Natural History, April 2002
“The eye to this day gives me a cold shudder,” Charles Darwin once
wrote to a friend. If his theory of evolution was everything he thought
it was, a complex organ such as the eye could not lie beyond its reach.
And no one appreciated the beautiful construction of the eye more than
Darwin-from the way the lens was perfectly positioned to focus light
onto the retina to the way the iris adjusted the amount of light that
could enter the eye. In the Origin of Species, he wrote that the idea
of natural selection producing the eye “seems, I freely confess, absurd
in the highest possible degree.”
But Darwin also knew that a cold shudder was not reason enough to abandon a scientific theory. He looked at the eyes of many different animals-from flatworms to crustaceans to vertebrates-and found among them a gradation of forms, from a simple patch of light-sensitive tissue all the way to an elaborate image-forming organ complete with lens, iris, and retina. He decided there was no reason that evolution could not have led gradually from one arrangement to another.
Darwin considered only the anatomy of the eye, because the biochemistry of vision was still a mystery in his day. Of course, an eye is far more than just what can be seen by another eye. All the work involved in vision-bending and catching light, fine-tuning the image that gets sent to the brain, keeping the eye clear and firm over the years-is carried out by an army of specialized molecules, produced in turn by specialized types of cells. And these cells contain genes and proteins that interact with one another in a dense web of cooperation and control, of feedback and inhibition. If Darwin could have seen the molecular complexity of the eye, his shudder might well have turned even colder.
But before too long, the shudder would fade. As scientists have uncovered the biochemical intricacies of the eye, they’ve also made great strides in understanding how it has evolved. In the process, they’ve come face-to-face with evolution’s remarkable laziness. Instead of giving rise to entirely new genes, evolution has in many cases simply borrowed old ones.
The story of this discovery begins in the 1960s, when scientists started to study the molecules that make up one important part of the vertebrate eye, the lens. The lens is essentially a blob of clear skin cells. As an embryo develops, a patch of cells on each side of its head begins to differentiate from the surrounding tissue. These cells start producing protein molecules called crystallins, which make up 90 percent of the protein in the lens. Soon the cells become little more than bags of crystallin.
Thanks to their structure, crystallins make a lens act as if it were made of glass. They bend the light as it passes through, and because they pack tightly together in an orderly way, rather than sticking together in irregular clumps, they don’t scatter the rays randomly. Crystallins are also incredibly tough-and they need to be, because they can’t be replaced once they’ve been formed. They are the longest-lived proteins in the body; many of the crystallins in the eyes of a centenarian were there when he or she was an embryo. If they become damaged and start clumping together, cataracts form.
When scientists first began investigating various vertebrate lenses, they expected to find at most just a few kinds of crystallins. Other molecules that carry out equally specialized jobs-light-sensitive rhodopsin in the retina, for example, and oxygen-ferrying hemoglobin in red blood cells-are pretty much identical in any vertebrate you care to examine, from a parrot to a python. But by the 1970s it became evident that crystallins are unusual in this respect: they come in a surprising variety of different structures, each of which interacts with light in a unique way. And the different crystallins are not mixed together randomly; as the lens grows, it builds up rings like an onion, and each new ring is made up of distinctive proportions of the various crystallins. These different combinations give each ring the ability to bend light at a particular angle. As a result, the entire lens can focus light onto a small spot on the retina.
But an even bigger surprise was in store for researchers. It turns out that certain vertebrates possess unique types of crystallins, present in no other eyes. Birds and reptiles, for example, all have lens proteins (dubbed delta-crystallins) that mammals and amphibians lack. Amphibians have crystallins of their own, as do mammals. These discoveries hinted that the vertebrate eye has had a turbulent evolutionary history.
The fossil record suggests that the first full-blown vertebrate eye arose 530 million years ago in a primitive fish. As the fish’s descendants diverged into new forms, new crystallins evolved in their eyes. Birds and reptiles, for example, descend from a common ancestor that lived about 300 million years ago-an ancestor that amphibians and mammals do not share-and this ancient creature presumably evolved delta-crystallins and passed them on to its descendants. New crystallins did not simply evolve as minor variations on old ones, however. Biologists can group the proteins into distinct “families,” and these families bear little resemblance to one another. Somehow, evolution devised transparent proteins again and again.
