Eyes, Parts 1 & 2 - [a two-part essay on eye evolution]

February 15, 2005
Eyes, Part One: Opening Up the Russian Doll

(The first of a two-part post)

The eye has always had a special place in the study of evolution, and Darwin had a lot to do with that. He believed that natural selection could produce the complexity of nature, and to a nineteenth century naturalist, nothing seemed as complex as an eye, with its lens, cornea, retina, and other parts working together so exquisitely.The notion that natural selection could produce such an organ "seems, I freely confess, absurd in the highest possible degree," Darwin wrote in the Origin of Species.

For Darwin, the key word in that line was seems. He realized that if you look at the different sort of eyes out in the natural world, and consider the ways in which they could have evolved, the absurdity disappears. The objection that the human eye couldn't possibly have evolved, he wrote, “can hardly be considered real.”

The more scientists study the eye, the more they recognize that Darwin was right. This is not to say that they know everything about how the eye evolved. Evolutionary biology is not an automatic answer machine that can instantly tell you every detail about how eyes--or any other organ--evolved. Instead, scientists study eyes of different animals, the proteins they are made of, and the genes that store their recipe. They come up with hypotheses about how evolution could have produced these results. Those hypotheses then point the way to new experiments. In this way, evolutionary biology is no different from geology or meteorology, or any other science that illuminates the natural world.

To be precise, I should say that scientists study the evolution of "the eye." There are millions of different eyes (and other light-detecting organs), each built by a different species from its own unique set of genes. Closely related animals tend to have similar eyes, because they descend from recent ancestors. Some scientists study how eyes can adapt over a few million years to the special circumstances of a particular species. Other scientists step a little further back, to look at how the different types of eyes have evolved from simpler precursors. And other scientists step even further back in time, to find clues about where those simpler precursors came from. In this post, I will move back through time through these different stages of eye evolution (a la Richard Dawkins's The Ancestor's Tale.)

Humans have what's known as a camera eye. Light first passes through a cornea, which refracts the light. It then passes through a lens, which refracts the light further, so that it forms a focused image on the retina. We are primates, and so it's not surprising that all other primates have a similar type of eye. But different primates have important differences in the shape of their eye. Nocturnal primates have wider, more curved corneas than primates that are active during the day. A wider cornea lets nocturnal primates make the most of the moonlight by allowing more of it into the eye. Primates active during the day benefit from small flat corneas probably because the lens can sit further forward in the eye, producing a sharper image. This arrangement doesn't let as much light in, but during the daytime, that's no great loss. Chris Kirk of the University of Texas analyzed primate eyes in the December 2004 issue of The Anatomical Record (he has posted the paper on his web site).

For the most part, nocturnal and diurnal primates fit the same patterns as other mammals. But monkeys and apes (including humans) turn out to have extremely small, flat corneas, even compared to other primates that are active in the daytime. Kirk argues that this particular group of primates (called anthropoids) has experienced natural selection that has produced even sharper vision than found in other mammals active in the daytime. Other aspecsts of the anthropoid eye also make it sharp, including its fovea, a small spot on the retina that's incredibly dense with photoreceptors. In fact, anthropoids are matched only by raptors for their sharp vision. It's possible that our ancestors evolved such sharp eyes for hunting insects; monkeys and apes are also extremely social animals, and they rely on their keen eyes to look at one another and pick up subtle cues in their faces. Our ability to make sophisticated tools may have been made possible by the evolution of tiny corneas.

Changing the shape of an eye requires changing the molecules that make it up. Molecular fine-tuning can also alter an eye's ability to block out UV rays, to refract light at different angles, or to become more sensitive to different colors. Despite the fact that all vertebrates share the same basic eye plan, you can find a wide range of molecules inside them. Some are found only in fish, some only in lizards, some only in mammals.

How does one group of animals evolve one of these new molecules? One way is to borrow it. Joram Piatigorsky of the National Eye Institute and his colleagues have identified many of the molecules that make up the lens and cornea of humans and other animals. These molecules are practically identical to molecules found elsewhere in the body. Some are essential for the development of the head in an embryo. Others protect our cells from heat and other stress, others detoxify poisons that would otherwise build up in the blood.

Originally, the evidence indicates, many of the molecules found in eyes today were only produced in other parts of the body. But then, thanks to a mutation, the same gene began producing its molecule in the developing eye. It just so happened to have the physical properties that made it well suited to being in an eye. In later generations, natural selection favored mutations that made it work better in the eye.

