| How genetic inputs are used to build complex structures and, furthermore, how genetic inputs are modified to generate intraspecies and interspecies variation in complex structures are two fundamental questions of evolutionary---developmental biology. The craniofacial skeleton is a complex structure in which anatomical variation among species is related to changes in functions as diverse as communication, diet, defense, olfaction, intelligence, vision, posture, and locomotion. Adaptive changes to the skull are well documented in the fossil record. However, until the molecular factors underlying these changes are identified, our understanding of craniofacial evolution is limited. The goal of this dissertation is to use comparative genomics methods to improve current understanding of the molecular underpinnings of craniofacial growth and development.;While change in protein coding sequence undoubtedly affects craniofacial growth and development, it has long been speculated that adaptive changes in gene regulation underlie many of the phenotypic differences between humans and other primates. Thus identification of non-protein coding regions bearing the signature of either adaptive or constrained selection is an important step in understanding the evolutionary history of humans and of other species. I have modified the standard McDonald-Kreitman test (1991) to examine molecular evolution in non-protein coding regions of human DNA, using repeat sequences ancestral to catarrhine primates to represent neutrally evolving sequence.;Results of this test (called the MKAR test) are used to find non-protein coding regions throughout the human genome that are candidate targets of natural selection over the past 23Myr (comparisons with macaque) or 6Myr (comparisons with chimpanzee) of primate evolution. These are potentially functional sequences, some of which may be regulatory regions. Non-neutral MKAR regions show several intriguing associations. For example, segmental duplications are highly enriched in MKAR regions with footprints of adaptive non-protein coding evolution. Additionally, non-neutral MKAR regions show little overlap when divergence is determined with chimpanzee versus macaque, reflecting different adaptive histories among these lineages, as well as highlighting the episodic nature of natural selection. Follow-up of the candidate MKAR regions in the vicinity of craniofacially affecting genes promises to be a fruitful area of further investigation.;In collaboration with the Genomics of Cranial Morphology Project, I have identified human regions homologous to the genomic regions for a group of pedigreed and genotyped baboons, Papio hamadryas, that significantly correspond to landmarked craniofacial distances. These regions are interrogated for genes demonstrated to affect craniofacial growth and development in model animal, human association, and/or familial linkage studies. These candidate craniofacially affecting genes are then examined for both coding and noncoding DNA sequence changes, using available multi-species genomic alignments and the MKAR test. Additionally, genes within 25kb of non-neutral MKAR regions are examined for expression in the head. This project uniquely combines comparative genomic data with phenotypic observations of a nontraditional model organism highly relevant to human biology.;To examine how craniofacially affecting genes have changed over phylogenetic time, I use whole-genome sequence alignments of human, chimpanzee, macaque, mouse, and dog to estimate rates of amino acid change in orthologous craniofacial genes across approximately 75 million years of evolution. As a set, these craniofacially affecting genes show strong evidence for unusual evolutionary constraint, with significantly fewer amino acid changes than other genes used as a comparison set. Amino acid sites in individual craniofacial genes that have a high probability of having undergone adaptive evolution across these five lineages tend to be located on the surface, or most solvent accessible, regions of a protein. Known human disease-associated mutations are found at the most highly constrained sites, and most amino acid changes associated with dysmorphogenesis are radical with respect to chemical property.;In this dissertation, I show how modern genomic and bioinformatics approaches, combined with experimental results, can begin to fill a gap in our understanding of craniofacial genetic mechanisms and their evolution. This work highlights the predictive power of comparative genomics and demonstrates the utility of an interdisciplinary approach to examining DNA sequence affecting a highly complex and evolutionarily critical structure. Furthermore, while highlighting genomic regions associated with craniofacial phenotypes that may have been targets of selection in primate evolution, this dissertation also highlights the many challenges of decipering genotype---phenotype relationships. |
