It took over a year and a half to get skilled enough in working with the technical aspects of studying butterflies so that he could design interesting experiments in molecular development. At first, the experimental strategy wasn’t that directed. The hope was that by using monoclonal antibodies made for studying fruit flies, they might be able to identify similar molecules at work in the developing butterfly. As Carroll explains, “It was sort of a round up the suspects approach, which was if all of these genes did interesting things in fruit flies, then maybe they did something interesting in butterflies. And one of them, it turned out, did something spectacularly interesting.” Carroll’s technician and co-author on the important paper, Julie Gates, called Carroll to the lab to look at the development of butterfly wings stained with a monoclonal antibody to the fruit fly gene Distal-less. A homeo-transcription factor (a factor implicated in what William Bateson called homoeosis), Distal-less is the first gene expressed in the development of limbs in drosophila. It had recently been shown to affect the formation of imaginal discs during drosophila development. These discs then become the regions for further differentiation in larva, where, depending on their placement, they can become wings, legs, part of the mouth, or even the antennae. What Gates and Carroll saw, was that this “gene was turned on in spots during the caterpillar stage” that directly corresponded with the development of wings spots in the butterfly. As the New York Times described: “These disks are positioned just where the eyespot forms on the adult's wing, and presumably lay down its foundation.”
The image at right is the original image from Carroll, et. al.'s original paper. The arrows in frames C and D of the illustration point to spots that stained positive with anti-Distal-less monoclonal antibody. Frames E and F show the stain locations as matured butterfly wing spots in the fore-wing and hind-wing respectively.
Although Carroll and company were not prepared for the results (he later described it as if a “2000 pound anvil had been dropped on his head”) by Monday they had worked up its evolutionary significance. The same molecules implicated in the development of limbs in flies, were used for a “novel” developmental purpose in butterflies, to develop spots on the wings.” This evolutionary feature, known as co-option, helped explain how complex features were developed from pre-existing traits. “Novelty”, Carroll explained “doesn’t come from scratch”. “It comes from using available bits in new ways.” As he explained to the NY Times shortly after the publication of his lab’s paper, “There's a limited set of genes and proteins in the universe, and the way you generate diversity is by making small variations on these standard regulatory programs."