Drosophila Comparative Genomics: The Evolution Of Proteincoding Genes, Sex Chromosomes, And An Ancestral Yautosome Translocation In D. Pseudoobscura
The recent sequencing of ten new genomes, bringing the total number of sequenced Drosophila genomes to 12, allowed Drosophila comparative genomics to be done now in the context of a phylogeny. This dissertation describes several studies that take advantage of these advances in Drosophila comparative genomics. A central theme in the dissertation is how properties of a gene and its genomic environment affect rates of protein evolution. We find that many genic factors influence protein evolutionary rate, especially the level and breadth of gene expression. The genomic environment, such as local recombination rate and genomic location, can also influence evolutionary rates. Selection at one site can influence selection at linked sites, especially in regions of the genome with low recombination rates. We are able to detect these effects on a genome-wide scale. The sex chromosomes have particularly interesting effects on the evolution of genes. Natural selection is expected to be more efficient on the X chromosome for new, recessive mutations because the X is hemizygous in males. We do not find consistent signals of more efficient positive selection on the X chromosome than the autosomes; however we do find more efficient purifying selection on the X chromosome. A striking example of the impact of genomic location on gene evolution is in D. pseudoobscura, where the ancestral Drosophila Y chromosome translocated to an autosome. We mapped this translocation to the dot chromosome. We find that the rDNA repeats, which are responsible for X-Y pairing in male meiosis, were lost from this ancestral Y chromosome, and the current Y chromosome of D. pseudoobscura has acquired and amplified the intergenic spacer repeats (IGS) of the rDNA. We hypothesize that the new location of the IGS functions to maintain X-Y pairing in male meiosis in the absence of rDNA. The most interesting feature of the Y-to-dot translocation is that the genes shrank 10-fold after moving. A survey of polymorphism and divergence on the dot revealed significantly reduced levels of variation and frequency spectra skewed towards rare variants. We hypothesize that this is due to selective sweeps from positive selection favoring the shortening of introns and that the most recent selective sweep was approximately 228,000 years ago.
dissertation or thesis