But this new job in the eye may have posed a trade-off for the molecule's original job. Further fine-tuning may have only been possible when the gene went through a particularly drastic (but common) mutation: it duplicated. Now one copy of the gene could adapt to the eye, while the other continued specializing in its original job. (I wrote an essay a couple years ago about some of Piatigorsky's work in Natural History.)

Darwin didn't know about gene sharing or gene duplication, but he still managed to make some important observations about how the human eye could have evolved from a simpler precursor. Early eyes might have been nothing more than a patch of photosensitive cells that could tell an animal if it was in light or shadow. If that patch then evolved into a pit, it might also have been able to detect the direction of the light. Gradually, the eye could have taken on new functions, until at last it could produce full-blown images. Even today, you can find these sorts of proto-eyes in flatworms and other animals.

The closest invertebrate relatives of vertebrates fit nicely into Darwin's predictions. Amphioxus, which looks like a sardine with its head cut off, lacks a true brain or camera eyes. But the front end of its nerve cord is slightly swollen, and is built by many of the same genes that build a human brain. What's more, they grow a pit lined with light-sensitive cells which they seem to use to navigate through the water. The genes that build this pit are nearly identical to the ones that build our own.

The fact that Aphioxus has such a simple precursor to the vertebrate eye might suggest that this organ evolved from scratch. Yet eyes can be found on many other animals--which was how Darwin first figured out what a precursor to the vertebrate eye might have looked like. Eyes can found in insects, squid, and many other animals. Did they evolve independently?

The answer is yes and no. In the 1990s, Walter Gehring of the University of Basel and his colleagues discovered an essential eye-building gene called Pax-6 that was shared by insects and humans. If he inserted the human version of the gene into a fly larva, he got fly eyes popping up all over the fly's body. Gehring has proposed that Pax-6 is a master control gene, switching on an entire circuit of eye-building genes. In insects and in humans (and in all of the animals that share a common ancestor), this circuit builds eyes. But in each lineage, a different set of genes have been incorporated into this circuit, so that they can build eyes as different as the compound eye of an insect and the camera eye of a human.

The simplest explanation for so many animals sharing this same circuit is that they all inherited it from their common ancestor--a small worm-like creature known as a bilaterian that might have lived 570 million years ago. Exactly what sort of eye these genes produced in the Precambrian mists of time isn't clear, though. And until last fall, another feature of the eye didn't seem to fit this hypothesis: its photoreceptors. Invertebrate eyes and vertebrate eyes use different photoreceptors to sense light. But researchers have found that both kinds of photoreceptors grow on a humble animal known as a ragworm, which is believed to have branched off very early in the evolution of bilaterians. It's possible that the ancestor of living bilaterians produced both kinds of photoreceptors. One kind was lost in the vertebrate lineage, and the other was lost in the lineage that led to insects and other invertebrates with full-blown eyes.

Yet eyes are not limited to bilaterians. Jellyfish belong to a branch of animals known as cnidarians that split off from the ancestors of bilaterians some 600 million years ago. Some species have simple photoreceptors, while others have full-blown camera-eyes hanging from their tentacles. Biologists want to know whether these eyes evolved independently, or share some of the ancestral toolkit that produced human eyes and fly eyes. One hint that they share a common heritage is the fact that some of the genes that jellyfish use to build eyes bear a striking similarity to Pax-6 and other genes that build bilaterian eyes. On the other hand, most cnidarians (such as sea aneomones and corals) don't have eyes. What's more, jellyfish eyes are pretty weird compared to bilaterian eyes--for one thing, they don't wire up to a brain. The larvae of one species grow photoreceptors that don't even connect to a neuron. The photoreceptors link instead to hair-like structures in the same cell. Presumably light triggers these cells to flail their hairs to make the larva swim.

In years to come, the search for the roots of eye evolution will push even further back in time. In a paper in press at the Journal of Heredity, Walter Gehring points out that the first component of animal eyes to have evolved was the photoreceptor--a molecule that could catch light and turn it into a signal. One model for the origin of animal photoreceptors comes from colonies of algae, many of which have "eyespots" that allow them to swim towards the light so that they can photosynthesize. Perhaps early animals lived in colonies as well and had similar eyespots. Later, these simple photoreceptors evolved pigments and other molecules that helped capture more light, and eventually became able to form images.

