Sight Unseen

Researchers run blind mice up the circadian clock, finding more to photoreception than meets the eye



Click here to download a PDF file illustrating the process of photoreception.




“Brightness information is used in brain systems below the level of consciousness. These systems help synchronize your sleep/wake cycle...”


The eye’s service to the brain appears more complex than the old analogy of a lens-and-film camera.



People who lose this second system of "sight" are at risk for other medical problems, such as sleep disturbance, or even heart attack.


In future studies, Van Gelder hopes to test whether treatment with melatonin will help regulate sleep patterns in children with optic nerve disease.



SCIENTISTS USED TO ASSUME that the eye functioned like an old-style camera. Basically, light came in through the cornea, was focused by the cornea and the lens, and wound up on the retina — which, following this metaphor, is the camera’s “film.” An image was made there as the light was converted into electrical signals by photoreceptor cells — rods and cones — and those signals transmitted to the brain.

“But a few years ago, a group led by Dr. Russell Foster found that mice that lack rods and cones still had sensitivity to light,” says Russell N. Van Gelder, MD, PhD, assistant professor of ophthalmology and visual sciences and of molecular biology and pharmacology at the School of Medicine. “Even visually blind mice had the ability to dilate and constrict their pupils in response to light.”

Russell N. Van Gelder, MD, PhD

But how? How would an animal without the ability to see still open and close its pupils in response to light? That question has driven much of Van Gelder’s research over the past few years. And he’s learned that there’s a whole other pathway in the eye. The eye’s primary job may involve vision, but that’s not all it does.


Despite how well the “eye as camera” metaphor works, it’s probably time for an update. Modern cameras also have light meters, and Van Gelder’s research suggests that the non-visual role of the eye might be compared to a light meter, which cannot form images but can determine how bright the environment is. In a camera, that information helps the photographer determine how to set the shutter speed and whether to use a flash. In the eye, that information is used for much more.

“Brightness information is used in brain systems below the level of consciousness,” Van Gelder says. “These systems help synchronize your sleep/wake cycle, reset your internal body clock to jet lag if you travel across time zones, control the pupil of your eye and how it responds to light, and regulate the release of hormones such as melatonin.”

Part of the second system’s job is to protect vision by controlling the dilation and constriction of the pupil. But it also communicates with parts of the brain that don’t receive much input from the visual system. Whereas the visual system sends messages to the thalamus, the cells that perform the eye’s light meter function hook up with the hypothalamus, an older part of the brain that’s involved with basic physiologic functions like eating and sleeping—and the circadian clock.

The retina’s primary visual system consists of the rods and cones, which convert light signals into nerve impulses processed in the brain. The non-visual system in the retina relies on different kinds of cells called intrinsically photosensitive retinal ganglion cells (ipRG cells). These cells don’t appear to be involved in vision, but they are directly light sensitive and play a crucial role in other functions.
Photoreception in the retina begins with light striking a photo-pigment molecule. Light induces a chemical change in the photopigment, which then is amplified into a signal the photoreceptor cell uses to communicate. Van Gelder is one of several scientists who have worked to identify the photopigments that ipRG cells use.

In a study published last year in the journal Science, his team reported that a family of proteins called cryptochromes is important in the pupil’s response to light in blind mice.

“First, we showed that blind mice lacking cryptochrome lost about 99 percent of their light sensitivity compared to mice that could see and about 90 percent of their light sensitivity compared to blind mice that still could make cryptochrome,” Van Gelder says.

The researchers demonstrated the importance of cryptochrome by exposing blind mice to light. Although the mice could not see, their pupils dilated and constricted in response to light. It took about 10 times more light to make pupils constrict in blind mice with cryptochrome than in mice that could see. In mice without cryptochrome, it took 100 times more light.

In the months following that discovery, Van Gelder and colleagues from the Novartis Gene Research Institute, the Uniformed Services University, and other centers demonstrated that blind mice lacking a second protein called melanopsin were even worse off than those without cryptochrome. They reported in a subsequent issue of Science that visually blind mice without melanopsin lost all pupillary responses and had other problems, too.

“These mice not only are blind, they also are circadianly blind, meaning they can’t synchronize their behavior to the day/night transition,” Van Gelder says. “It appears melanopsin is absolutely required for the regulation of that function.”
The work supports the notion that the eye is responsible for more than just vision, that it regulates circadian rhythms, pupillary responses and hormone secretion, functions that are very important in animals. (In sheep, varying hormonal levels signal the proper breeding time.)

At present, melatonin is the only hormone linked directly to this system; Van Gelder believes others also may interact with the eye’s light meter. The stress hormone cortisol is released by the adrenal glands every morning. Its regulation can be disrupted in mice that carry mutations in so-called clock genes. Van Gelder and collaborator Louis J. Muglia, MD, PhD, associate professor of pediatrics and of molecular biology and pharmacology, are investigating whether mice without melanopsin or cryptochrome experience similar disruptions.

