SCIENCE . Vol. 275 . 28 MARCH 1997
Human eyes, fly eyes, and horseshoe crab eyes, to name a few, differ so greatly that it would seem nature invented eyes dozens of times in the course of evolution. A blow to this argument came 2 years ago when a mouse eye gene spliced into fruit flies prompted them to form extra fly eyes on their bodies, suggesting that a single, ancient genetic program kicks off eye development throughout the animal kingdom.
Now this feat has been duplicated in fruit flies, using a gene from an animal even more evolutionarily remote from a fly than is a mouse: a squid. The finding, reported in the 18 March issue of the Proceedings of the National Academy of Sciences, is the strongest evidence yet that all animals with eyes inherited them from a common ancestor.
Eyeful. Eye produced on a fruit fly's wing by a squid gene (A) resembles one produced by a mouse gene (B).
A group led by molecular biologists Stanislav Tomarev and Joram Piatigorsky of the National Eye Institute in Bethesda, Maryland, and developmental geneticist Walter Gehring of the University of Basel in Switzerland, found that both the mouse and the squid Loligo opalescens carry a gene called Pax-6. In squid embryos, they found, the gene is active in the brain, olfactory organs, and the iris and lens of the eye - just where it's active in mice. What's more, when the researchers engineered Drosophila fruit flies to express squid Pax-6, extraneous fly eyes sprouted in locations where the flies' wings, legs, and antennae normally grow. That's just what happens when the fly version of Pax-6, called eyeless, is activated in these areas.
The results strengthen the idea that eyes evolved only once, and suggest that Pax-6/eyeless has acted as a key regulator of eye development since before vertebrates and invertebrates parted ways some 500 million to 600 million years ago, the researchers say.
"If [the gene] were active only in flies and mammals, then you could suggest there was independent recruitment of the same gene to make eyes" in different species, says Tomarev. "But when it happens many times in many organisms, that's harder to imagine."