been twitched—the more often the column entered the sleep state.

Rector had discovered the basic biological unit that sleeps: not the whole brain, and not individual cells, but groups of cells related by function. That was already a huge breakthrough, but Rector went even further. He taught his rats a skill. They learned to lick when a particular whisker was twitched, and to refrain from licking when a different whisker was twitched. This behavior had been used in studies with rats before, but not in conjunction with the sleep-wake assessment. It became, in essence, the world’s first PVT test for rats.

Once a rat was adept at its new skill, Rector twitched one whisker repeatedly, to “tire out” the neurons connected to it. Lo and behold, when the column went to sleep, the rat’s score on the lick test dropped—just like Belenky’s human subjects several minutes into a PVT.

Last fall, Lynn Churchill added another piece of the puzzle. She found that after a prolonged period of whisker twitching, the worked column produced TNF, the chemical Krueger had earlier shown is related to sleepiness.

Rector had put the pieces together. Krueger and Obal were right. Different parts of the brain can be asleep at different times. When enough groups enter the sleep state, the whole animal goes to sleep. There’s no central control, no on-off switch. Sleep emerges from the individual clusters of neurons that work together on some task. As they work, they accumulate TNF and other chemicals, and eventually they tune out.

What sleep does for the neuronal groups remains an open question. Krueger & Co. have ideas about it—but that’s a story for another day.

It’s Better in a Group

Rector presented his whisker-twitch results at the national Sleep Research Society meetings in Denver last summer.

“He changed sleep research forever,” says Krueger. “I think David’s paper on this will probably be the single most important paper ever published in sleep research, because he identified what it is that sleeps.”

After the symposium, all the members of the team, plus Van Dongen, who was not yet at WSU, and another scientist who visits Krueger’s lab yearly, got together for a lunch that became a brainstorming session.

“That was such a treat,” says Krueger. Their discussion brought out ideas that culminated in a grant proposal, currently under review by the National Institutes of Health, to look at chronic fatigue syndrome as a disorder of the sleep-regulation system.

Krueger says that’s just one example of how a group of people working in the same area can make headway where a lone investigator would struggle.

“Having isolated individuals working on their areas by themselves doesn’t work,” he says. “It’s so hard for them to compete nationally and internationally.”

He’s especially pleased that the WSU team spans two generations of researchers.

“Our strength is in David and Hans,” says Krueger. “They’re young, and they are the very best. Just watching these guys interact, already it’s just spectacular. Harvard doesn’t have anybody that age group that’s even close to these guys. The future’s with us.”


To learn more about sleep, sleep deprivation, and your own sleep health, check the website of the National Sleep Foundation,