“If you’re blind, you probably think that although you have no vision, everything else should be fine,” Van Gelder says. “But if you lose this second system, you might be at risk for other serious problems.”


One of those problems appears to be sleep disturbance. In the February issue of the journal Ophthalmology, Van Gelder’s group reported that young people with eye diseases that damage the inner part of the retina and optic nerve are significantly more likely to have sleep disorders than those with other types of eye disease or those with normal vision.

“This study was really an opportunity to go ‘bench to bedside’ and take the results from our animal studies to generate a testable clinical hypothesis,” he says.

His research team studied 25 students, ages 12 to 20, from the Missouri School for the Blind and 12 students with normal sight from the Thomas Jefferson School, a boarding school in suburban St. Louis. The visually impaired students received eye exams and were divided into two groups: one in which visual problems were related to optic nerve disease and another group in which vision loss did not involve the optic nerve.

That’s a key difference, because the ipRG cells form part of the optic nerve and are likely damaged in diseases that affect the nerve. It would be logical to assume that blind children with damage to that part of the retina might have impairments in their non-visual function.

“There were no real differences in vision between the two groups,” says Van Gelder. “For the most part, the children in both groups were barely able to read the big ‘E’ on the eye chart, but those with optic nerve disease were 20 times more likely to be pathologically sleepy —as indicated by napping 20 or more minutes per day — than subjects with normal sight, and nine times more likely to have pathologic sleepiness than the kids who were blind from non-optic nerve diseases.”

Why? Extrapolating from the animal work, the hypothesis is that children with damage to the optic nerve also might have damaged, or missing, ipRG cells. Animals without those cells don’t make melanopsin and cryptochrome and cannot regulate their circadian clocks.

To measure the impact the loss of those cells might have in humans, Van Gelder’s team had study subjects wear a device known as an actigraph. Worn like a wristwatch, the actigraph measures every movement a person makes. A sophisticated computer algorithm then takes the movement data and determines whether a subject was awake or asleep, active or inactive. Study subjects wore the actigraphs continuously for two weeks.

Those with optic nerve disease had highly variable wake-up times and also had trouble falling asleep. The children who had optic nerve disease napped, on average, about 28 minutes a day, or eight minutes more than the definition of pathologic sleepiness. Actually, most subjects with optic nerve disease actually napped for almost an hour, but they took naps only every other day.

None of the children in the study had any other conditions that might contribute to sleep disorders, such as taking sedative drugs, attention-deficit hyperactivity disorder (ADHD), or being treated with Ritalin or other stimulant medications. So, the researchers believe the sleep problems these children experienced were directly related to their eye disease.

“Taken together, these results lead to the unexpected conclusion that eye disease can be a risk factor for sleep disorders, and the health of the optic nerve strongly influences risk,” Van Gelder says.

And there may be other risks, too. Heart attack, for example. For reasons not well understood, most heart attacks occur between 4 and 6 o’clock in the morning. Van Gelder says the body’s circadian clock somehow interacts with other systems to influence risk. It’s possible, he says, that by controlling the release of hormones, this non-visual system in the eye plays a role.

Another group at risk for loss of the non-visual system is patients with the eye disease glaucoma, which affects at least 2 million Americans and is the leading cause of blindness in African Americans. Glaucoma targets retinal ganglion cells like the ones that make melanopsin and cryptochrome. In severe cases, patients can lose 90 to 95 percent of their retinal ganglion cells. That could affect their ability to sense light with the non-visual system that Van Gelder and colleagues have been studying.

“We need to determine whether patients in the early stages of glaucoma show signs that they’re losing this second system,” he says. “If so, it’s possible they should be treated more aggressively.”

In future studies, he hopes to test whether treatment with melatonin will help regulate sleep patterns in children with optic nerve disease. But even before he learns whether it’s possible to help these patients synchronize their internal clocks to the outside world, Van Gelder believes it is important for health professionals to begin considering the impact of eye disease on sleep.

“Physicians and other health care professionals should be sensitive to the possibility of daytime sleepiness or insomnia, particularly in patients with severe optic nerve disease,” he says. “In the future, your eye doctor might want to make a point of asking you how you’ve been sleeping.”

Measuring the pupillary light responses of mutant mice

Infrared video recordings taken before and during exposure to relatively bright blue light show the responses of “blind” mutant mice—the yellow circles show the outlines of the pupils. Mouse “A” carries two copies of the retinal degeneration (rd) gene, and is visually blind; Mouse “B” carries this gene and is additionally mutant for both cryptochrome genes.

The “lights out” images were taken in darkness under infrared light. After lights are turned on, the pupil constricts in the retinal degenerate mouse, “A,” but not in the retinal degenerate mouse lacking cryptochromes, “B.” Studies in Russell N. Van Gelder’s lab have shown these mice to be only about 5 percent as sensitive to light as the “blind” retinal degenerate mice.