A clue to how this reinvention took place came in the 1980s, when scientists learned how to read the sequence of amino acids in crystallins. They discovered that alpha-crystallins, which are found in all vertebrate eyes, are strikingly similar to the common molecules known as heat-shock proteins. Heat-shock proteins help other proteins function at high temperatures and under other kinds of stress. Without the help of heat-shock proteins, a stressed-out protein may unfold and lose its intricate shape. But a heat-shock protein can cradle other proteins and protect them from assaults on their structure. Heat-shock proteins are crucial to every organism on earth; in our own bodies they are at work in almost every cell.
Before long, researchers found that, just as alpha-crystallins have a similarity to heat-shock proteins, other families of crystallins bear a striking similarity to other common proteins. Their closest matches are enzymes that help with the basic metabolism of cells. Some of these enzymes turn food into energy, while others detoxify poisonous wastes that would otherwise build up in the blood.
At first, scientists had only a rough idea of the similarity between crystallins, on the one hand, and these enzymes and heat-shock proteins on the other. But by the end of the 1980s they had an answer, and a surprising one at that. In many cases, a crystallin and its related protein aren’t produced by two similar genes; both are encoded by a single gene.
These double-duty genes did not start out making crystallins. Heat-shock proteins can be found not just in mammals but in other animals as well as in plants and microbes. They must have evolved early in the history of life, long before eyes came into existence. The same goes for the crystallin-like metabolic enzymes. So how could proteins adapted for these primitive sorts of functions later wind up moonlighting as transparent crystallins? The gulf may seem wide, but the evolution required to cross it is actually minimal.
A gene consists of more than just a code for building a specific protein. At the beginning of its sequence are short stretches of DNA that act like on-off switches. The gene’s code can be read only if certain proteins grab onto its switches and activate the gene, and these particular proteins are produced only under particular circumstances. This system guarantees that genes will become active only at the right place and the right time. If not for these switches, our bodies would be a proteinaceous mess. Red blood cells might be filled with bone instead of hemoglobin; our teeth might be made of hair instead of enamel.
Because these switches are made of DNA, they can mutate like any other part of a gene. In one common kind of mutation, these switches are accidentally sliced off one gene and pasted onto another. Getting an extra switch this way gives the recipient gene the power to become active under an additional set of conditions. This seems to be how lens crystallins came into existence. Some of the genes that make heat-shock proteins and metabolic enzymes accidentally picked up extra switches, and as a result they started to become active inside the cells of developing eyes.
In most cases, these moonlighting genes produced proteins whose structure was poorly suited for a lens. But in a few cases, they made a protein that could bend light without absorbing it. At first the protein may not have done a very good job, but even a lousy crystallin is sometimes better than none at all. Natural selection would have favored these genes with their new switches, and they would have become widespread. Through later mutations, these genes might have produced proteins better adapted to the eye, with a structure that was more durable or that improved their owner’s eyesight.
But Graeme Wistow and Joram Piatigorsky, two crystallin experts with the National Eye Institute of the U.S. National Institutes of Health, have pointed out that such two-switch genes may face a conflict of interest. Say that a detoxification gene has a part-time job producing lens crystallin. A mutation that improves the crystallin-making it more durable, for example-could make the protein do a worse job at detoxifying cells. The benefit brought by the mutation might not be favored by natural selection because of the harm it caused at the same time. This trade-off, which Wistow and Piatigorsky have dubbed “adaptive conflict,” may make genes evolve more slowly than scientists would otherwise expect.
A striking example of this sort of adaptive conflict can be found in the blind mole rat (superspecies Spalax ehrenbergi), which lives in dark underground tunnels in the Middle East. This mammal’s ancestors began living underground about 40 million years ago and probably lost the ability to see fairly soon thereafter. As an embryo, a blind mole rat begins to develop eyes, but they quickly degenerate and get buried in the surrounding tissue. If a blind mole rat acquires a mutation that damages one of its lens genes, it won’t pay a penalty. So you’d expect that after millions of years underground, a lot of mutations would have accumulated in Spalax’s crystallin genes.
Yet that’s not the case. The crystallin produced in the blind mole rat’s eyes is still surprisingly similar to the crystallin found in the eyes of other rodents that can still see. The genes haven’t changed much, apparently because they remain active in other parts of Spalax’s body. They must provide some other sort of crucial service that would be wiped out by too many mutations.
