RESEARCH N EWS

Focusing on the eyeless Gene

Activating eyeless induces formation of full-fledged eyes in fruit fly wings, legs, and other tissues. This suggests it may be a "master control gene" for eye development.

You've heard the expression, "He's got eyes in the back of his head." Of course, it isn't meant to be taken literally-people can't sprout eyes on the back of the head, or any other part of their anatomy. But the same can no longer be said for fruit flies. On page 1788, Walter Gehring's group at the University of Basel in Switzerland reports that, when a gene called eyeless is turned on in parts of flies where it wouldn't normally be active, the flies grow extra eyes. The Gehring team's experiment produced flies with eyes on their wings, eyes on their legs, and eyes on their antennae. Some flies had as many as 14 fully developed eyes.

"It is really remarkable that you can take a tissue that would normally make a wing or an antenna and by turning on one [gene], make that into a complex thing like the eye," says fruit fly geneticist Gerald Rubin of the University of California (UC), Berkeley. "This is like someone finding a gene that would turn a kidney into a liver."

That kind of capability makes eyeless the best candidate yet to be the elusive quarry that developmental biologists call a "master control gene," a gene that single-handedly triggers the formation of an organ or structure. The protein produced by the eyeless gene has all the hallmarks of a transcription factor, a protein that turns genes on or off. It apparently "binds to a distinct set of genes that starts the whole process to make eyes," says Larry Zipursky, who studies fruit fly eye development at UC Los Angeles. With the help of eyeless, Zipursky says, researchers should be able to "piece together the steps" by which eyes are made.

All eyes. The consequences of abnormal eyeless activation can be seen in these eyes on the antenna (left) and leg (right) of a fruit fly.

The work has evolutionary implications as well. Gehring's group has shown that the mouse counterpart of eyeless also causes eye formation when it is put into fruit flies. That suggests that the mouse gene functions in the same way as the fly gene, which in turn implies that the gene's role as a master regulator of eye development is very ancient dating back half a billion years to the common ancestor of flies and mammals.

Gehring's group didn't set out to plumb the history of eye evolution, or even to find the master gene for eye formation. In fact, the team's venture into eye development began "quite accidentally," says Gehring. In 1993, graduate student Rebecca Quiring was searching for fly genes that code for transcription factors. In the course of her search, she pulled out the gene for one such factor, and postdoc Uwe Walldorf found that it mapped to the same chromosomal site as the eyeless gene. Although not yet cloned, eyeless had been identified years earlier through its effects in genetic studies: When mutated, the gene produces flies with deformed eyes or no eyes at all.

Upon analysis of the gene, Quiring and Walldorf made a startling discovery: The fruit fly eyeless gene is very similar to genes that play roles in eye formation in mice and humans. The mouse gene is called Small eye because mutations in one of its two copies yield underdeveloped eyes; damage to both copies eliminates the eyes altogether. The human version of the gene, Aniridia, was found because mutations in one copy cause defects in the iris, lens, cornea, and retina.

The Gehring group's discovery that eyeless is the fruit fly counterpart of Aniridia and Small eye, published last August in Science (5 August 1994, p. 785), came as a surprise, says Charles Zuker of UC San Diego, who wrote a Perspective that accompanied the paper, because the compound eyes of insects and the single-lens eyes of vertebrates are so different that biologists have generally assumed they evolved independently. "There are a dozen ways you can make image-forming eyes, just based on optics," says Zuker. "If you ask how many of those are found in evolution, you find that each and every one is." Such diversity had led researchers to conclude that eyes evolved independently 40 or more times. But the common requirement for the eyeless gene stood out as a striking link between such divergent eyes, suggesting that the eyes of insects and mammals didn't spring from different origins, but somewhere in the distant past shared a common ancestor.

Aside from its evolutionary implications, the finding raised in Gehring's mind a possible connection to a phenomenon he had studied 30 years ago as a graduate student with Ernst Hadorn at the University of Zurich. In fruit flies and other insects, embryonic tissues called imaginal disks give rise to adult structures such as wings, eyes, and legs. When imaginal disks are removed from fruit fly larvae and induced to develop in culture, they usually retain their identity-wing disks form wings, leg disks form legs, and so on. But Hadorn found that occasionally a disk would shift identity - a wing disk, for example, might form an eye - a process he dubbed transdetermination.

The mystery of transdetermination was never solved, but Gehring and others came to suspect that it resulted from the accidental switching-on of master genes that had the power to turn on a different developmental program. When Quiring cloned eyeless, Gehring had a hunch it might be one of these putative master control genes. Gehring postdoc Patrick Callaerts and graduate student Georg Halder tested that hunch by engineering flies to express the eyeless gene in imaginal disks destined to form wings, legs, and antennae.

