The heads of primates, including humans, all have the same basic parts, yet every head is different. Craniofacial variation has important implications for diet, defense, posture, balance, hearing, vision, and the accommodation of the expanding primate brain. The skull is a complex adaptive structure, and the question naturally arises pertaining to the genetic mechanisms that assemble and coordinate its shape. This project will determine whether variation in head shape occurs along similar anatomical axes among living monkeys and apes, their fossil ancestors, a pedigree of 800 extant baboons, and in laboratory mice. Computerized tomography (CT) scans will be acquired from these heads and used to measure variation in three dimensions. The genes responsible for this variation will be identified using data from baboon and mouse. Developmental pathways are shared among mammals, so mice provide an important comparative and experimental basis for primate findings. The manner by which genes structure variation in head shape will be investigated by combined comparative DNA sequence and bioinformatic analysis, and by engineering similar changes in mice to those seen in baboons. Many genes doubtlessly contribute to structures like the head, and a general challenge to modern evolutionary and developmental genetics generally is to understand whether some genes are repeatedly involved in similar changes in complex traits over the millions of years and different branches of evolution, or whether different genes in the shared developmental pathways are responsible in each instance.
The project will generate a large amount of data which will be posted on a publicly accessible website along with information on analytical approaches and results. The PIs have a strong record of training women and minority students and they will continue this tradition at all levels from undergraduate to postdoctoral. Human evolutionary studies have a strong public appeal and this project will be exceptionally popular to the public media.
Using two model species, baboon and mouse, the objective of this project was to choose various measures of the skull that seem to have evolutionary/biological importance, and to identify regions of the genome that contribute in sufficiently detectable ways to variation of those measures. Follow-up was planned with experimental models to characterize the tissue of expression and possible relevance of genes so identified. We did not expect to find single genes with dominating influence on these traits but to understand the nature of genomic causal complexity involved in these traits and their evolution. We acquired 3D computed tomography scans of the heads of >800 baboons of a known geneology and micro computed tomography scans of the heads of >1200 mice from the F32 generation of a known cross. 3D measures of every skull were used to quantitatively estimate statistical parameters of the size and shape of the skulls of these two samples. The images and all 3D landmark data are avaiable to any scientist interested in analyzing these data. We expected as outlined in our original application that these traits are complex, affected by multiple genes, but that some shared contributing genome regions would be found in both model systems and inform us about evolution of these traits in humans. During the years of our funding, the general genome-mapping technology made huge advances that we could not (for funding and logistic reasons) incorporate. There has been a massive proliferation of genomewide mapping studies of various traits, confirming our general findings and expectations about the basically polygenic nature of craniofacial traits. Even in our well-controlled study conditions, craniofacial dimensions reflect the effects of many genome regions, regulatory and non-coding and most individual genomic contributors have only weak effects. Multiple weak effects is what we have found and the reason is simply that major effects (like mutations) are likely to be seriously damaging and not compatible with life. Indeed, the genes that are known to cause disease when seriously mutated generally did not appear as causal ‘hits’ in our mapping results. These findings clearly support what should become the working model in anthropology, and more broadly in understanding the genomic causation and evolution of complex traits. Indeed, it is problematic whether defining even so simple a trait as a craniofacial distance measure can be used as a well-posed causal question because the underlying genomic contributors to these traits typically serve many different functional roles in modern organisms and probably evolved, piece-meal, serving diverse roles during their evolution. The bottom-line lesson here is that we should not expect to have neat, enumerable genetic causation of traits that have evolved in the usual slow, gradual Darwinian way by natural selection (or genetic drift). We had hoped to identify candidate genome regions in the baboon that could be compared in the mouse, with the ability to engineer changes in the mouse to test their effects. For the same reason as just described—causal complexes of individually very small effects—this was not practicable. In addition, the timing of the mapping simply could not provide such leads early enough in the project. Instead, and to address an objective of our 2007 proposal, we took a different approach to understand the genomic aspects of craniofacial development. The idea is to try to get a sense of which tissues are most important during development in relation to achieved head shape. To do this, we have taken an available transgenic mouse model of craniofacial disease (Apert syndrome), and used crosses with other mice to engineer transgenic mice that expressed the anomalous Fgfr2 mutation only in specific tissues during embryological development (osteoblasts, neural crest, mesenchyme, endochondral bone, intramembranous bone) to see which altered tissue had which effects on the resulting head shape. This work continues using other funds after the current Hominid funds were exhausted. Results indicate that some specific tissues have notable shape effects when using only a defective Fgfr2 signal receptor (intramembranous bone), while other tissues similarly altered have no shape effect (endochondral bone). This work will provide evidence of how the complex tissue environment works to generate head shape. In summary, our results champion a rethinking of our ideas about the genomic nature and processes of evolution of complex traits that constitute our species. Genomic variation affects these traits, and they have evolved with some assistance from natural selection, but enumerating individual contributors is not important nor central to the objectives of anthropology. When so many genomic contributors affect a trait, it is likely that each individual contributing variant evolves largely by drift even when the trait is evolving adaptively. While there will be some exceptions, identifying selective reasons for a trait’s evolution at the genome level, will be problematic at best. Our results reveal the staggering need for new ways to think about these important, long-standing questions.
