Despite the ubiquity of transposons in eukaryotic genomes, their evolutionary role remains controversial. The discovery of several domesticated genes has suggested that transposons can gain host functions, and thus contribute to the evolution of their host. Here, I present the results of a genome-wide screen for transposon-derived host genes, which was based on the idea that, once domesticated, the open reading frame of such elements would be maintained, while terminal structures necessary for transposition would be lost. Eight-hundred-and-sixty-three such transposon-dissociated elements were mined from the genome of Arabidopsis thaliana var. Columbia-0, of which less than 10% are associated with expression data. Phylogenetic analysis of Mutator superfamily genes in the genomes of Oryza sativa ssp. japonica (cv Nipponbare) and Arabidopsis, including 121 Mutator-derived transposon-dissociated elements, found that only two gene families are taxonomically widespread. MUSTANG1, a member of one of these families, appears to be under purifying selection. Thus, despite the dearth of taxonomically widespread and/or expressed transposon-dissociated elements, MUSTANG1, as well as three transposon-dissociated elements that may be associated with mutant phenotypes, might be newly discovered transposon-derived host genes.

Transposons are DNA segments that can move or transpose themselves to new positions within the genome of an organism. Biologists need to predict preferred insertion sites of transposons to devise strategies in functional genomics and gene therapy studies. It has been found that the deformability property of the local DNA structure of the integration sites, called Vstep, is of significant importance in the target-site selection process. We considered the Vstep profiles of insertion sites and developed predictors based on Artificial Neural Networks (ANN) and Support Vector Machines (SVM). We trained our ANN and SVM predictors with the Sleeping Beauty transposonal data, and used them for identifying preferred individual insertion sites (each 12bp in length) and regions (each 100bp in length). Running a five-fold cross-validation showed that (1) Both ANN and SVM predictors are more successful in recognizing preferred regions than preferred individual sites; (2) Both ANN and SVM predictors have excellent performance in finding the most preferred regions (more than 90% sensitivity and specificity); and (3) The SVM predictor outperforms the ANN predictor in recognizing preferred individual sites and regions. The SVM has 83% sensitivity and 72% specificity in identifying preferred individual insertion sites, and 85% sensitivity and 90% specificity in recognizing preferred insertion regions.

The sequencing of eukaryotic model organisms has provided us an unprecedented opportunity for a genome-wide characterization of Transposable Elements (TEs) and the study of TE-host relationships. By developing methodologies on database mining, we explored the existence of MUtator (Mu)-Like Elements (MULEs) in Arabidopsis thaliana and Oryza sativa. Mu elements were first discovered in Zea mays; so far, a dozen of the elements have been identified in the genome. We identified a total of 1392 MULEs from the sequenced Arabidopsis genome. They represent one of the most abundant, diversified, yet still mobile DNA transposon families in eukaryotes. The Arabidopsis MULEs are composed of not only the elements showing the typical Mu-family-specific terminal structure (that is the long Terminal Inverted Repeat, TIR), but also a novel type of non-TIR MULEs. Some of this latter type of elements was found to be active both transcriptionally and transpositionally. To understand host-mediated genome-wide regulation(s) on the MULE system in Arabidopsis, we characterized 235 MULE mobility-specific genes (or mudrA-like genes) by mapping them on the sequenced Arabidopsis chromosomes and performing a genome-wide expression assay utilizing Arabidopsis METHYLTRANSFERSE1 (MET1) mutant (met2) plants, we showed that MET1-mediated global CpG methylation can only repress a portion of the gene family; its efficiency depends largely on the gene locations within the context of Arabidopsis chromatin remodeling: stronger in heterochromatic regions but weaker in euchromatic ones. This finding suggests that the Arabidopsis heterochromatic regions are not just a graveyard for the accumulation of defective elements; rather, they may have been playing an important role on the repression of TE activity via, at least in part, exerting MET1-mediated silencing effect. Our expression analysis also suggested that a TIR structure is not necessarily required for the MET1-mediated si

No description available.

Thesis (M.S.)--Oregon State University, 2007. / Printout. Includes bibliographical references (leaves 59-63). Also available on the World Wide Web.

No description available.

No description available.

No description available.

No description available.

No description available.