In the examples above of pigmentation evolution in vertebrates and insects, genes such as MC1R, bab, or yellow were frequently involved in phenotypic divergence, but there were other genetic routes to similar phenotypic changes. This illustrates that an apparently simple developmental readout (such as melanic pigmentation) can be modified by a variety of mechanisms.
Other characters, however, may be governed by genetic regulatory architectures that may have fewer or greater possible genetic inputs, and therefore different degrees of genetic constraints on evolution. For example, innervated bristles and noninnervated hairs are two types of epithelial projections that adorn most surfaces in insects and other arthropods, and whose formation requires the activity of the regulatory genes of the Achaete-Scute Complex (see Chapter 2) and the shaven-baby gene, respectively. Each of these genes acts as a nodal point (see Chapter 3) that integrates spatial information and whose expression and function is absolutely required for formation of the respective pattern elements. Therefore, diversification in the number or pattern of these structures will always be associated with differences in the expression of these regulatory genes. The key question then is the degree to which diversity is achieved through changes at these regulatory loci, or at other loci whose products affect the expression of these regulatory genes.
In most Drosophila species, there are intricate patterns of hairs covering large portions of the larval body. In D. melanogaster, the formation of these hairs requires the activity of the aptly-named Shavenbaby (Svb) transcription factor (mutants in svb lack hairs, appearing "shaven"), whose expression precisely foreshadows the position of denticles and hairs in the larval epidermis. In the melanogasterspecies group, all members except for D. sechelia bear rows of fine hairs on the dorsal part of larval segments (Fig. 7.7). The close relationship of members of this group allow for some interbreeding in the laboratory, which revealed that the difference in D. sechelia is entirely due to changes at the svb locus, and is correlated with the lack of svb transcription in the region where the hairs form in other species. These findings indicate that cis-regulatory evolution at svb underlies this discrete phenotypic difference.
As in the case of pigmentation pattern diversity, the loss of larval hair patterns is not restricted to just one species, but has also occurred repeatedly in the distantly related D. virilis group (which some 40-60 Ma diverged from D. melanogaster). In each instance, regulatory changes within the svb locus are associated with the phenotypic change. This is not expected because changes in the deployment of an activator of svb would also be a plausible mechanism for hair loss. The parallel, independent loss of dorsal hairs via regulatory evolution at the svb locus demonstrates that morphological evolution may occur through the same developmental mechanism and that such parallel changes in different lineages may be more prevalent than has been anticipated.
The bristle patterns of adult fruit flies also exhibit considerable intraspecific variation and interspecific divergence. Drosophila melanogaster is covered with sensory bristles that serve mechano- and chemoreception functions. Both the number and pattern of bristles on various body parts can vary between individuals as well as between populations. The most successful approaches to understanding intraspecific variation exploit genetic methodologies for assessing the number and identity of genes involved in morphological differences between populations. Much is known about the genes that control the bristle pattern and their functions in the developmental processes that give rise to it. Consequently, the variation and evolution of bristle patterns offers great potential for exploiting our emerging knowledge of developmental genetics so as to better understand morphological variation.
Evolution of trichome patterns via regulatory evolution at the shavenbaby gene
Evolution of trichome patterns via regulatory evolution at the shavenbaby gene
The dorsal hair patterns of five members of the melanogaster species group shown schematically. The distribution of short denticles (small blue projections), fine hairs (curved thin lines), and large denticles (large blue projections) on an individual segment are depicted (anterior cells are open rectangles, posterior cells are shaded blue). In D. sechelia, the fine hairs are absent due to differences at the shavenbaby locus (see photographs on right).
Source: Courtesy David Stern and adapted from Sucena E, Stern DL. Proc Natl Acad Sci USA 2000; 97: 4530-4534.
Bristle number is an example of a quantitative trait that varies continuously in populations. Other quantitative traits include features such as body size, organ size, and life span. Quantitative genetic techniques seek to elucidate the number and identity of those loci (quantitative trait loci (QTLs)) that contribute the bulk of the genetic variance underlying the morphological variance in characters under study. A useful distinction is to discriminate between genes of "small effect," many of which may combine to account for some portion of the variance in a trait, and genes of "large effect," which may account for 5-10% (or much more) of the variance in a given trait.
Studies of Drosophila bristle variation have shown that many loci affect bristle number but only a few loci have large effects that cause most of the variation. Some QTLs are genes known to be involved in bristle patterning, including the achaete-scute, scabrous, and Delta genes. The achaete and scute genes encode transcription factors required for neural precursor formation, whereas the Delta and scabrous genes encode ligands involved in the Notch-mediated signaling pathway that regulates cell interactions in proneural clusters, and ultimately, achaete and scute expression. Because mutations in these genes have dramatic effects on bristle development, their identification as QTLs in bristle number variation implicates them as playing a role in the evolution of bristle number.
The variation at all three loci appears to occur within regulatory regions. In the achaete-scute region, for instance, DNA insertions are strongly associated with variation in bristle number in natural populations. These insertions do not disrupt the protein coding regions, but they do appear to affect the expression of the achaete and scute genes. Two sites within the Delta gene are also associated with bristle variation. Each of these sites is located in an intron, rather than a coding segment of the gene. Similarly, sites associated with bristle variation in the scabrous gene appear within particular regulatory regions. These studies of Drosophila bristle variation have also revealed that the molecular basis for the genetic variation in loci that contribute to morphological variation can involve two or more sequence differences in a gene, not just a single nucleotide substitution.
Another emerging and powerful experimental approach to studying the genetics of morphological variation involves artificial selection on traits of interest. Rather than relying solely on the expressed quantitative variation present in natural populations, it is possible to derive populations with much greater divergence in traits through repeated selection over several generations for individuals with character states at either end of a continuum.
Experiments have been performed to select lines of flies with greater or fewer bristles. Genetic crosses between fly lines with "high" and "low" bristle numbers can be used to estimate the number of loci involved in the divergence and to map the quantitative contributions of genetic regions and individual loci. Many of the candidate loci identified in these experiments encode proteins known to play roles in the patterning and genesis of bristles. Although the specific molecular bases of the genetic differences responsible for the selected morphological divergence remains unknown, we do understand one crucial finding in both natural and artificially selected populations—that is, many genetic loci typically contribute to differences in such modest traits as bristle number on a single fruit fly body part. Because the products of these genes interact through regulatory networks, the effects of genetic variation are usually not strictly additive. Combinatorial interactions—so powerful in developmental processes—clearly play major roles in morphological variation and evolution.
Was this article helpful?