Pattern Formation: A Focus on Notchin Butterfly Eyespots

Vernon French1 and Paul M. Brakefield2

New observations of early and dynamic expressionof Notch in developing lepidopteran wings suggeststhat this signalling pathway may function in definingthe central focus that will specify the butterflyeyespot colour pattern.

Most animals, from flatworms to primates, developcharacteristic body colour patterns, as their surfacecells occupying different positions synthesise oraccumulate different pigments. Among insects, theLepidoptera (moths and butterflies) display the mostspectacular array of colour patterns on their wings [1],and we are now beginning to understand developmentof one striking element of these patterns— theconcentric eyespot [2,3]. Crucial to eyespot formationare the cells at the future centre, or focus, whichsignal outwards to specify the surrounding colourrings. There has been little information about how thefocus is formed, but a study published recently inCurrent Biology [4] indicates that the Notch signallingpathway has a role in establishing the focus and thusin initiating development of the eyespot.

In metamorphosing insects like butterflies, thewing originates as a flap of epidermis, the wing disc,which grows inside the larva and then everts to formthe immobile pupal wing and, finally, the differenti-ated, functional wing of the adult. The pattern ofwing veins starts to appear in the last (fifth) larvalstage, as tracheae extend out along the spacedhaemolymph channels, or lacunae, that separatedorsal and ventral layers of the wing disc. Colourcomes much later, towards the end of the pupalstage, when pigment is deposited in the cuticle ofthe wing scales and eyespot patterns appear,centred midway between veins on the distal part ofthe wing blade [1].

The eyespots, however, form in response toepidermal cell interactions that occurred much earlierin development. Surgical experiments, pioneered byNijhout [5], showed that the eyespot rings arespecified on early pupal wings by signals coming fromthe central focus. Thus, removal of the focal cellseliminates the eyespot, while grafting them elsewhereon the wing epidermis produces a displaced eyespotpattern. These results imply that the focus must havebeen specified earlier, by cell interactions in the larvaldisc and, although surgery has not proved informative,studies of gene expression are giving some cluesabout how the eyespot focus is formed.

Many genes have known functions in developmentof the (colourless) Drosophila wing, and orthologs ofsome of these are expressed in butterfly wing discs asin fly discs, but with extra features seemingly relatedto formation of colour pattern [6–8]. Hence, in mid fifthlarval stage, Distal-less (Dll) is expressed in a broaddistal band, as in Drosophila, but with rays extendingin along the midlines of wing-cells — the regionsbetween successive lacunae. These rays either fadeor, in wing-cells that will bear eyespots, resolve intosmall focal spots [7]. Subsequently, in the early pupalwing epidermis, Dll and genes encoding other tran-scription factors become expressed in concentriczones that prefigure the eyespot colour rings [8].

Spatial patterns are specified by cell interactions,however, so signalling pathways must be involved.Reed and Serfas [4] have now studied Notchexpression in the larval discs of three eyespot-bearingspecies of nymphalid butterflies. They found thatNotch expression in these species is up-regulated in adynamic pattern — with midline rays which resolveinto focal spots — that resembles, but distinctly pre-cedes, expression of Dll (Figure 1). The focal expres-sion of Notch probably also precedes that ofhedgehog (hh) in the adjacent cells [4,9].

Notch encodes the transmembrane receptor for theconserved, short-range Notch signalling pathway.Binding of an external signal, such as Delta, activatesthe Notch receptor and ultimately changes gene tran-scription in the responding cell. One consequence canbe the up-regulation of Notch (and repression of Delta)transcription, forming a positive feedback loop [10]. Inthis way, the Notch pathway operates as a lateralinhibition mechanism in the segregation of individualneural cells from their neighbours. In the Drosophilawing, the Notch pathway acts in the formation of linearboundaries, between dorsal and ventral surfaces and,later, along the differentiating veins [10]. In butterflywings, the distribution of signals is still unknown, butthe pattern of Notch up-regulation does suggest thatthis signalling pathway may be involved in defining themidline between lacunae and in positioning stable Dllexpression along it, establishing the focus [4].

