20.4 N&v 999 Mh

1002 several occasions as well. For instance, the unadorned wings of the most famous species in the genus, D. melanogaster, derive from a distant spotted ancestor. Prud’homme et al. took advantage of this chequered history to study the genetic basis for convergence. To validate their approach, they examined two spotted species that derive from the same spotted ancestor and whose spots would therefore be expected to share the same genetic basis. Their previous work with one of these species, D. biarmipes, identified mutations in a part of the regulatory region of the yellow gene (called the spot element) as responsible for one origin of the wing spot. This gene encodes an enzyme involved in pigment synthesis, and mutations affecting its expression have contributed to the evolution of certain pigmentation patterns in other fruitfly species. During wing development in D. biarmipes and D. elegans, the yellow gene is expressed specifically in a region that prefigures the black spot of the adult wing, implying that there is probably a similar underlying genetic basis for the spots in these species. So much for similarity by common descent; what about convergent similarity? The repeated spot losses and gains turn out to have a surprisingly similar genetic basis. In two independent cases of spot loss, represented by D. gunungcola and D. mimetica, different sets of mutations in the spot element of the yellow regulatory region abolish expression of yellow specifically in the spot region of the wing. The spot element comprises just a few hundred of the roughly 180 million base pairs of the fruitfly genome, making this a ‘similar’ genetic basis by any reasonable criterion. Even more striking is the case of D. tristis, a species that underwent an independent, convergent acquisition of the wing spot. Here, the spot regulatory element does not harbour the mutations responsible for spot production in this species. Instead, Prud’homme et al. discovered a different regulatory element that activates yellow expression in the spot region of the wing. The convergently similar wing spots of D. biarmipes and D. tristis are therefore the product of mutations in the same gene, but involve co-option of different regulatory elements. These results hint at generalities in the genetic basis for anatomical evolution. The spot gains and losses all result from mutations that affect the expression of the gene, reinforcing the notion that such regulatory mutations constitute a major component of the genetic basis for anatomical evolution. The fact that the resulting anatomy is so similar conveys a more subtle and interesting message. The developing wings of fruitflies are patterned by signalling molecules that activate the expression of transcription factors. These proteins in turn activate the spatially restricted expression of genes such as yellow, which encode proteins that convert the pattern into anatomical structures. The spatial scaffold for producing a spot is therefore present even in species lacking spots. If, through random mutation, a gene acquires a new binding site for a transcription factor in its regulatory region, its expression is likely to change in ways that reflect the existing spatial scaffold. This explains similarities in the size, shape and position of independently evolved spots. But why was yellow involved each time, given that many fruitfly genes affect pigmentation? The authors argue that a gene already expressed in the wing (such as yellow) is much more likely to be recruited to produce a new wing pattern through a small number of mutations. This is because it already possesses regulatory elements that interact with some of the transcription factors required to produce the new pattern. Of course, none of this explains why mutations that generate or erase spots become established within fruitfly populations. The reason for that seems to be sex. Male fruitflies woo potential mates using elaborate courtship dances that involve a good deal of wing waggling (Fig. 1d). Accordingly, shifts in mate-choice preferences among females could drive rapid changes in wing patterns. Perhaps because it is already expressed in wings, yellow has repeatedly provided genetic variation that produces anatomical differences, making it central to the diversification of fruitfly coloration. !