Qualitative And Quantitative Aspects Of Skeletal Evolution In Stickleback Fish

Pigmentation, hairs, and bristles are all characters displayed on the body surface, so perhaps it is not surprising that extensive variation and divergence occurs in these traits on very similar or identical body plans. One might expect that traits such as elements of the vertebrate skeleton would be more constrained and slower to evolve. This notion is shattered by the remarkable evolutionary history of the threespine stickleback fish (Gasterosteus aculeatus) in North America.

This species colonized many newly created lakes and streams at the end of the last ice age. In the course of just 15,000 years, rapid parallel evolution has taken place that has produced similar pairs of differentiated species in numerous locales. In general, in each location benthic forms with reduced body armor, increased body size, and reduced gill rakers (for filtering ingested food) have differentiated from limnetic forms with more extensive body armor, a longer body, and increased number of gill rakers (Fig. 7.8). The two forms are reproductively isolated in the wild, but can be intercrossed in the laboratory to allow for detailed investigation of the genetic architecture of anatomical and behavioral differences between body forms.

Figure 7.8

Evolution of limnetic and benthic forms of the threespine stickleback

Limnetic (top) and benthic (bottom) forms of the threespine stickleback from Priest Lake, British Columbia. Differences in spine length, armor plate number, and gill raker number have evolved as adaptation to different niches. Source: Photographs courtesy of Katie Peichel and David Kingsley.

Figure 7.8

Evolution of limnetic and benthic forms of the threespine stickleback

Limnetic (top) and benthic (bottom) forms of the threespine stickleback from Priest Lake, British Columbia. Differences in spine length, armor plate number, and gill raker number have evolved as adaptation to different niches. Source: Photographs courtesy of Katie Peichel and David Kingsley.

While some of these differences appear to be due to variation at many loci of small effect, a small number of loci were found to account for much of the variance in other characters. Interestingly, the sets of QTLs affecting the length of the first and second dorsal spine were distinct, while those affecting the second dorsal spine and pelvic spine overlapped with one another. These observations reveal that genes with a range of magnitude of effects have contributed to evolutionary change, and that very similar morphological features can be affected differently by the same gene. A key outstanding question presented by the threespine stickleback, in addition to the identity of genes involved in divergence, is whether the same loci are involved in the independent evolution of similar forms in different locales.

MORE VARIATION THAN MEETS THE EYE: CRYPTIC GENETIC VARIATION AND THE POTENTIAL FOR MORPHOLOGICAL EVOLUTION

In addition to the overt, phenotypic variation in the characters described here, developmental geneticists have uncovered more variation lurking in individuals than is generally appreciated ("cryptic variation"). The importance of gene interactions to morphological variation has been underscored by some fascinating studies of the phenotypes that arise when mutations are introduced into individuals with different genetic backgrounds.

While a single mutation is often described as causing a particular phenotypic change in laboratory studies, the effect, in practice, depends on the context of the genetic background in which a mutation is studied. For example, it is well established that in laboratory lines of D. melanogaster, introduction of dominant mutations in the Sevenless tyrosine kinase receptor and EGF receptor (DER) induces a roughening of the surface of the adult eye. This outcome occurs because of a perturbation of developmental events in eye patterning that are dependent on the Sevenless and DER pathways. By contrast, introduction of these mutations into wild-type flies of different origins (and different genetic backgrounds) produces a considerable range of severity in phenotypes. In some genetic backgrounds, the mutant phenotypes are suppressed as compared to their effects in laboratory strains; in others, they are enhanced (Fig. 7.9). The enhancement observed in some backgrounds sometimes exceeded that caused by the combination of Sevenless or DER mutations, and mutations in additional components of the Sev or EGF-R signal transduction pathways. This finding indicates that considerable genetic variation occurs in the wild in loci affecting the function of major pathways.

Similar observations have been made from introducing homeotic mutations into different genetic backgrounds. Both extreme modification and suppression occurs. In some cases, this disparity reflects the presence of single genes of large effect that modify the homeotic pheno-type, but have no stand-alone effect on wild-type development. Such studies demonstrate that there may be widespread genetic variation that does not discernibly affect phenotypes unless certain other interacting mutations are present. Evolutionary and developmental geneticists are only now beginning to truly appreciate this cryptic variation. Importantly, this hidden variation implies that underlying phenotypic stability, the quantitative aspects of genetic regulatory inputs may vary extensively. Furthermore, it suggests that the existing genetic variation available for the selection of new phenotypes during evolution may be significantly greater than previously thought.

One striking illustration of this latter idea has come from artificial selection for homeotic phenotypes in Drosophila. It has long been known that homeotic phenotypes can sometimes

Figure 7.9

Cryptic variation affecting eye development in flies

Figure 7.9

Cryptic variation affecting eye development in flies

(a—c) Scanning electron micrographs of eyes. (d-f) Sections of eyes viewed to reveal the architecture of ommatidia. (a,d) The wildtype D. melanogaster eye has approximately 800 smooth ommatidia that possess one central R7 cell among eight photoreceptor cells. (b,c,e,f) In mutants bearing constitutively active Sevenless proteins, eye and ommatidia development are altered to different degrees, depending upon the genetic background.

Source: Photographs courtesy of Greg Gibson, from Polaczyk PJ, Gasperini R, Gibson G. Dev Genes Evol 1998: 207: 462-470.

be obtained by environmental insults to embryos during sensitive periods. Treatment of developing wild-type flies with ether vapor, for example, can induce mimics or phenocopies of homeotic transformations caused by mutations in the Ubx gene. C. H. Waddington showed decades ago that if one repeatedly selected for individuals demonstrating the bithorax phenocopy, the resulting populations would show a greater frequency of response to the treatment, including individuals that displayed the phenotype independent of treatment. The selection therefore uncovered genetic variation that affects development.

Developmental and molecular genetic analysis of the Ubx gene in response to selection for the ether-induced bithorax phenocopy revealed that ether induces loss of Ubx expression in patches of cells in the haltere imaginal disc. In selected populations, the frequency of lost gene expression increases and is heritable. Therefore, genetic differences must exist, either in cis or trans to the Ubx gene, that influence the susceptibility of the Ubx gene to loss of expression following exposure to ether. One site that responded to selection for phenocopy induction was mapped downstream of the Ubx coding region in a large cis-regulatory region. This finding suggests that genetic variation in a cis-regulatory element of the Ubx gene affects the fly population's susceptibility to loss of gene expression in response to ether.

Cryptic variation is also detectable between species. D. melanogaster and D. simulans adults exhibit identical patterns of thoracic bristles, a pattern that is highly stereotyped and rarely variant among individuals. Yet, when interspecific hybrids are produced, the hybrids lack a variable number of bristles. The loss of bristles is associated with decreased transcription of the achaete and scute genes and genetic experiments suggest that divergence in achaete/scute cis-regulatory elements and in interactions with trans-acting genes (regulators of achaete/scute) has occurred between species. These observations suggest stasis in adult patterns is maintained by compensatory interactions within each species and that this compensation is disrupted when alleles of each species are combined in hybrids.

The phenotypic variation unmasked by artificial selection or interspecific hybrids suggests that there is a reservoir of cryptic genetic variation among developmental genes that can be tapped during evolution by natural selection. Changes in the environment or introduction of a new mutation into a population can quickly uncover genetic variation as observable pheno-typic variation. The increase in phenotypic variation provides a broader range of fitnesses upon which natural selection may act.

Was this article helpful?

0 0

Post a comment