But Gehring also proposes a weird but compelling alternative: our ancestors stole their eyes. Many times over the course of evolution, organisms have been engulfed by larger organisms, and the two have become integrated into a single being. Our cells, for example, contain mitochondria that we rely on to generate energy; originally, these were free-living oxygen-consuming bacteria. Another important fusion took place over two billion years ago, when bacteria that could carry out photosynthesis were consumed by an amoebae-like host. The bacteria then became a structure called the chloroplast, which can be found today in trees and other plants, as well as various sorts of algae. Increidbly, some of these algae were engulfed by other algae, which also came to depend on the photosynthesis carried out by the bacteria. Gehring likens these organisms to Russian dolls, with the original bacteria nestled deep within other organisms.

It's likely that before the bacteria were consumed again and again, they had already evolved a light-sensing molecule that helped them harness sunlight--perhaps by acting as a biological clock. The algae that devoured the bacteria may have retained the ability to sense light for the same purpose. Gehring points out that one group of these algae--dinoflagellates--have fused with corals, jellyfish, and other animals. It's possible that early animals may have incorporated the genes for light-sensing in their own genomes. If he's right, we gaze at the world with bacterial eyes.

February 16, 2005

In my last post, I went back in time, from the well-adapted eyes we are born with, to the ancient photoreceptors used by microbes billions of years ago. Now I'm going to reverse direction, moving forward through time, from animals that had fully functioning eyes to their descendants, which today can't see a thing.

This may seem like a ridiculous mismatch to my previous post. We start out with the rise of eyes, a complex story with all sorts of twists and turns, with gene stealing, gene borrowing, gene copying; and then we turn to a simple tale of loss, of degeneration, of a few genes mutating the wrong way and--poof!--billions of years of evolution undone.

In fact, loss is never such a simple matter. I can illustrate this fact with two disparate beasts: fleas and cavefish.

Cavefish were familiar to Darwin, as were the many other blind cave dwellers, such as salamanders and insects. Darwin saw cavefish as yet another example of an animal carrying around the vestiges of its ancestors, just as we carry around the stump of a tail. As for how cavefish lost their eyes, he set natural selection aside. Darwin could not imagine how a fish in a cave would get any benefit from eyes that did a worse job than its ancestors' eyes. "I attribute their loss soley to disuse," he wrote. By disuse, Darwin may well have been thinking along the lines of his precursor, Lamarck. As fish stopped relying on their eyes in the dark, somehow their eyes degenerated, and that degeneration was passed down to the next generation of fish.

Once scientists began to decipher the molecules of heredity, such an explanation became obsolete. Instead, some scientists translated the notion of "disuse" into the language of mutations. Like any animals, a cavefish has a small but real chance of undergoing a mutation to its DNA. In some cases, these mutations can impair the fish's eyes. In a population of surface-dwelling fish, this sort of mutation would probably make it hard for a fish to find food, and might even make it an easy target for predators. The chances of the fish passing down that mutant gene to a new generation of fish would be pretty slim. But in a cave, such a mutation would have no effect on the reproductive fortunes of a fish. Over time, the population of cavefish would accumulate lots of eye mutations, until their eyes were rendered useless.

But this "neutral mutation" hypothesis isn't the only possibility. Scientists have also proposed an "energy conservation" hypothesis. Mutations that prevent cave fish from developing eyes let them save energy, boosting their odds of survival.

Scientists have tested this hypothesis in recent years by studying the fish Astyanax mexicanus. You can find perfectly normal populations of this fish in surface waters in the U.S. , but if you go into caves, you can also find some 30 populations that are blind. This transformation has happened overnight, biologically speaking: scientists estimate that it was only 10,000 years ago that populations Astyanax moved into the caves. One vivid demonstration of just how recent this move was is the fact that a cave fish and a surface fish can mate and produce healthy hybrids. The lion's share of research on Astyanax has been carried out in the laboratory of William Jeffery at the University of Maryland, and he offers an excellent summary in a paper in press in the Journal of Heredity.

