By Héloïse Dufour, Shigeyuki Koshikawa and Cédric Finet
In this post we will discuss our recent paper entitled “Temporal flexibility of gene regulatory network underlies a novel wing pattern in flies” . We initiated the present project in Sean Carroll’s lab where the pigmentation in drosophilids was used as a model to study evolutionary genetics. When we started back in 2010, we knew that the black dots on Drosophila guttifera wings result from the co-option of the protein Wingless during evolution . However, D. guttifera does represent a particular case: the black dots on the wing are associated with campaniform sensilla, which is not the case for other wing colour patterns in drosophilids. The most vivid examples of wing pigmentation patterns are found among the Hawaiian flies, from simple pigmentation around the veins to complex black and white patterns , but we were aware that the Hawaiian flies are hardly tractable in the lab. In the meantime, we came across the species Samoaia leonensis which was maintained at the National Drosophila Species Stock Center. We decided to investigate the making of the wing colour pattern in this species.
The genus Samoaia: a case study of pigmentation and its evolution
Samoaia is a small genus of seven described species endemic to the Samoan Islands in the central South Pacific. For example, the main model of our study S. leonensis exhibits a beautiful and complex black and white spot pattern on its wings (Figure 1). This species is a perfect case of pattern matching, with white spots not only covering the black wings, but the legs, the head, the abdomen, and the notum as well. [For etymology lovers, it is tempting to see in leonensis a reference to the lion, then to the jaguar and its spotted fur. Well, entomology systematics can be sometimes down-to-earth: the species S. leonensis was simply discovered around Leone on Tutuila Island.]
The entire genus Samoaia is an excellent model to study the evolution of pigmentation, especially the modularity of pigmentation. The species S. attenuata has indeed white spots on the legs only (Figure 1), suggesting that the whole body ‘camouflage’ of S. leonensis might result from the stepwise gain of spots in different organs during the course of evolution.
Key finding nº 1
We showed that the co-option of the transcription factor Engrailed underlies the making of the white wing spots in S. leonensis wings. Whereas the expression of Engrailed is restricted to the posterior wing compartment in early pupal stages, its expression becomes spotty in both anterior and posterior compartments in later stages. This result is a big surprise for Drosophila developmental biologists for whom engrailed is the posterior identity gene by definition. Would it mean that there is a critical point beyond which engrailed is no longer required for AP patterning?
We tested this hypothesis in D. melanogaster. We silenced or overexpressed engrailed specifically in the wing at different time points. We were able to precisely define three different time windows during which (i) disturbing engrailed expression results in severe morphological defects, (ii) disturbing engrailed results in minor vein defects, (iii) disturbing engrailed expression does not lead to visible phenotype. We argue that the co-option of a given gene is possible beyond its corresponding critical point.
Key finding nº 2
We found that the co-option of Engrailed in the S. leonensis wing is partially independent from the other genes of the AP specification network. For instance, hedgehog and engrailed are not co-expressed in the anlagen of the future white spots, whereas these two genes are co-expressed in the posterior wing disc. Would it mean that the interactions between players of the AP specification network are labile over development? Is it a specificity of S. leonensis late development?
Again, we tested this hypothesis in D. melanogaster. In the D. melanogaster wing disc, Engrailed acts in coordination with other players of the anterior-posterior (AP) specification network. For example, Engrailed both activates the expression of Hedgehog and represses Cubitus interruptus in the posterior compartment. We found that the AP network is only partially maintained in late pupal stages in D. melanogaster. For instance, the depletion of engrailed transcripts in the developing late pupal wing no longer causes the expansion of Cubitus interruptus expression into the posterior compartment. Our results suggest that the temporal flexibility of regulatory gene networks might facilitate gene co-option during evolution.
Key finding nº 3
We further investigated the function of several genes of the AP specification network over wing development in D. melanogaster. Except for cubitus interruptus and hedgehog that behave in a similar way, we found that every single gene of the gene regulatory network has its own functional time window. This property might facilitate the co-option of a single gene independently from the whole gene network. We followed the same approach for supplementary genes, and found that wingless and Distal-less have a critical time point particularly early. Wingless and Distal-less have been co-opted to generate wing pigmentation in D. guttifera  and D. biarmipes , respectively. We propose that genes with an early critical point might be more easily co-opted during evolution.
A perfect model but…
The only fly in the ointment is the suitability of Samoaia flies for functional developmental biology. Let us now share a little bit of the unsaid B-side of the project. We injected S. leonensis embryos for two years and finally obtained a single transgenic line that turned to be useless on its own. The members of the Carroll lab wondered how it was possible to accumulate so many impediments. The “curse of Samoa has struck again” became quickly the new leitmotiv in the lab. It may be useful for the community to know a few of the technical challenges. First, the chorion of S. leonensis embryo is very thick and hard. After the first attempts, the injection post was scattered with dozens of broken needles like the French lines after the battle of Waterloo. We therefore dechorionated the embryos by bleaching prior to microinjection, and let them hatch in halocarbon oil to prevent them from desiccation. The larvae looked pretty healthy but 90% of them died within the next few days. After one year of laborious work we finally got transgenics carrying the transgene UAS-engrailed. Believe us, GFP never appeared so beautiful and ecstatic. We were halfway and Christmas was almost there. Such a gift! For us the coming new year meant more and more injections. Again, we encountered numerous issues and got a single transgenic larva carrying the transgene nab-Gal4. The larva died, bringing away our last hope of seeing a pigmentation phenotype in S. leonensis.
Fly wing coloration: future directions
The Samoan Islands and their colorful flies would have been a real change of scene. But the call of the genetics is too strong, we need tractable species… Back to the real world, we wonder what mystery is still unsolved in the pigmentation patterns of flies . Many genes are probably involved in the pattern formation, and many of them still await discovery . We have to clarify how the coordinated regulation of many genes was acquired to make novel pigmentation patterns, and how pre-existing gene regulatory networks were (or not) modified and recruited. In addition to the on/off regulation of gene expression, the transport of signaling molecules, hormones and precursors of pigments could be also key factors for pigmentation evolution. Besides pigmentation, the structural coloration of membranous wings region has been proposed to be play a role in visual communication. Little is known about how the wing structural coloration is formed during development and how it evolved. Fly wing coloration has many chapters, some still being written as time flies…
 Dufour HD, Koshikawa S, Finet C. (2020). Temporal flexibility of gene regulatory network underlies a novel wing pattern in flies. PNAS 117: 11589-11596.
 Werner T*, Koshikawa S*, Williams TM, Carroll SB. (2010). Generation of a novel wing colour pattern by the Wingless morphogen. Nature 464: 1143-1148.
 Edwards KA, Doescher LT, Kaneshiro KY, Yamamoto D. (2007) A database of wing diversity in the Hawaiian Drosophila. PLoS ONE 2(5): e487.
 Arnoult L*, Su KF*, Manoel D*, Minervino C, Magriña J, Gompel N, Prud’homme B. (2013). Emergence and diversification of fly pigmentation through evolution of a gene regulatory module. Science 339: 1423-1426.
 Koshikawa S. (2020). Evolution of wing pigmentation in Drosophila: diversity, physiological regulation, and cis-regulatory evolution. Development, Growth & Differentiation (in press).
 Fukutomi Y, Kondo S, Toyoda A, Shigenobu S, Koshikawa S. (2020). Transcriptome analysis reveals wingless regulates neural development and signaling genes in the region of wing pigmentation of a polka‐dotted fruit fly. The FEBS Journal (in press).