Naturally occurring species show spectacular differences in morphology, physiology, lifestyle, and behavior. They also differ in disease susceptibility and life span. Although the genomes of many organisms have now been completely sequenced, we still know relatively little about the specific DNA sequence changes that underlie interesting species-specific traits. My laboratory is using a combination of genetic and genomic approaches to identify the detailed molecular mechanisms that control evolutionary change in vertebrates, with a focus on five fundamental questions:
1. Are new evolutionary traits controlled by countless genetic differences of small effect, or by a few genetic changes with large effects?
2. What specific genes have changed to produce interesting evolutionary differences seen in nature?
3. What kinds of mutations have occurred in these genes (e.g., dominant or recessive, coding or regulatory, preexisting or de novo)?
4. How predictable is evolution? If you know how evolution has occurred in one population, is it possible to predict the genes and mutations that also underlie the same trait in different populations?
5. How has evolution produced the unique characteristics of humans?
We use a variety of methods in mice, sticklebacks, and humans to study these questions. Mice are often the best system available for asking detailed mechanistic questions in mammals and for testing the phenotypic effects of particular sequence changes seen in other species. We have used classical genetics in mice to identify genes controlling formation of bones, patterning of joints, and relative sizes of forelimbs and hindlimbs. Our genetic studies of the mouse short ear and brachypodism genes provided the first genetic evidence that bone morphogenetic proteins (BMPs) play an essential role in formation of both bone and joints, as well as in the repair of bone fractures in adult animals.
BMPs have the remarkable ability to stimulate the entire process of cartilage and bone formation when expressed at new sites in mammals. Since BMPs are the key signals used by vertebrates to initiate skeletal formation in vivo, much of the pattern of the skeleton may be encoded by the regulatory sequences that lay out the expression of BMPs in specific patterns during embryonic development. We have used genetics, genomics, and comparative sequencing to identify a remarkable array of long-distance, modular regulatory elements surrounding the Bmp5, Gdf5, and Gdf6 genes. These sequences correspond to individual "anatomy" elements that help control the size, shape, and number of individual bones and joints. Further study of these anatomy regulatory elements should provide a much more detailed picture of the molecular mechanisms that control the formation and patterning of the vertebrate skeleton.
Stickleback fish offer an unusually powerful system for studying the molecular basis of evolutionary change in natural environments. These fish have undergone one of the most recent and dramatic evolutionary radiations on earth, colonizing countless new freshwater lake and stream environments generated by widespread melting of glaciers at the end of the last ice age. Many freshwater sticklebacks have evolved dramatic differences in morphology, physiology, and behavior in the last 15,000 years. Although often reproductively isolated in nature, recently evolved stickleback species can still be crossed using artificial fertilization in the laboratory. This provides a unique opportunity to identify the key chromosome regions, genes, and mutations that control repeated evolution of new traits and new species under a full range of fitness constraints in the wild.
Thousands of papers and several full-length textbooks have been written on the ecology, morphology, paleontology, and adaptive significance of stickleback traits. We have developed a complete set of genetic and genomic resources for this classic system, including the first genome-wide linkage maps, transgenic methods, expressed sequence tag (EST) collections, large-insert bacterial artificial chromosome (BAC) libraries, and physical maps useful for positional cloning (molecular work done in collaboration with Jane Grimwood, Jeremy Schmutz, Devin Absher, and Richard Myers [HudsonAlpha Institute]; Chris Amemiya [Benaroya Research Institute at Virginia Mason, Seattle]; Pieter de Jong [BACPAC Resources, Oakland, California]; and Marco Marra and Jacqueline Schein [University of British Columbia]). We also nominated threespine sticklebacks for complete genome sequencing to the National Institute of Human Genome Research. We have worked with the Broad Institute and Ensembl on development, assembly, and annotation of the first whole-genome sequence for Gasterosteus aculeatus, which was released in 2006.
Using genetic crosses and genomics, we have now successfully identified the molecular mechanisms that control repeated evolution of armor plate patterning, pelvic reduction, and skin color changes in nature. Our studies show that big evolutionary changes can be controlled by single-chromosome regions. These regions produce large morphological effects because of genetic changes in key transcription factors and developmental signaling molecules (Pitx1, Ectodysplasin, Stem Cell Factor/Kit ligand). Although null mutations in these essential developmental control genes are typically deleterious or lethal in laboratory animals, natural stickleback populations have made regulatory alterations in these genes, thus producing large morphological effects in particular tissues, while preserving overall viability. The key altered genes show strong signatures of positive selection in natural populations, providing a definitive link between evolution of new traits in wild environments and the origin and selection of particular sequence variants at the DNA level.
The widespread evolution of sticklebacks also offers an opportunity to test whether the same or different genes are used when the same traits evolve in widely separated locations. Genetic mapping, complementation tests, and gene expression studies suggest that similar genetic mechanisms are used when the same traits are selected in multiple populations around the world. How far might such reuse of particular genes extend? Our recent studies suggest that the genes underlying major morphological change in sticklebacks are also reused when similar morphological changes evolve even in distantly related animals. Thus, we find that pelvic reduction in marine mammals shows the same morphological features previously associated with Pitx1-mediated pelvic reduction in both mice and sticklebacks. And the genes and mechanisms that we first identified in mice or sticklebacks also control major differences in human morphology, human disease incidence, and differences in skin color in billions of people around the world.
Building on this work, we have now begun a variety of new projects to identify genomic mechanisms that underlie the evolution of key human traits. These studies combine human-specific sequence changes detectable by comparative genome analysis (with Gill Bejerano's lab at Stanford University), patterns of evolutionary change we have previously learned in sticklebacks, signatures of selection in human populations, and functional tests of human-specific sequence changes using mouse models. We are still a long way from knowing the genomic mechanisms that have made us human. We believe, however, that molecular mechanisms contributing to human-specific traits can now be studied and that progress in this area will lead to important new insights into both human health and human disease.