Population genetics and transcriptional evolution of genes with male-biased expression in Drosophila

Much of our understanding of evolution by natural selection has come from comparing observed patterns of genetic and phenotypic variation in nature with theoretical predictions of the effects that selection should have on such variation. Due to the ease of DNA sequencing and the relatively simple process by which mutation changes DNA and protein molecules over time, the greatest advances of this kind have predictably come from inferring the historical action of natural selection from patterns of nucleotide and amino acid sequence variation. Following this model, chapter one of this thesis uses a standard set of population genetic and molecular evolution tools to interpret patterns of nucleotide variation in populations of Drosophila simulans. An unusually low amount of nucleotide polymorphism is found among six contiguous genes, and the polymorphism that does exist in this region is partitioned in a highly non-random way. While certain aspects of the data are consistent with the recent action of positive selection, the persistence at low frequency of relatively old alleles with neutral levels and patterns of polymorphism indicate that a simple model of a selective sweep is inadequate. This suggests more complex forms of selection, such as frequency-dependent or epistatic selection. In the second chapter, the same heuristics that guide the inference of selective forces on nucleotide sequences are applied to a completely different type of data - gene expression profiles obtained using cDNA microarrays. Approximately 1/3 of the predicted transcripts in the D. melanogaster genome are assayed for changes in their relative cellular abundance between eight strains of D. melanogaster from around the world. By analogy with nucleotide evolution, the patterns of expression variation found on these arrays is suggested to reflect the differential action of natural selection on subsets of the genome. Integration of these data with other microarray experiments investigating interspecific divergence and sexual dimorphism of gene expression encourages further inferences regarding selection on global gene expression patterns. In chapter three, a verbal model is proposed to explain one aspect of the microarray data, namely, the unusual and rapidly evolving patterns of gene expression in the testes in Drosophila.