366 Alex Rosenberg
the various ectopic and gene deletion experiments employed to formulate the why-
necessary explanation for the major pattern).
The developmental molecular biologists, S.B. Carroll and colleagues, who re-
ported the beginnings of the proximal explanation sketched above, eventually
turned their attention to elucidating the ultimate explanation. Carrollet al.,
write:
The eyespots on butterfly wings are a recently derived evolutionary
novelty that arose in a subset of the Lepidoptera and play an important
role in predator avoidance. The production of the eyespot pattern is
controlled by a developmental organizer called the focus, which induces
the surrounding cells to synthesize specific pigments. The evolution of
the developmental mechanisms that establish focus was therefore the
key to the origin of butterfly eyespots. [Carrollet al., 1999, 532]What Carroll’s team discovered is that the genes and the entire regulatory path-
way that integrates them and which control anterior/posterior wing development
in theDrosophila(or its common ancestor with butterflies) have been recruited
and modified to develop the eyespot focus. This discovery of the “facility with
which new developmental functions can evolve... within extant structures” [p. 534]
would have been impossible without the successful why-necessary answer to the
proximate question of developmental biology.
Besides the genes noted above, there is another,Hedgehogwhose expression
is of particular importance in the initial division of theDrosophilawing imagi-
nal disk into anterior and posterior segments. As in the fruit fly, inPrecisthe
hedgehoggene is expressed in all cells of the posterior compartment of the wing.
but its rate of expression is even higher in the cells that surround the foci of the
eyespot. InDrosophila,Hedgehog’s control over anterior/posterior differentiation
appears to be the result of a feed back system at the anterior/posterior bound-
ary involving four other gene products, and in particular one,engrailed,which
represses another,cubitus interruptus(hereafter ‘ci’ for short), in the fruit fly’s
posterior compartment. This same feed back look is to be found in the butterfly
wing posterior compartment, except that here theengrailedgene’s products do not
repressciexpression in the anterior compartment of the wing. The expression of
engrailed’s andCi’s gene-products together result in the development of the focus
of the eyespot.. One piece of evidence that switching on theHedgehog-engrailed-
cigene system produces the eyespot comes from the discovery that in those few
butterflies with eyespots in the anterior wing compartment,engrailedandciare
also expressed in the anterior compartment at the eyespot foci (but not elsewhere
in the anterior compartment). “Thus, the expression of theHedgehogsignaling
pathway andengrailedis associated with the development of all eyespot foci and
has become independent of the [anterior/posterior] restrictions [that are found in
Drosophila]” [p. 534]
Further experiments and comparative analysis enabled Carroll and co-workers to
elucidate the causal order of the changes in theHedgehogpathway as it shifts from