"If you rounded up 10 people and asked, 'Do you think you could actually [make eyes with such an experiment],' you would probably get 10 different reasons why it wouldn't work," says Berkeley's Rubin. But Gehring would not have been among the skeptics. Recalling the transdetermination experiments of 30 years ago and the "brilliant red pigmented eye facets" that arose in his wingdisk cultures, Gehring says he believed the experiment would generate eyes.

He was right. "Flies started to hatch with huge eyes on their wings," Gehring says. And on their legs. Some even had eyes on the ends of their antennae that looked "like little crab eyes" on stalks. The eyes appeared to be perfect copies of normal eyes. "These eyes have everything a [normal] eye has, from the bristles to the lens to the pigment cells," says UCSD's Zuker. And Gehring says his team recently confirmed that the eyes are functional as well: Their photoreceptor cells respond to light.

These results stand in stark contrast to what usually happens when developmental biologists coax a tissue to take steps down the wrong developmental path. Typically the wayward tissue may make a protein or a partial structure characteristic of the path it is forced to take-but not a whole organ. In the case of eyeless, says Rubin, "you are definitely not talking about the tissue just expressing an eye [protein]. You are making a whole tissue with all its complexities."

This is not, however, the first case in which genetic engineering has been used to design flies with complete structures in the wrong place. In 1987, for example, Gehring's group turned on a gene called Antennapedia in the heads of flies, where it is not normally expressed, and as a result the flies grew legs where their antennae should have been. But that experiment and others like it were fundamentally different from the eyeless experiment, says William McGinnis, who studies fruit fly development at Yale University. Antennapedia is not a master gene for leg formation he says, but instead is concerned with "assigning a spatial position" to a tissue. Turning on Antennapedia in the head tells the head to develop as the midbody would, and that includes the formation of legs instead of antennae. In contrast, the eyeless gene, McGinnis says, specifies "a functional organ rather than a spatial identity."

The ability to induce a complete organ is what appears to make eyeless a "master control gene." And that is particularly exciting for researchers who study eye development in fruit flies, says UCLA's Zipursky. A great deal is known about the genes that cause the production of the different cell types in the eye-genes such as sevenless, which triggers the formation of one of a specific type of photoreceptor cell, or the genes that code for the rhodopsin pigments that detect light. But little is known about the hierarchy of regulatory genes that turn on these and the hundreds of other genes necessary to form eyes. eyeless may help change that. "What's very exciting here is that eyeless is really high in that hierarchy," Zipursky says. Using a variety of experimental methods, researchers should be able to find the genes that are activated by eyeless and thus begin to fill in the complex cascade of genes turning on genes that leads to the eventual activation of the structural genes that make eyes.

The current paper also strengthens the evolutionary connection that was made when the link between eyeless, Small eye, and Aniridia was discovered, says Nipam Patel, a developmental biologist at the University of Chicago who studies the evolution of genes that control development. As different as the eyes of flies and mice are, says Patel, the finding that they not only share a common control gene, but that the mouse form of the gene can function in flies, is powerful evidence that they have a common ancestry.

Patel says evidence of a common ancestry in no way contradicts the view that image-forming eyes evolved independently. It merely means they diverged from a common ancestral eye and that the development of that eye was likely governed by the ancestor of the eyeless gene. That early eye, says Patel, may have been a mere eyespot, a cluster of light-sensitive cells with no image-forming ability that is common in lower animals. But as nature improved on that eye in all the ways that led to various image-forming eyes, it apparently continued to use the eyeless gene to control the ever-more complicated process of eye development.

That suggests that eyeless should participate in eye formation in other species as well. And that may well be true. Gehring's group has already found counterparts of eyeless in a diverse range of animals including squid- which have very advanced image-forming eyes-and planarians, tiny flatworms with rudimentary eyespots, although the gene has not yet been shown to be essential for eye formation in these animals.

As researchers use eyeless to probe the mysteries of eye formation in fruit flies and other animals, they will undoubtedly make comparisons up and down the phylogenetic tree. And the similarities and differences they find will lead to a better understanding of just how much our eyes do have in common with those of our distant animal relatives-as well as how our eyes are formed. The one thing it won't do is give us eyes in the back of our heads, useful as that might be.

-Marcia Barinaga

SCIENCE · VOL. 267 · 24 MARCH 1995 pp. 1766-1767

Return To Molecular Biology Main Page

Return To Course Materials