Intrinsic and extrinsic evolution of Helitrons in flowering plant genomes

Abstract

Helitrons are a recently discovered class of eukaryotic transposable elements that are believed to transpose by a rolling circle mechanism. Because Helitrons frequently acquire and fuse fragments of multiple genes, they may be major contributors to the creation of novel genes by exon shuffling. This dissertation provides a comprehensive study of Helitrons in flowering plant genomes including their identification in several different genomes, their structural features, and their roles in genome evolution. Helitrons can be difficult to identify because they have few and tiny conserved structures. We developed a new approach to effectively identify Helitrons (tested in Arabidopsis and nematodes) and further refined and demonstrated this approach on a few completely sequenced plant genomes (Medicago, rice and sorghum). We discovered a large number of new elements and new element families, and also a few new Helitron characteristics. We identified initiation and termination bypass events that led to new 5′ or 3′ ends that created newly active families and subfamilies of Helitrons. We found that Helitrons preferentially insert into AT-rich regions, that they prefer to insert near other Helitrons, and that the predicted hairpins near their 3′ ends would have high predicted melting temperatures. Maize Helitrons are known to acquire gene fragments frequently. With the completion of the maize genome sequencing project this year, we were able to perform a large-scale search for Helitrons in the maize genome. We discovered 1930 intact elements in the maize genome, and were able to predict more than 20,000 total elements that account for just over 2% of the sequence assembly. We found 1194 intact Helitrons that contain fragments of regular nuclear genes, from 840 independent acquisition events. A total of 4% of the captured gene fragments appeared to be under negative selection and another 4% under positive selection. The results also indicated that gene fragments acquired in the same orientation as Helitron genes persist longer than gene fragments acquired in the antisense orientation. Finally, we identified candidates for active elements from the rice and maize genomes by transposon display of sibling plants and of plants derived from tissue culture. The results suggest that, in most rice and maize lines, Helitron activity is non-existent or quite low.