Summer 1995 Volume 4, Number 3

he human eye is an extraordinary optical instrument. Its function is to focus the image of objects on the retina. One may have a sense of the precision of the optics of the eye if one thinks that the eye's lenses, the cornea and crystalline lens, project an image free of distortions on a photoreceptive layer whose thickness is less than 1/1000 of an inch. Thus, the eye matches the accuracy of the lenses of a good camera or a microscope. Optical instruments, however, are made by complex, high precision machines. How does the body manage to manufacture out of flesh and water such a precise organ?

At the heart of our research is the following question: Is eye growth totally programmed by our genes, or does the eye behave like a self-focusing camera, in which the sharpness of the retinal image somehow controls the expansion of the eye globe? Surprisingly, it looks like the answer may be in the nervous system.

We are large, fast animals and it is appropriate that our sight is adjusted to scan the horizon for large predators. Therefore, in the normal, or emmetropic, eye at rest, a light pencil from an object at an infinite distance is focused by the eye's optical (or dioptric) media onto the photosensitive cells of the retina. When, however, we need to observe nearby objects, a muscle contracts in our eye and the crystalline lens increases its power, in a process called "accommodation."

In myopia, the image of distant objects is formed in front of the retina, and, in most cases, this is due to the fact that the eye is too large from front to back. In hyperopic subjects, the opposite is true: The image of distant objects is formed behind the retina because the eye is too short. Both errors of refraction represent anomalies of development: In myopia the eye grew too much after birth and in hyperopia it grew too little.

In the U.S., the prevalence of myopia is about 25 percent of the population, but in Asian countries such as China or Japan it may exceed 70 percent, and it approximates 90 percent in selected populations, such as Chinese university students.

The cause of myopia has long been the subject of a spirited, nature-versus-nurture controversy. There is little doubt that heredity plays an important role in the genesis of refractive errors. Twin studies have been especially convincing in this regard, as well as the fact that myopia occurs in families and is more frequent in certain ethnic groups and races.

Equally compelling is the evidence supporting the idea that environment plays a role in the development of myopia: For instance, among Eskimos the incidence of nearsightness has considerably increased with the advent of electricity and schooling.

To understand the mechanism of myopia, it was necessary to find an animal model in which the refractive state could be experimentally manipulated. This became possible when we observed that suturing shut, or fusing, the lids in neonatal macaque monkeys induces excessive growth of the posterior segment of the eye globe, thus causing a refractive error that is very similar to intermediate myopia in humans.

At that point, we asked ourselves the crucial question: Is lid-fusion myopia caused by some trivial effect of the sutured lids on the growing eye, such as pressure or an increase of the temperature of the orbit? If so, our model would have no counterpart in the human condition, and its clinical value would be slight.

Another possibility, however, was that the closed lids were interfering with vision at a critical stage of development: Somehow the nervous system was responding to the abnormal visual experience by causing axial elongation. And, in fact, our next investigations did persuasively implicate the nervous system in axial elongation.

How can the nervous system influence the growth of the eye and therefore determine its refractive state? First, we chose to interfere with the peripheral control of accommodation. We closed the lids of one eye in monkeys belonging to two species, the stump-tailed macaque and the rhesus macaque. Then, we administered atropine daily to the closed eye by introducing it into the conjunctival sac. (Atropine is a drug that paralyzes accommodation.) To our surprise, we obtained different results in the two macaque species.

In the stump-tailed macaque, myopia did not develop. The rate of growth was the same in both eyes and, after one year of atropine administration, the closed eye had about the same axial length and refraction as the open one. This experiment indicates that accommodation, which is controlled by the brain, has, in addition to its role of focusing on close objects, a role in axial elongation in the stump-tailed macaque. Perhaps the macaque's central nervous system responds to the blurred images transmitted by the closed lids by trying to focus them through accommodation and this in turn induces excessive growth of the eye and subsequent myopia.

