Hans Berger could do nothing as the huge field gun rolled toward him.
In 1892, the 19-year-old German had enlisted for military service. One spring morning, while pulling heavy artillery for a training session, Berger’s horse suddenly threw him to the ground. He watched, helpless and terrified, as the rolling artillery came toward him, only to stop at the very last minute.
At precisely the same moment, Berger’s sister — far away in his hometown of Coburg — was struck by a premonition, an overwhelming sense that something tragic had befallen her brother. She begged her father to send him a telegram to make sure he was OK. Berger was stunned by the coincidence. “It was a case of spontaneous telepathy,” he later wrote of the incident.
Determined to make sense of the event and what he called “psychic energy,” Berger began to study the brain and the electrical signals it gave off during wakefulness. In a sense, he succeeded. His efforts to record the small electrical signals that escape from the brain and ripple across the scalp have given us one of the key tools for studying sleep, the electroencephalogram (EEG), or, as Berger described it, “a kind of brain mirror.”
In 1929, Berger published his discovery. As others looked to replicate Berger’s work, they realized the EEG revealed electrical activity during sleep, too. Based on the EEG signature, researchers could show there were several different stages of sleep, and the sequence and timing of them underpins the diagnosis of many sleep disorders. But in the first few decades of using the EEG, there was one stage of sleep nobody noticed.
In The Sleep Chamber
During a long train journey in the 1940s, Robert Lawson, a physicist at the University of Sheffield in the U.K., made an interesting observation. He was sitting in a carriage with a young man and his wife, and as the train rattled along, both his fellow travelers fell asleep several times. Lawson began to collect data, recording the frequency of blinking when his fellow passengers’ eyes were open and when they were closed. “The subjects were quite unaware that they were under observation,” he wrote in a short letter to Nature in 1950. With their eyes open, both the man and the woman blinked roughly once every two seconds. When they closed their eyes, Lawson could see their eyelids twitching at the same frequency for a time. Then, quite suddenly, the blinking stopped altogether, suggesting to Lawson that the transition from waking to sleeping was not gradual but sudden.
Nathaniel Kleitman — then “the most distinguished sleep researcher in the world” — read this casual observation. The University of Chicago professor then gave a graduate student named Eugene Aserinsky the task of finding out more about blinking. He buried himself in the literature, with the aim of becoming “the premier savant in that narrow field.”
As Aserinsky tinkered away with his equipment, he was often joined in the lab by his young son, Armond. “The building was old and dark,” recalls Armond, now a retired clinical psychologist in his 70s, living in Palm Harbor, Florida. “It was like something out of the horror movies of the 1930s.” This might have put off an ordinary 8-year-old, but for Armond, these were exciting times. He lived on campus with his parents and his sister, with the University of Chicago as his playground.
More often than not, Aserinsky would involve young Armond in his research, bouncing ideas off him, asking him to read through a manuscript or using him to calibrate the EEG equipment. “Electricity is coming out of your brain,” he explained to his son, “and this machine is going to measure it. It will be interesting to see what’s produced when you’re asleep.”
Armond remembers one session in particular. It was the afternoon, and he was in the sleep room. It was a chamber furnished only with a cotlike bed, with an intercom as the sole means of communication with the outside world. On his scalp and his eyelids were electrodes with wires that would transmit his brain waves and the movement of his eyeballs to the recording equipment outside. Instructed to lie down and try to sleep, Armond — like a good boy — did as he was told.
An hour or so into his nap, the readout suggested his eyeballs had suddenly gone crazy, jerking rapidly from left to right. “My father woke me and asked me what was going on.” Armond had been dreaming. “There was a chicken walking through a barnyard,” he still recalls 65 years later.
As Aserinsky’s studies progressed, his sleeping subjects appeared to enter a categorically different state. Poring over the reams of paper that had spooled out of the machine — up to half a mile of paper per session — Aserinsky found that there were times when the brain signals during sleeping looked almost indistinguishable from the signals during waking. And though their eyeballs were jerking, the subjects were obviously still asleep.
