People Publications Links Teaching Pictures Protocols Movies Art

Research summary

We seek to understand the evolution of morphological novelties by focusing on the evolution and development of butterfly wing patterns. Research in the lab addresses the selective factors that favor particular wing patterns as well as the mechanisms that generate those patterns. We combine tools from ethology, population genetics, phylogenetics, and developmental biology to understand the nature of the variation underlying developmental mechanisms within and between species, and why species display their particular color patterns. Our model organisms (so far) have been African satyrid butterflies in the genus Bicyclus, other nymphalids, pierid butterflies, and saturniid moths.

Prospective students

If you are interested in pursuing a graduate project (Marster's or PhD project in the lab) you will need to apply to the graduate program at the Department of Biological Sciences at the National University of Singapore. You can also learn about PhD scholarships via Yale-NUS College (here).

If you are an undergraduate, there are several ways that you can participate in research in the lab. Either by doing an honors project (two semester's worth of research), an independent project (one semester), or a summer project (three months).

Projects in the lab

Identifying the eyespot gene regulatory network:

In order to understand how novel traits and their underlying gene regulatory networks originate we should begin by describing these networks. An obvious place to start is to identify the genes that are differentially expressed in tissues where the novel trait develops versus homologous tissues that don't develop the trait. This is what we are currently doing for the eyespot gene regulatory network using RNA-seq.

The developmental origin of eyespots

Jeff Oliver's work (Oliver et al. 2012, 2014) indicated that eyespots probably originated with the co-option of a pre-existing network of interacing genes into the middle of wing compartments. The question now is: which network was co-opted? We are currently testing out the model proposed in Monteiro and Podlaha (2009), and Monteiro (2012) to discover the identity of the ancestral network that gave rise to eyespots using FAIRE-Seq and CRISPR tools applied to candidate cis-regulatory regions of eyespot genes.

Investigating the origin of eyespot plasticity:

iWe currently know something about the physiological and molecular mechanisms that lead to phenotypic plasticity in eyespot size and eyespot center brightness in response to rearing temperature (Monteiro et al. 2015). The next step is to perform comparative work across nymphalid species to investigate how these mechanisms evolve. Students/postdocs interested in this project are also welcome to join the lab.

Sex-specific development of butterfly wing patterns

A new research direction aims to understand the evolution of sexual dimorphism in butterfly wing patterns. We have begun by localizing the expression of doublesex, and important gene in the sex determination pathway, in wings of males and females at multiples stages of development.

Testing gene function in color pattern formation:

In order to test whether a gene has a role in the development of butterfly wing patterns it is important to be able to manipulate its expression using transgenic tools. We have successfuly transformed the genome of Bicyclus anynana with Hermes and piggyBac transposable elements. We initially used EGFP and DsRed under the control of different promoters (actin and 3xP3) as marker genes to test germ line transformation. Later we tested whether the heat-shock promoter from Drosophila Hsp70 could be used as an inducible promoter. More recently we have tested the role of several developmental genes in eyespot development by either over-expressing or down-regulating these genes at particular times during development using the Hsp70 promoter. People that did this work were formal post-doctoral researcher Jeffrey Marcus, graduate student Diane Smith, former visiting professor Bin Chen, and myself.

Postdoctoral fellow Andrew Stoehr also produced transgenic Pieris rapae butterflies using piggyBac constructs.

We are currently using CRISPR-Cas9 to disrup genes and cis-regulatory elements in B. anynana and Pieris canidia.

New over-expression and RNAi plasmid construct:

Bin Chen produced a new expression plasmid, Pogostick, that can be used for gene over-expression (or ectopic expression once activated in a new location) or gene knock-down. This plasmid uses piggyBac as the transposable element, and EGFP under the control of 3xP3 as the transformation marker. Candidate genes, inserted once, can be over-expressed. When inserted twice, in opposite orientation, can start the process of RNA interference (RNAi) inside the cells. We tested this system in our recent work on Dll and Ubx ectopic expression in B. anynana (Monteiro et al. 2003; Tong et al. 2014). We are currently using a Wg-RNAi construct to study the role of Wg in eyespot development.