Wistow and Piatigorsky have suggested that one escape from an adaptive conflict is for a crystallin gene to get duplicated. It’s relatively common for an extra version of a gene to be accidentally produced while DNA is being copied. Once there are two copies of a shared gene, one of them can become better adapted for one job, while the other can devote itself to the second. This seems to have happened many times over in the vertebrate lens. It’s not unusual for scientists to come across two versions of a crystallin gene, one of which still makes its protein in other parts of the body, and a nearly identical copy that is active only in the lens.
Scientists are finding that evolution has borrowed genes to create new crystallins many times in the history of life-not only in vertebrates but also in squid, octopuses, and insects. Invertebrates have radically different kinds of eyes; insect eyes, for example, are made up of hundreds or thousands of columns, each of which captures a tiny fragment of the surroundings. To fill those columns with crystallins, insects have borrowed the gene for one of the proteins in their exoskeleton. Scientists have even found shared genes making crystallins in the eyes of jellyfish, although these eyes-tiny photoreceptors connected not to a brain but to the surrounding muscles-barely fit the meaning of the word.
Once vertebrate and invertebrate eyes were established hundreds of millions of years ago, evolution continued borrowing genes and fine-tuning them for new situations. For example, fish have hard lenses focused for seeing at close range, which makes them well suited to the ocean, where visibility is limited. When vertebrates came on land, they acquired new crystallins that softened up their lenses and allowed them to focus on objects much farther away.
One of the most fascinating examples of a newly borrowed lens gene can be found in the eyes of geckos. Perhaps 100 million years ago, the ancestors of today’s geckos became nocturnal predators. Their eyes changed dramatically in the process. Their pupils became permanently locked at their widest opening, and their eyelids fused to their eyes to become transparent spectacles. (Unable to blink, geckos use their tongues to clean their eyes.) Later on, a few lineages shifted back to life in the sun. Their descendants can be found in Africa, the Middle East, and Latin America. Living in deserts and other open spaces, they are bombarded by damaging ultraviolet rays. But as a result of their nocturnal heritage, they can’t use their eyelids to shade their eyes, nor can they narrow their pupils to cut down on the ultraviolet rays.
Beate Röll, a biologist at Ruhr-Universität Bochum in Germany, and her colleagues have discovered the gecko’s solution: just borrow another gene. The one in question, named CRBP1, makes a protein (found in all animals) that can store vitamin A and ferry it around a cell. Röll discovered that in several different gecko lineages, CRBP1 has been slightly altered to produce new crystallins. Thanks to their structural heritage, these proteins can hold on to a version of vitamin A, which absorbs ultraviolet rays. Equipped with the new crystallins, the lens acts as a filter, shielding the eye from harmful radiation and letting visible light pass though.
Röll has shown that after the CRBP1 gene was co-opted by the gecko eye, it adapted to its new role by producing a more compact protein, one that may last longer and refract light more effectively. But these changes haven’t interfered with the protein’s old job as a vitamin ferry; Röll can detect the gene at work within other organs in the gecko.
The remarkable history of crystallins offers many important insights. By tracing how crystallins have come into existence, scientists can learn how they function in the eye. It turns out, for example, that some crystallins descended from heat-shock proteins may still be able to help protect other proteins in the lens from stress. At the same time, the story of double-duty proteins such as crystallins should be taken as a warning by those who hope to treat diseases with gene therapy. One can’t assume that each gene does only one job. If scientists try to cure a defect in one part of a patient’s body by tinkering with a faulty gene, they may be surprised to discover that the gene actually plays a completely different role in another part of the body-and this role may be compromised by the therapy.
But the most profound lesson is that evolution works in quirky ways. Almost half our genes are duplicates, and many of them have been copied so often that they now form giant gene families. The kinds of gene borrowing and adaptive conflict that have created our lenses have probably been at work throughout our bodies. This sort of evolution is both playful and pragmatic: it experiments with new uses for existing genes and leaves the precision tuning for later. It’s enough to turn a cold shudder into a smile.
Copyright © 2002 Carl Zimmer. Reproduction or distribution is prohibited without permission from the author.