A genetic, biochemical, and population analysis of MGL, a non-LTR retroelement from the plant pathogenic fungus Magnaporthe grisea

This dissertation describes the characterization of a novel transposable element isolated from the plant pathogenic fungus Magnaporthe grisea. The sequence of MGR583, a previously reported repeated DNA fragment, was completed and shown to have features characteristic of non-LTR retroelements (LINEs). These include an element length of 5.9 kb, the lack of flanking long terminal repeats, the presence of short (6-13 bp) direct repeats flanking many element copies, and two principal open reading frames (ORFs). The first ORF is 570 amino acids in length and contains homology to the gag ORFs found in many retroelements. The second ORF is 1,295 amino acids in length and has strong homology to reverse transcriptases (RT) ORFs found in non-LTR retroelements (LINEs). In accordance with these results, the name of the repeat was changed to MGL for Magnaporthe grisea LINE. Analysis of the 3' terminus of MGL showed 90% homology to the 3' terminus of Mg-SINE, suggesting an evolutionary relationship between these two elements. A survey of the distribution of MGL in populations of M. grisea showed the element to be present in all isolates tested. Copy number was not uniform between isolates, with approximately fifty copies present in rice isolates and between less than 10 and up to 50 copies in the 17 non-rice isolates tested. A PCR-based assay was designed and used to screen M. grisea isolates for polymorphic MGL insertion loci. Thirteen polymorphic MGL insertions were scored and used to construct a phylogenetic tree that included 11 non-rice isolates and 20 rice isolates. The results strongly suggested that development of virulence on rice was a single event correlated with the acquisition of virulence on several other grass species. In addition, the observation that rearrangements occurred at one of the insertion loci in some rice isolate strains support the proposal that there is considerable plasticity in the genomes of these isolates. Finally, a yeast transposon ( Tyl) system was used to express and test the second ORF for RT activity. No activity was detected for any of the MGL RT constructs tested