This became known as rapid eye movement, or REM. Aserinsky and Kleitman wrote up these findings for Science in 1953. They were so focused on the eyes, though, that they failed to notice something significant. REM is also accompanied by a complete loss of muscle tone throughout the body, most likely to prevent you from acting out your dreams.
The discovery of REM inspired a flurry of research. University of Chicago psychology graduate William Dement, who eventually became an iconic Stanford University sleep researcher, had recently worked his way into Kleitman’s circle and helped Aserinsky with his work. He was excited by the possibility that REM might be an objective way to study dreaming.
Within a few years, Dement and Kleitman had come up with an EEG-based description of the stages of a normal, healthy night’s sleep. All of these stages — 1, 2 and 3 — are collectively referred to as “non-REM” sleep. Then, all of a sudden, the brain passes, as if through some cognitive portal, into the REM state. A hurricane of ocular activity kicks in, lasting a matter of minutes before the brain returns to relative non-REM calm and the entire cycle starts again. The duration of this cycle — from the start of stage 1 to the end of REM — typically lasts around 90 minutes and repeats throughout the night.
“I believe the study of sleep became a true scientific field in 1953, when I was finally able to make all-night, continuous recordings of brain and eye activity during sleep,” wrote Dement. “For the first time, it was possible to carry out continuous observations of sleep without disturbing the sleeper.”
Dement eventually had sufficient data to suggest the human brain needs dream time of around 80 minutes a night — and if it doesn’t get it, it attempts to catch up. This implies REM is serving some vital physiological purpose crucial to the proper functioning of the brain.
Austrian neurologist and founder of psychoanalysis Sigmund Freud believed dreams must be a form of wish fulfillment, revealing repressed and often sexual desires. Although few people now buy into such Freudian thinking, a recent study suggests that most still believe there is meaning to be gleaned from dreams.
But this popular pastime is probably a colossal waste of time and energy. In the 1950s, Dement pored over the EEG signals of his subjects in the hope of finding the function of REM and, hence, dreams. At the same time, Michel Jouvet at the University of Lyon made an intriguing discovery suggesting dreams might not be the raison d’être of REM.
When he stripped back the brains of cats, removing the organ’s thick outer layer called the cortex, where most of dreaming action is thought to take place, the animals still slept perfectly well, with a regular cycle of both non-REM and REM. It turned out REM had its origins in an ancient region of the brainstem called the pons.
This observation is at the heart of an article published in the American Journal of Psychiatry in 1977 by Allan Hobson and Robert McCarley, both psychiatrists at Harvard Medical School. They proposed REM begins with some kind of activation in the pons, a content-free pulse that acquires meaning only as it ricochets through the cortex, synthesizing vivid imagery, crazy plotlines and intense emotions. This sequence of events effectively demotes dreaming to something of a secondary, perhaps inconsequential afterthought.
In The Promise of Sleep, Dement captured the essence of this phenomenon with a sparkling analogy:
It might help to think of a stained-glass window. … White light, which is a jumble of colors, enters on one side, but what comes out on the other side has a definite pattern of colors that is often very meaningful. Like the stained-glass window (which is a filter for light), the brain acts as a filter that imposes order on the random signals passing through it.
That’s not to say researchers and clinicians have reached a consensus on why we sleep. They haven’t. However, sleep most likely performs more than just one function.
For example, one idea is that sleep evolved to save energy — a hypothesis that Jerry Siegel, a sleep researcher at the University of California, Los Angeles, refers to as “adaptive inactivity.” Some people have countered by pointing out the brain is active during sleep and the energy saved is minimal — in humans, it’s the equivalent of what you’d get from a piece of bread. “My take,” says Siegel, “is that saving a little energy is not trivial. If you could give half of the people on Earth a piece of bread once a day, they would do a lot better than the half that doesn’t have a piece of bread.”
There are plenty of other ideas, too. It could, for instance, be a way to purge pointless information. In 2003, biologists at the University of Wisconsin-Madison developed this notion. The brain is so busy making connections when an animal is awake, they argued, that sleep is needed to pare back this neurological noise. More than a decade later, there is now compelling evidence that some kind of neuronal editing takes place during the non-REM stages of sleep.