Long before gene expression was examined inbutterflies, Nijhout [1] suggested that an eyespotfocus could become established at a midline locationwithin the wing-cell, by the operation of a lateralinhibition/reaction–diffusion patterning system [11].Briefly, if haemolymph in the proximal and laterallacunae of a wing-cell provides an initial source of thereactants, interactions within the epidermis can causeconcentrations to build up along the distal part of themidline and, with appropriate parameter values, tostabilise as a discrete central peak which could thenspecify the focus [1]. While the match between thismodel and the observed dynamic expression patternsis not exact — for example, expression is initiallyalong the distal, but not the proximal or lateral, lacuna

Dispatch

Current Biology, Vol. 14, R663–R665, August 24, 2004, ©2004 Elsevier Ltd. All rights reserved. DOI 10.1016/j.cub.2004.08.013

1School of Biological Sciences, University of Edinburgh,Edinburgh EH9 3JT, UK. 2Institute of Biology, LeidenUniversity, 2300 RA Leiden, The Netherlands. E-mail: vernon.french@ed.ac.uk

— it is very intriguing, for Dll [12] and, particularly, forNotch, which is known to mediate lateral inhibition inother contexts [10].

Strictly, the analysis of gene expression patternscan provide only circumstantial evidence —consistent with, but not proof of, gene function ineyespot specification. In the case of Dll, further strongcircumstantial evidence comes from the demonstra-tion [13] that allelic variation at the Dll locus isassociated with differences in eyespot size producedby artificial selection. The analysis of developmentalmechanisms, however, requires ways of experimen-tally manipulating gene expression and, so far,attempts ectopically to express candidate genes inlarval wing discs by using viral vectors, or to abrogateexpression with injected oligonucleotides, have notbeen successful. The most promising route to estab-lishing the function of Notch (and other genes) maynow be to use germline transformation to manipulateexpression during eyespot development [14].

Lepidopteran wings are distinctive in their richdiversity of colour patterns. Nonetheless, a longtradition of comparative analysis has proposed acommon ‘groundplan’ of basic pattern elements, suchas eyespots and central transverse bands, which maybe variously absent, distorted or merged on the wingsof different species [1]. Furthermore, some specieshave rows of eyespots fused together, or a bandbroken into eyespot-like swirls, leading Nijhout [1] tosuggest that even the different elements develop via acommon focal signalling mechanism that is used atmany different locations on the wing. Once evolvedwithin the Lepidoptera, this mechanism may then havebeen modified in only rather minor ways to generatethe great range of patterns [1,15]. So far, however,there has been little direct support for this attractiveunitary view. Thus the discrete spots of Notch [4] and

Dll [7] expression correspond to centres of eyespotsbut, apparently, not of the transverse bands. The focifor other elements might function much earlier — and,of course, only a few genes of unproven significancehave been studied — but it is also possible that theeyespot was an evolutionary novelty, developing by amechanism unrelated to that of other patternelements.

The eyespot is a frequent pattern element onlywithin the nymphalid butterflies. Reed and Serfas [4]also examined gene expression in lepidopteranspecies coming from other groups and lacking eye-spots. They observed that, in a representative of thepierids — the ‘whites’ — Notch and Dll are expressedin persisting midline patterns, but not in discrete foci,whereas midline expression is absent in two mothspecies [4]. This may indicate that the mechanism ofresolving expression into foci evolved afternymphalids diverged from other families, while the useof Notch signalling to define the midline of the wing-cell may have originated earlier, but after the split ofthe butterflies from moths [4]. These suggestions arenecessarily very tentative: few species have beenstudied and, after all, midline expression does notform foci in wing-cells lacking eyespots — even in thenymphalids! Furthermore, striking eyespot patterns doadorn the wings of some species within other butterflyfamilies — such as the papilionids, the ‘swallowtails’— and even of some hawkmoths and saturnid moths,and their development has not yet been examined [4].

The Lepidoptera have evolved a glorious range ofwing colour patterns. Understanding development ofthe eyespot, and of the other pattern elements, willrequire information on the expression of many moregenes, coupled with the results of manipulating thatexpression. Deciphering the evolution of the devel-opmental mechanism(s) will necessitate a much

DispatchR664

Figure 1. Gene expression in butterflyeyespot patterning.

