Development and evolution of the eye: Pax-6 and the compound eye of Drosophila melanogaster

It is no longer nature's secret that many proteins involved in basic mechanisms of development are highly conserved between organisms as different as humans and flies. Nonetheless, the discovery that the vertebrate transcription factor Pax-6 and the Drosophila Pax-6 homolog eyeless both play crucial roles in the developing eyes of mammals (1-3) and flies (4) came as a big surprise. Why? In contrast to the established view of eye evolution, these findings raise the intriguing possibility that the vertebrate and fly eye share a common evolutionary history that included regulation by Pax-6. Traditionally, the vertebrate and fly eye have been considered to be independent evolutionary innovations for several reasons (5). Most conspicuously, the architectures of a fly compound eye and of a vertebrate single-lens eye are so strikingly different that no structural homology can be claimed. Even the ultrastructure of their photoreceptor cells appears to be fundamentally different. Consequently, it was unexpected to find the same transcription factor regulating eye development in both groups of animals. However, compatible with an independent evolutionary origin of eyes, the Pax-6 discovery could also be explained by an unrelated recruitment of Pax-6/eyeless into separately evolving eye developmental pathways, i.e. by convergent evolution. To distinguish between these alternatives we must understand and compare the function of Pax-6/eyeless during eye development in diverse animals.

My thesis centered on the study of eyeless function during Drosophila compound eye development. Remarkably, eyeless is not only necessary for eye development, but it is able to induce the entire cascade of gene activities sufficient to generate complete and properly formed compound eyes in Drosophila (6). The early expression of eyeless in the eye primordia (4) and the observation that mutations in other genes required during initial steps of eye development do not affect the expression of eyeless (6) indicated an early role for this gene. Loss-of-function mutations in eyeless produce flies with reduced eyes or no eyes at all (4, 7). In these mutants, the eye primordia degenerate early in development (6), consistent with an early and determinative function for eyeless.

The potential of eyeless to act as a developmental switch was tested by targeted expression of eyeless in tissues that do not normally express it (6). To do so, a transposon carrying the eyeless gene under the control of a transcriptional enhancer that stimulates expression in wing, leg, and antennal primordia was introduced into the fly genome. As a consequence, extra eyes developed on wings, legs, and antennae! These eyes consist of the full complement of different cell types normally found in a compound eye, including photoreceptors, pigment cells, cone cells, and bristles. In addition, the arrangement of the different cell types is the same as in a normal eye and the photoreceptors depolarize upon illumination. Evidently, eyeless can switch on the eye developmental program in which several thousand genes may act, thereby directing the formation of an organ with all its complexity.

Is Pax-6 a master regulator of eye development in other animals? The position of Pax-6 in the genetic hierarchy of vertebrate eye development is not yet known, but the early expression in the eye and the loss of eye structures in the human Pax-6 disorder Aniridia and the mouse Small eye mutants are consistent with an early and determinative role (8). Squid possess single-lens eyes very similar in basic design to vertebrate eyes, but again, clear differences between their modes of development have led to the hypothesis that they originated independently in evolution (5). Nonetheless, a Pax-6 homolog was recently found to be expressed in the developing eye of the squid (9)! Moreover, in spite of the divergence of mammals, molluscs, and insects more than 500 million years ago, targeted expression of the mouse and mollusc Pax-6 genes in Drosophila also induces the formation of extra eyes (6, 9). Therefore, the development of different types of eyes appears to be controlled by conserved Pax-6 proteins despite their enormous morphological divergence.

Genes acting downstream of Pax-6 are also conserved. I found that the two regulatory genes sine oculis (10, 11) and eyes absent (12) play critical roles downstream of eyeless during Drosophila compound eye development. Work from other laboratories has shown that homologs of sine oculis and eyes absent are active in the developing mouse eye and that mouse eyes absent acts downstream of Pax-6 (13, 14). In addition to these two genes acting at intermediate levels, genes at the bottom of the genetic cascades that are involved in the terminal differentiation of the eye are also conserved (15). All visual systems analyzed so far, including those of vertebrates and insects, share homologous proteins called opsins that are the protein component of their visual pigments (15). This expands the parallels beyond a single transcription factor and suggests that the development of the different types of eyes is indeed under the control of similar genetic cascades.

What do these findings tell us about the evolution of eyes? The parallels in the eye developmental programs lead us to favor the idea that the common ancestor of all higher animals, including vertebrates and insects, already had a primitive eye and that the development of this ancestral eye was regulated by Pax-6 (16). This eye may have been a simple eyespot consisting of a cluster of photo-sensitive cells with no ability to form an image, a type of organ found in many animal phyla. Once a functional light sensing organ had evolved, nature apparently improved on its optical performance in many different ways, leading to the incredible variety of eyes seen today. During this process, Pax-6/eyeless continued to be used to control the development of the evolving eyes.

More genetic parallels in eye development of animals as distinct as mammals and flies may well be discovered. But it also remains to be seen how the pathways diverged in the course of evolution to produce eyes that are nevertheless remarkably different.

References

1. C. C. T. Ton, et al., Cell 67, 1059 (1991).
2. R. E. Hill, et al., Nature 354, 522 (1991).
3. C. Walther, P. Gruss, Development 113, 1435 (1991).
4. R. Quiring, U. Walldorf, U. Kloter, W. J. Gehring, Science 265, 785 (1994).
5. L. v. Salvini-Plawen, E. Mayr, Evolutionary Biology 10, 207 (1977).
6. G. Halder, P. Callaerts, W. J. Gehring, Science 267, 1788 (1995).
7. M. A. Hoge, American Naturalist 49, 47 (1915).
8. P. Callaerts, G. Halder, W. J. Gehring, Ann. Rev. Neuroscience 20, 483 (1997).
9. S. I. Tomarev, et al., Proc. Natl. Acad. Sci. USA 94, 2421 (1997).
10. B. N. R. Cheyette, et al., Neuron 12, 977 (1994).
11. M. A. Serikaku, J. E. O'Tousa, Genetics 138, 1137 (1994).
12. N. M. Bonini, W. M. Leiserson, S. Benzer, Cell 72, 379 (1993).
13. G. Oliver, et al., Development 121, 4045 (1995b).
14. P.-X. Xu, I. Woo, H. Her, D. R. Beier, R. L. Mass, Development 124, 219 (1997).
15. T. H. Goldsmith, The Quarterly Review of Biology 65, 281 (1990).
16. G. Halder, P. Callaerts, W. J. Gehring, Curr. Op. Gen. & Dev. 5, 602 (1995).

To Advertise   |   Find Products