Evolutionary biology

The problem of variation

Article metrics

One genetic source of the sex-specific variation in pigmentation patterns of different fruitfly species has been identified. This study illustrates the power of bringing together developmental and evolutionary biology.

Put simply, evolution is the result of natural selection promoting or eliminating particular heritable 'phenotypic' differences in a population, such as a behaviour or a physical characteristic. Over the past 150 years, many aspects of this process have been explored. But one factor in the equation has remained enigmatic — the source of the heritable phenotypic variation itself.

On page 553of this issue1, Kopp and colleagues illustrate how developmental genetics can help to solve this problem. Their first finding is interesting enough: they have identified a key genetic regulator of sex-specific differences in abdominal pigmentation in the fruitfly Drosophila melanogaster. Even more impressive, they show that changes in the regulation of this gene may have contributed to the variety of pigmentation patterns seen among different fruitfly species. Finally, they investigate how these findings relate to mating behaviour.

Nowadays it is thought that mutations in DNA sequences cause changes in phenotype. But it has proved difficult to identify evolutionarily relevant mutations — those that alter phenotypes and persist in natural populations. There are three core difficulties. First, the DNA regions that encode genes contain vast quantities of so-called neutral variations, which have no effect on either the phenotype or the fitness — the ability to survive and reproduce — of an organism.

Second, many evolutionarily relevant mutations may occur outside the protein-coding regions of genes. Genes can be divided roughly into the DNA sequence that codes for RNA, which may in turn be translated into protein, and regulatory regions, which encode instructions for regulating gene expression. These instructions are deciphered by DNA-binding proteins called transcription factors, which recognize specific DNA motifs and enhance or repress transcription. In doing so, they switch genes on or off in specific spatial and temporal patterns throughout development2. Although regulatory regions have a critical role in development, we still have a poor understanding of how their structure relates to their function. The third difficulty is that even dramatic evolutionary alterations in the DNA of regulatory regions can result in no change in function3,4.

The upshot of all this is that many evolutionarily relevant changes in DNA sequences are probably buried within vast quantities of neutral variation, within both protein-coding sequences and poorly understood regulatory regions. Another obstacle to identifying evolutionarily relevant mutations is the paucity of experimental systems that allow physical variations to be related to DNA variations.

The new discipline of evolutionary developmental biology is ideally poised to tackle these problems, by applying the experimental approaches of developmental biology to the study of variation between closely related species. Developmental biologists typically work out the normal function of genes by examining the physical consequences of experimentally induced mutations within those genes. Evolutionary geneticists, by comparison, focus on the natural genetic variation found within and between species. Natural genetic variation normally causes more subtle phenotypic differences than the mutations studied in developmental biology. But it is becoming clear that researchers in these two fields have been studying different kinds of variation at the same genes5,6,7,8.

The work of Kopp et al.1 illustrates how developmental biology can help to identify the source of heritable phenotypic variations. The authors have used a model experimental organism — D. melanogaster — to gain a detailed genetic understanding of a phenotypic trait that varies between fruitfly species, and have determined whether variations in this gene between species correlate with phenotypic variations.

Kopp et al. first show that differential regulation of the bric-a-brac (bab) gene, which codes for a putative transcription factor, is required for the development of the sex- specific abdominal pigmentation patterns in D. melanogaster. In females, bab is expressed in all abdominal segments, repressing pigmentation. In males, bab is not expressed in the most posterior segments, leading to pigmented posterior abdomens (see Fig. 1 on page 553). The authors discovered that the bab gene integrates inputs from the sex-determination pathway, which limits repression of bab to males, and from the Hox genes, which limit bab repression to the posterior abdominal segments.

The authors then show that this sexually 'dimorphic' regulation of bab sometimes correlates with sex-specific pigmentation patterns in other Drosophila species. The implication is that the evolution of the regulation of bab was important in the evolution of this sexually dimorphic feature. However, in some species of the montium subgroup, bab is repressed in the posterior abdominal segments of males even though the species do not have sex-specific pigmentation. So, the genetic circuitry linking bab to pigmentation has also evolved, presumably as a result of alterations in the regulatory regions of genes downstream of bab. This link was probably either lost in the montium subgroup, or gained in the other subgroups.

Kopp et al. also examined other sexually dimorphic characteristics in D. melanogaster, such as the shape of abdominal segments and the patterns of bristles and of small projections of the cuticle called trichomes. They find that sexual dimorphism in these features also depends on the regulation of bab. In addition, in those species most closely related to D. melanogaster, male-specific expression of these characters correlates with the repression of bab. In contrast, species of the ananassae and montium subgroups have diverse bristle and trichome patterns that vary independently of bab expression patterns. In these lineages, there may have been evolutionary changes in unidentified genes that in D. melanogaster are regulated by bab, the sex-determination pathway and the Hox genes.

Finally, the authors have investigated the role of pigmentation patterns in sexual behaviour. They find that D. melanogaster females do not prefer males with a normal pigmentation pattern over males with a mutant, unpigmented posterior abdomen. Males, however, do discriminate against females with a 'male' pigmentation pattern. So males — but not females — may use pigmentation cues to discriminate between the sexes. Kopp et al .'s use of mutations within D. melanogaster to study sexual behaviour is innovative, and potentially links selection acting on mating systems to the molecular mechanisms underlying phenotypic evolution. But it is not yet clear whether these results are relevant to sexual behaviour in flies with natural variations in pigmentation.

This difference between the variation generated in the lab and that seen in natural populations illustrates one historical difference between developmental and evolutionary biology9. But Kopp et al.'s work1 shows how these two fields can be brought closer together. No doubt more progress will come from further recognition that both fields are interested in studying the same genes, and from the identification of the specific genetic changes that cause natural phenotypic variations.


  1. 1

    Kopp, A., Duncan, I. & Carroll, S. B. Nature 408, 553– 559 (2000).

  2. 2

    Arnone, M. I. & Davidson, E. H. Development 124, 1851–1864 (1997).

  3. 3

    Ludwig, M. Z., Bergman, C., Patel, N. H. & Kreitman, M. Nature 403, 564–567 ( 2000).

  4. 4

    Takahashi, H., Mitani, Y., Satoh, G. & Satoh, N. Development 126, 3725–3734 (1999).

  5. 5

    Gibson, G. & Hogness, D. S. Science 271, 200–203 (1996).

  6. 6

    Stern, D. L. Nature 396, 463–466 ( 1998).

  7. 7

    Long, A. T., Lyman, R. F., Morgan, A. H., Langley, C. H. & Mackay, T. F. C. Genetics 154, 1255–1269 (2000).

  8. 8

    Sucena, E. & Stern, D. L. Proc. Natl Acad. Sci. USA 97, 4530–4534 (2000).

  9. 9

    Stern, D. L. Evolution 54, 1079–1091 (2000).

Download references

Author information

Correspondence to David L. Stern.

Rights and permissions

Reprints and Permissions

About this article

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.