Firdous Kamal developed a new laser-mediated heat-shocking mechanism using a IR laser that enables us to heat-shock specific areas of a transgenic butterfly in order to induce ectopic expression of candidade wing patterning genes. Bin Chen and Diane Smith cloned a series of candidate genes and introduced them in the germ line of B. anynana. These transgenes were under the control of the heat-shock promoter. These experiments allowed us to test the role of these candidate genes in the differentiation of colored wing scales. For instance, ectopically expressing the Distal-less transcription factor on the pupal wing was sufficient to induce black scale development, showing that Distal-less is regulating the melanin pigment biochemical pathway.

The ability to ectopically express genes on the wing, combined with tissue dissections, and powerful transcriptomic approaches, may allow us to test how individual genes connect to downstream targets over the course of trait development.

Reconstructing and animating ancestral wing patterns:

Undergraduate student, Sam Arbesman, worked on a web based animation (Ancient Wings) that reconstructs the putative ancestral ventral hindwing patterns of 54 of the 80 species of Bicyclus butterflies, and morphs these patterns across the phylogenetic tree of Bicyclus.

Jeff Oliver also reconstructed presense and absence of eyespots across the Bicyclus phylogeny and estimated rates of eyespot evolution separately for males and females. He found that eyespots on the dorsal surface or forewing surfaces were more likely evolving via sexual selection because rates of evolution were high and different between the sexes. Eyespots on the exposed ventral or hindwing surfaces were evolving slowly and at similar rates across the sexes. Jeff's subsequent projects extended this work to use wing pattern data to all nymphalid butterflies (using the extensive collection of Lepidoptera housed at the Peabody museum). Essentially we tested whether eyespots evolved once or multiple times within the nympahlidae, and whether eyespot number evolved by the addition of eyespots to more and more wing compartments throughout evolution (via network co-option), or whether eyespots appeared in all wing compartments originally and were later lost from particular wing compartments? Jeff's data showed that eyespots evolved once, roughly 85-90 MY ago, in the lineage sister to the Danainae. Eyespots also originated as a cluster of 4 or 5 units on the ventral hindwing. Eyespots later appeared on the forewing and dorsal wing surfaces.

The genetics of eyespot number

We are currently following up on this work by identifying the genetic locus that may be regulating variation in eyespot number. We are using RAD-Tags and a natural eyespot number mutant to map the locus in the genome of B. anynana.

Testing the role of sexual selection in maintaining species (and season) specific wing patterns:

Graduate students Kendra Roberston and Katie Costanzo discovered that wet season female B. anynana choose males on the basis of their UV-reflective dorsal eyespot pupils as well as their pheromones. When one or both of these two traits are blocked or removed, females don't like to mate with these males. Postdoctoral researcher Katy Prudic discovered that dry season males also care about a female's UV-reflective dorsal eyespot pupils. Females display their dorsal eyespots to males during the dry season, but not during the wet season. Males do the opposite. Females are the choosy sex in the wet season and males are the choosy sex during the dry season. This phenotypically plastic courtship reversal is controlled by larval rearing temperature. Females court males in the dry season because the spermatophore they receive from these males allows them to live almost twice as long as females that don't mate or mate with a wet season male. It is still unclear why males are reticent to mate with dry season females.

Testing the role of natural selection in maintaining species (and season) specific wing patterns:

Katy Prudic tested the role of mantid predators is explaining the plasticity in eyespot size on the ventral surface of the wings of B. anynana in response to the wet and dry seasons. Vertebrate predators such as birds and lizards appear to have a more difficult time detecting the dry season form relative to the wet season form in cage experiments in the lab. But, once they detect the butterflies, attacks are usually directed towards the wing margin in both forms. So, it is unclear why ventral eyespots become large in the wet season form, given that in interactions with these predators they don't appear to have any advantage. Katy showed that large eyespot size may have evolved to deflect attacks from insect predators such as mantids that have a different visual system (cannot detect UV). These predators attack the wing margin of butterfies with large contrasting rings of color, but attack more vulnerable parts of the insect if eyespots are small. We are currently following up on this work with additional experiments with mantids in the lan and natural predators in the field.