Much in genetics, development, and neurobiology has been learned from studying a small winged invertebrate, the fruit fly Drosophila melanogaster. Because of its general significance as molecular genetic model system, it is often overlooked that Drosophila is a highly specialized insect. Its maggot like larva is headless. As a consequence, the large compound eyes of the adult fly are not formed during development of the embryo, but inside the larva. This distinguishes Drosophila greatly from primitive insects and most animals in general, where eye development begins in the embryo. To elucidate the sequence of events that were responsible for the origin of the unique form of Drosophila visual development, we investigate the phylogenetic history as well as the developmental evolution of the Drosophila visual system integrating molecular phylogenetics, comparative development and genomic approaches. Our animal models include the red flour beetle Tribolium castaneum and the small carrion cave beetle Ptomaphagus hirtus.
Cave adaptation in a beetle
Caves represent one of the most extreme habitats on earth. And yet, hundreds of species succeeded in adapting to these light-deprived and nutrient-poor biota. Already Darwin discussed whether the reductive evolution of visual organs is of adaptive significance or the consequence of neutral degeneration along the lines of ‘use it or lose it’. One school of thought explains the loss of eyes in cave animals as the consequence of the relinquishing selection. Another argues that, since photoreceptors are exceptionally energy consuming, the loss of visual organs may be driven by positive selection because cave animals face the challenge of coping with the nutrient scarcity of the cave environment (for review see Culver and Wilkens, 2000).
We have begun to study cave adaptation in the fungus beetle Ptomaphagus hirtus. Surface fungus beetles possess pronounced compound eyes, which are reduced to residual lens patches in Ptomaphagus hirtus. An early histological study failed to detect photoreceptors in P. hirtus, which is therefore considered blind (Packard, 1888). This conclusion is challenged by the more recent discovery that the specification lens cells in insects is dependent on inductive signals from differentiating photoreceptor cells (Wolff and Ready, 1993).
Recent momentum to our work with P. hirtus has come by winning the $500 award of the Groundwater and Caves Challenge Grant competition with the project "Exploring the temperature tolerance of a cave beetle" through the generous support of 109 backers and the kind endorsement by highly accomplished peers in the fields of cave biology and vision research:
Functional genomic analysis of visual system development in Tribolium
The early retinal genes dachshund (dac), eyes absent (eya) and sine oculis (so) are key regulators of adult eye development in Drosophila. Expression data implicate homologs of all three transcription factor genes in vertebrate eye development (for review see Hanson, 2001). However, functional confirmation has thus far only been reported for homologs of so (Donner and Maas, 2004). We therefore investigated expression and function of these genes in the red flour beetle Tribolium castaneum (Yang et al., 2009a). Our results show that Tribolium so and eya are essential for both larval and adult eye development. dac knockdown Tribolium exhibit severe but incomplete adult eye reduction, irregularities in ommatidial photoreceptor numbers, and pigment defects in peripheral ommatidia. In a parallel study, we further discovered that dac is essential for adult eye development in combination with the Pax6 transcription factors eyeless (ey) and twin of eyeless (toy) (Yang et al., 2009a).
ey and toy are upstream regulators in the gene regulatory network, which instructs the formation of the adult eye primordium in Drosophila. Most animals possess a singleton Pax6 ortholog, butthe dependence of eye development on Pax6 is widely conserved. A rare exception is given by the larval eyes of Drosophila, which develop independently of ey and toy. To obtain insight into the evolution of differential larval and adult eye regulation, we have also studied the function of toy and ey in the red flour beetle Tribolium castaneum (Yang et al., 2009b). Larval eye development is highly sensitive to single and combinatorial knockdown of toy and ey in Tribolium, while adult eye development is only mildly affected. Adult eye-loss was only provoked when ey and toy were RNAi silenced in combination with the early retinal gene dac. The results lead to a model, which differs from the largely epistatic interactions between toy, ey and dac in Drosophila (Pappu and Mardon, 2004). That is, the specification of the adult eye primordium occurs in the embryo under partially redundant control by ey and toy. The subsequent maintenance of eye primordium commitment during larval development is in part specifically and in part redundantly regulated by dac, toy and ey (Yang et al., 2009b).
We are currently testing these ideas and the function of further candidate eye genes in Tribolium in collaboration with Dr. Rui Chen at the Human Genome Sequencing Center of Baylor Colllge in Houston. This project is funded by NSF grant 0951886 ‘Pax6 and the genetic regulation of eye development in Tribolium’.
Gene duplication and the developmental evolution of Drosophila
In the course of the Tribolium genome annotation project, we discovered evidence of an exceptional gain of visual system related genes in the lineage leading to Drosophila compared to other insect genome model species (Tribolium, Apis) and major dipteran subgroups (mosquitoes) (Bao and Friedrich, 2009). Surveying published genetic data on these duplicated genes we were able to conclude that gene duplication very likely played an adaptive role during the evolution of visual performance in the fast flying higher Diptera. Developmental gene duplications, by contrast, predominantly preserved partial to complete redundancy consistent with a role in boosting developmental robustness.
Our work in the FlyTree consortium has provided us with tissue samples and expertise to study the dynamics of gene duplicate accumulation in the higher Diptera in more detail. The first question that we have been pursuing is whether the increase in visual gene duplications reflects a genome-wide or a gene function phenomenon. To this end, we extended our gene-by gene analysis to 399 developmental genes. This work revealed a two-fold higher number of unique gene duplications in the lineage to Drosophila compared to other lineages. Using this data matrix as gold standard, we developed a computer program that identifies lineage specific gene duplications with high accuracy. The preliminary results from this new program indicate that the lineage to Drosophila experienced a genome-wide approximately 2-fold increase in gene duplication rate compared to most other insect lineages.
We are now pursuing the question whether there is a correlation between gene duplication and body plan evolution in the higher Diptera, and if so, whether this correlation resulted from causal impact of gene duplication on body plan evolution or from independent factors that impacted both processes. Previous large-scale studies of gene duplication focused on scenarios that resulted from whole genome duplication. Our findings in Drosophila and related Diptera establish a resource-rich system to study of the impact of steady rate background gene duplication on development.
In parallel, we plan to take advantage of the rich repertoire of gene duplications in the Diptera by exploiting the phylogenetic information content gene duplications in the form of rare genomic event traits.