Much of Jeffery's work has gone into tracking the development of the fish from eggs. The most startling thing he has found is that cavefish grow eyes for quite a long time. Just as in surface fish, the brains of cave fish embryos bulge out to the sides, stretching into stalks that end in cups. A simple retina and lens begin to form, and growing nerves begin to link the retina to the visual centers of the fish brain. After about a day, however, the cavefish eye and surface fish eye begin to take different paths. The cave fish eye fails to develop an iris or a cornea, for example. Still, many parts of the cave fish eye continue to grow as their cells multiply.

These findings alone call into question both the neutral mutation hypothesis and the energy conservation hypothesis. If mutations were building up in the cave fish genome, you wouldn't expect that the fish could advance so far in the development of their eyes. And if energy conservation was the sole advantage driving the evolution of blindness, you wouldn't expect the fish to keep producing new eye cells, even as the eye begins to deteriorate.

Even the degeneration of the eye challenges both of these hypotheses. The eye doesn't collapse into a stew of chaos; it is dismantled in a stately choreography. The cells in the lens release some signal that instructs other eye cells to begin to commit suicide. In surface fish, the lens sends signals that do just the opposite, allowing the eye to develop fully. Jeffery and his colleagues found that if they transplanted just the lens of a surface fish into the eye of a cave fish, the cave fish grew a completely normal eye. What's more, the transplant triggered new nerve fibers to project from the retina to the brain, and the part of the cave fish's brain that handles vision even grew. It's possible that a transplanted lens allows a cave fish to see. Despite being blind, the cavefish still retains its original circuit of eye-building genes.

Jeffery and his colleagues have also tracked the degeneration of the eye at the level of genes. The neutral mutation hypothesis would lead you to expect that cave fish would express fewer genes in the eye than surface fish, because many of them would have been destroyed by mutations. But this is not the case, Jeffery and his colleagues have found. Instead, they're starting to identify some genes that make more of their proteins in the eyes of cave fish than in those of surface fish, and even some genes that aren't active in the eyes of surface fish at all.

One particularly important protein in the developpment of cavefish eyes is known as Hedgehog. In all vertebrates, Hedgehog plays a vital role in the development of the eye, starting at its earliest stage. Initially, the cells that will give rise to the eyes form a single cluster. Cells in the midline of the embryo start producing Hedgehog, which somehow signal the cells in the middle of this eye cluster to stop developing. As a result, only the cells on the far sides continue to develop, thus producing two separate eyes. Mutations that interfere with the production of Hedgehog can cause a gruesome birth defect in humans called cyclopia, in which a single cyclops-like eye develops.

Cave fish have evolved in the opposite direction: they produce more Hedgehog, rather than less. The extra protein stops the development of a wider expanse of the original eye-cell cluster, leaving few cells to progress. Jeffery and his colleagues confirmed this by boosting the production of Hedgehog in surface Astyanax. Not only do they develop smaller eyes, but they suffer the same lens-directed degeneration seen in cavefish. This means that the degeneration of cavefish eyes requires cells beyond the eyes to help coordinate the process.

What's most remarkable about this choreography is that it has evolved again and again. Studies on Astyanax DNA suggest that populations of surface fish have repeatedly invaded caves, and each time they have gone blind. Jeffery and his colleagues have started comparing the development of embryos from different populations, and they find the cavefish have evolved blindness through the same patterns of gene activity.

This parallel evolution is hardly what you'd expect from a random blast of neutral mutations. Nor does Jeffery believe that energy conservation can explain it. Males and females show no difference in the development of their eyes, despite the fact that females need a lot more energy to make their eggs. Likewise, some populations of cave fish get lots of energy because they live under colonies of bats that can drop food and guano into the water. Despite this luxurious conditions, these fish are no different than their leaner cousins.

Jeffery thinks that Hedgehog may be the key to understanding what's really driving the evolution of cavefish. Like many genes involved in development, Hedgehog has many different jobs. It is known to be essential for the development of tastebuds, for example, as well as teeth and the bones that make up the head. And in cave fish, all of these features are significantly different from surface fish. It's possible that these changes are adaptations that help the cave fish feed more efficiently. These changes were only made possible by cranking up the production of Hedgehog. A side effect of this increase was the destruction of the cave fish eyes. But because eyes aren't essential in the dark, this wasn't such a big price to pay. If Jeffery is right, Darwin's real mistake with cave fish wasn't falling back on a Lamarckian explanation. It was not recognizing how powerful natural selection could be.