In the rhesus macaque, administration of atropine in the closed eye did not prevent abnormal eye growth; myopia developed. We concluded that, in the rhesus macaque, accommodation does not play a role in axial elongation. Could it be possible that in the rhesus monkey the retina itself, rather than the brain, is involved in causing nearsightedness?

We designed the following experiment on both types of macaques to test this hypothesis: If we suppressed completely the visual input to the brain by interrupting the fibers of both optic nerves and then we sutured the lids of one eye, would myopia still develop?

If our hypothesis was correct, that the rhesus retina, not its brain, caused myopia, the closed eye should become nearsighted in the rhesus monkey, but not in the stump-tailed macaque. This was precisely the result obtained. We concluded that, upon eye closure in the rhesus monkey, an abnormal message from the retina to the other components of the eye causes excessive eye elongation.

The retina is designed to respond vigorously to sharp edges; we can imagine that, after birth, pattern discrimination modulates the release by the retina of regulatory molecules that are responsible for the fine tuning of eye growth. For the rhesus macaque, if the visual world degenerates to featureless shadows as a result of eye closure, altered amounts of these molecules are produced and the axial length of the eye abnormally increases. In the stump-tailed macaque, the retinal contribution to axial elongation may be slight, but the centers of the brain stem respond to the loss of sharp edges by inducing excessive accommodation.

Does our macaque model cast any light on human myopia? Does it suggest means of preventing this refractive error in humans? Although at the present stage of our investigations it is premature to draw any conclusion as to the human condition, it may be worthwhile to formulate a set of hypotheses that may be useful in designing clinical studies.

When young humans are exposed to a world without edges, as in cases of opacities of the dioptric media of the eye, they do develop myopia, just as our juvenile monkeys. Thus, our model of myopia is quite pertinent to the human condition.

On the other hand, prolonged near work in predisposed children and adolescents may cause myopia through excessive accommodation. These two mechanisms may operate independently of one another in different subjects and be combined in others. This would explain why attempts to prevent myopia in children by the administration of agents that paralyze accommodation met with various degrees of success. If, in certain individuals, the cause of myopia is a genetic defect of the retina, it does not seem very useful to administer atropine to all children who exhibit the tendency to become nearsighted. By the time one realizes that accommodation is not responsible for the abnormal refractive state, it will be too late.

However, if we could identify the molecules that are released by the retina and regulate eye growth, one can perhaps devise means to control the axial length in all individuals without paralyzing the muscle of accommodation.

We have made a promising step in this direction. In collaboration with Richard A. Stone and Alan M. Laties, two opthalmologists from the University of Pennsylvania, we observed that a particular neuropeptide is more abundant in the closed than in the open eye of lid-sutured rhesus macaques. This neuropeptide is the vasoactive intestinal peptide (VIP), and it is contained in a class of retinal neurons called the amacrine cells.

Thus it appears as though the neonatal retina, when presented with a visual world without edges, responds to this abnormal visual experience with an altered release of a neuromodulator, perhaps VIP, and that this may in turn affect eye growth. We are now investigating whether VIP is one of the molecules that control postnatal eye development either directly or by regulating the production of other growth factors. In addition, we are trying to identify the genes that regulate the growth of the eye.

The main achievement of our studies was the realization that the nervous system plays a fundamental role in the development of myopia and that a drastic change in the visual environment may affect the postnatal growth of the eye. Thus, both nature and nurture participate in determining the final shape of the eye and its refractive state.

The challenge now is to discover the molecular mechanisms that control eye growth. These mechanisms may eventually lead to methods of intervention that may be as exacting and elegant as the functioning of the eye itself. *

Dr. Raviola is Bullard Professor of Neurobiology and Professor of Ophthalmology at Harvard Medical School. Dr. Wiesel, formerly Chairman of the Department of Neurobiology at Harvard Medical School, is President of the Rockefeller University, NY. In 1981, Drs. Wiesel and David Hubel received the Nobel prize in Medicine or Physiology for their work on the visual cortex.