(A) The ventral wing colour pattern of thenymphalid Bicyclus anynana, consistingmainly of transverse bands and eyespotscentred between wing veins. (B–D) Thesediagrams each show part of fifth larvalinstar wing discs, roughly correspondingto the boxed area in (A). The discs are ofincreasing age, staged by the extent towhich tracheae have penetrated thelacunae [16]. Notch (N, yellow), and thenDistal-less (Dll, red dots), protein is up-regulated in midline rays that eitherresolve into stable spots at eyespot foci(lower wing-cell) or fade [4]. At the lateststage (D), Notch is also expressed alongthe lacunae.

Current Biology

A B

C

D

Lacuna

Midline ray

N

Focus

Focus

N, Dll

broader phylogenetic approach. Meanwhile, the newwork of Reed and Serfas [4] represents real progresson both fronts.

References1. Nijhout, H.F. (1991). The Development and Evolution of Butterfly

Wing Patterns. (Washington: Smithsonian Institution Press.)2. McMillan, W.O., Monteiro, A., and Kapan, D.D. (2002). Development

and evolution on the wing. Trends Ecol. Evol. 17, 125-133.3. Beldade, P., and Brakefield, P.M. (2002). The genetics and evo-

devo of butterfly wing patterns. Nat. Rev. Genet. 3, 442-452.4. Reed, R.D., and Serfas, M.S. (2004). Butterfly wing pattern evolution

is associated with changes in a Notch/Distal-less temporal patternformation process. Curr. Biol. 14, 1159-1166.

5. Nijhout, H.F. (1980). Pattern formation on lepidopteran wings: deter-mination of an eyespot. Dev. Biol. 80, 267-274.

6. Carroll, S.B., Gates, J., Keys, D.N., Paddock, S.W., Panganiban,G.E.F., Selegue, J.E., and Williams, J.A. (1994). Pattern formationand eyespot determination in butterfly wings. Science 265, 109-114.

7. Brakefield, P.M., Gates, J., Keys, D., Kesbeke, F., Wijngaarden, P.J.,Monteiro, A., French, V., and Carroll, S.B. (1996). Development,plasticity and evolution of butterfly eyespot patterns. Nature 384,236-242.

8. Brunetti, C.R., Selegue, J.E., Monteiro, A., French, V., Brakefield,P.M., and Carroll, S.B. (2001). The generation and diversification ofbutterfly eyespot patterns. Curr. Biol. 11, 1578-1585.

9. Keys, D.N., Lewis, D.L., Selegue, J.E., Pearson, B.J., Goodrich, L.V.,Johnson, R.J., Gates, J., Scott, M.P., and Carroll, S.B. (1999).Recruitment of a hedgehog regulatory circuit in butterfly eyespotevolution. Science 283, 532-534.

10. Artavanis-Tsakonas, S., Rand, M.D., and Lake, R.J. (1999). Notchsignalling: cell fate control and signal integration in development.Science 284, 770-776.

11. Meinhardt, H. (1982). Models of Biological Pattern Formation. (NewYork: Academic Press)

12. Nijhout, H.F. (1994). Genes on the wing. Science 265, 44-45.13. Beldade, P., Brakefield, P.M., and Long, A.D. (2002). Contribution of

Distal-less to quantitative variation in butterfly eyespots. Nature415, 315-318.

14. Marcus, J.M., Ramos, D.M., and Monteiro, A. (2004). Germlinetransformation of the butterfly Bicyclus anynana. Proc. R. Soc.Lond. B (Biol. Letters, Suppl) S1–S3.

15. Nijhout, H.F. (2001). Elements of butterfly wing patterns. J. Exp.Zool. 291, 213-225.

16. Miner, A.L., Rosenberg, A.J., and Nijhout, H.F. (2000). Control ofgrowth and differentiation of the wing imaginal disk of Precis coenia(Lepidoptera; Nymphalidae). J. Insect Physiol. 46, 251-258.

Current BiologyR665