Jeffery and his colleagues have managed learned so much about the evolution of cavefish eyes because they figured out how to turn Astyanax into a laboratory organism, which can be studied as carefully as a fruit fly or a lab rat. This sort of transformation takes many years, and only a few species have what it takes. Many other animals have lost their eyes, but in most cases, scientists can only glean less direct clues. Still, the stories they have to tell can be just as interesting. Most interesting of all is the fact that different evolutionary forces seem to have been at work.

Scientists know very little about the vision of fleas. As insects, fleas have inherited the standard insect eye, which consists of slender columns tightly packed together. But this standard insect eye has undergone drastic changes in fleas. Some fleas have what look like simple eyespots. Others seem to lack any eye at all. To learn about this transformation, a team of biologists from Brigham Young University have compared fleas to their relatives, which still have eyes.

This wouldn't have been possible even a few years ago, because scientists have only recently worked out the "flea tree." Fleas evolved from a group of insects with particularLY sharp vision. Their cousins include scorpionflies, which rely on their image-forming eyes to help them scavenge dead insects. Their closest relatives are "snow fleas" (Boreidae). These wingless insects live in mountains, where they feed on moss. They have small eyes, but can see well enough to jump away if you try to catch them. So it appears that fleas are the product of a long-term evolution towards simpler eyes.

The scientists used this tree to track the evolution of some of the molecules that are essential for vision. Known as opsins, they respond to light by triggering a chemical reaction that sends a signal from the eye to the brain. Opsins can be sensitive to different colors, depending on their shape, which depends in turn on the DNA sequence in their genes. The scientists isolated the gene for green opsins from 11 species of scorpionflies, snow fleas, and true fleas.

The scientists then compared the DNA sequences for signs of change. A mutation to an opsin gene may have no effect on the opsin molecule itself, or it may alter its structure dramatically. The difference depends on where in the DNA sequence that mutation strikes. The scientists found that most changes that occurred during the evolution of fleas had no effect on the actual opsins. They confirmed this by using the DNA sequence of the opsin genes to create computer models of the opsin molecules themselves. Even in fleas, the green opsin molecule has basically the same structure as in scorpionflies--despite their radically different eyes.

Just because a gene hasn't changed for millions of years doesn't mean that it hasn't been experiencing natural selection. The scientists found evidence that the opsin gene has been experience a special kind of natural selection in fleas and their relatives, known as purifying selection. Purifying selection occurs if even the slightest change to the structure of a molecule puts a serious dent in the reproductive success of an animal. The fact that fleas have experience purifying selection on their opsin gene means that it remains essential to their survival. (The details of their work appear in a paper in press at the journal Molecular Biology and Evolution.)

So what on Earth are the fleas doing with their opsins? The scientists doubt that the fleas are using them in their eyes. They point out that flea eyes are covered over in a tough layer of chitin, and they lack the lenses and other structures that would let them see. But in many animals, ranging from pigeons to salmon to butterflies, opsins have also been found outside the eye. In some animals, they grow inside the brain, while in others they grow on the abdomen or other parts of the body. Recent studies suggest that these opsins set the pace for biological clocks by registering the change of light from day to night.

This brings us back around to the very origin of eyes, which I described in my first post. Long before full-fledged eyes evolved, light-sensitive molecules may have existed in microbes, allowing them to change their movements during night and day. These molecules may have been incorporated into early eyes, making it possible for animals to see. But this transition didn't mean that photoreceptors could no longer serve their original function. Early insects may have used opsins both within their eyes to see and outside of their eyes as biological clocks. Later, some lineages of insects lost their eyes. Some may have lost them in dark caves. Fleas, on the other hand, lost their eyes as they became parasites. Instead of navigating through a complex landscape in search of a particular prey, they just hopped from one host to the next. But they still relied on opsins to run their biological clocks. The authors point out that scientists have also found opsins in other animals that have lost their eyes. The animals? None other than Astyanax.

What's particularly remarkable about the new study is how strongly the flea opsin resisted any evolutionary change--even after it was no longer being used in the flea eye. The molecule need the same functional structure for both jobs. As I mentioned at the beginning of my previous post, Charles Darwin recognized that the complexity of the eye might appear to pose a major challenge to his theory. To some people, it still does; they argue that the components of the eye cannot function on their own, and so they could never have existed on their own. By this reasoning, it would be impossible for one of these components--an opsin, for example--to do anything useful if it wasn't inside an eye.

Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.