UC Irvine

Transposable elements are DNA sequences capable of moving or copying and reinserting themselves in the genome of a host cell. The Hermes DNA transposon was originally found in the housefly Musca domestica. Since its discovery, Hermes has been adapted to multiple eukaryotic species for random mutagenesis or mapping of chromatin structure. The Ty3 retrotransposons of S. cerevisiae is a long terminal repeat (LTR) mobile element of the gypsy–like class. Ty3 is unique among retroelements for its highly specific targeting of RNAP3-transcribed genes for integration. In this dissertation, each of these mobile elements are utilized for different purposes. The Ty3 retrotransposon was used to investigate the relationship between integrase and host factors that direct insertion site selection. Ty3 integration events were quantitatively profiled genome-wide, demonstrating an unexplained bias within an already limited selection of RNAP3 targets. This bias was not defined by the host chromatin environment flanking target sites, and investigation of RNAP3 factors found poor agreement with Ty3 targeting bias. DNA sequence composition and structure of local target sequence were found to regulate integration frequency, in a motif-independent manner. The Hermes DNA transposon was used to establish a functional genomics system for the oleaginous yeast Yarrowia lipolytica, a quickly emerging and valuable microbe with many biotechnology applications. Identification of essential genes provides important information required for genome engineering. Hermes was used to mutagenize a population of Y. lipolytica. To classify essential and non-essential genes, transposon were profiled in surviving mutants by statistically testing under-representation of insertions relative to an expected number derived for each gene. The Y. lipolytica genome was categorized as 20.8 % essential, 71.1 % non-essential and 8.1 % poor essential. Many essential differed from S. cerevisiae, demonstrating the utility of a model obligate aerobe compared to current model yeasts. Pooled mutants were grown for 80 generations in glucose or glycerol, an industrial waste stream, to identify genes contributing to fitness. Glycerol media created a much more competitive environment than glucose. Gene classifications inform future metabolic studies by fine tuning flux analysis and genome scale models, while mutant libraries provide a pipeline for rescreening phenotypes relevant to industrial applications.