There is evidence, too, that non-REM sleep may be a time for the brain cells to carry out important housekeeping duties, for instance, replenishing stores of neurotransmitters, the brain’s chemical messengers. Brain cells also appear to shrink somewhat during non-REM sleep, allowing more room for cerebrospinal fluid to percolate and wash away toxic metabolic waste.
Despite the best efforts of Dement and others, we haven’t made the same headway with REM.
Emmanuel Mignot is Dement’s successor as director of the Stanford Center for Sleep Sciences and Medicine. In Mignot’s view, REM has all the hallmarks of being an ancient phenomenon, one that evolved amid vertebrate evolution before the forebrain, where most complex thinking happens, had a chance to expand. Perhaps REM was the primitive brain’s way of getting some rest, he suggests.
It also provides an explanation for one of the most baffling things about REM: why it sends most of the body’s core physiological functions offline. The skeletal muscles shut down, the body temperature free-runs, breathing becomes irregular, the heart races, blood pressure rises and the blood vessels dilate. “During REM, you become a little like a reptile,” says Mignot.
That’s why Philippe Mourrain, a developmental geneticist at Stanford, is hoping to find the secrets of sleep in a creature simpler than humans: zebrafish. Fish don’t have eyelids and they don’t move their eyes when they’re asleep, but they do experience a sleep state similar to humans, he says.
Mourrain believes the main feature of REM is the muscle paralysis, rather than the peculiar eye twitching that some species demonstrate. “Eye movement is not the best way to quantify this state,” he says. It would be much better to come up with a definition of sleep rooted in the parts of the brain that actually control these phenomena.
Thankfully, these ancient structures are completely exposed in zebrafish. As a bonus, larvae are transparent, so the entire nervous system — brain and all — is visible under a microscope. With some clever genetic engineering, it’s also possible to smuggle a certain protein into highly specific populations of neurons, so that when they fire, they emit a flash of fluorescent light.
In Mourrain’s lab, a paused video of a microscopic view of a larval zebrafish fills a laptop screen. Certain parts of its brain are fluorescing, some more than others. There is a bright signal of neural activity coming from the hind-brain, a fainter glow in the vicinity of the eyes and a still weaker, more diffuse emission from the tail muscles.
Mourrain hits the play button. In the first few seconds, not much happens. The fish, with its head in a blob of transparent gel to keep it under the microscope’s lens, is awake. But when a drop of a hypnotic drug known to trigger REM is plipped into the water, there is a neurological reaction. Boom! A burning flash of light starts in the fish’s pons. A wave washes from the brainstem forward through the brain, bleaching the eyes and petering out at the tip of the nose; it’s just like the waves that drive REM in mammals.
“Concentrate on what happens to the focus of the image,” he says. As the pons flashes, the zebrafish blurs. “It’s going out of focus because of the muscle relaxation.” The beautifully coordinated wave of light radiating from the pons and accompanied by muscle paralysis is exactly what occurs in mice, cats and humans during REM sleep.
In the basement of the building, he and his team look after about 20,000 fish. He makes his way into one of several windowless labs, this one containing around 1,400 shoebox-sized tanks stacked on racks like books in a library. “They all house different mutants and different transgenic lines,” says Mourrain.
On a work surface in the corner of the room, there’s a small platform tilting in a circular fashion. On top of it are two objects wrapped in tinfoil, each about the size of a small bullet. “It’s a sleep deprivation experiment,” explains Mourrain. His words draw an image of the tiny zebrafish in the darkness of their little plastic tubes, the water around them swirling in constant, sleep-disturbing motion.
Mourrain’s work on zebrafish strongly suggests that non-REM-like and REM-like states are both extremely ancient phenomena that arose more than 500 million years ago and have been conserved throughout the course of evolution. It’s possible the REM-like sleep of fish could even trigger some kind of dreamlike experience in their sliver of a cortex.
It’s understandable, of course, that humans should be interested in human sleep. But trying to run before being able to walk is rarely a successful endeavor. “Studies of non-mammalian vertebrates like fishes, but also amphibians, reptiles and birds, may bring more light than originally expected on mammalian sleep and REM,” says Mourrain.
“If I could have a transparent human being, easy to manipulate and easy to image, then maybe I’d consider it as a model species,” he says.