Molecular paleontology of transposable elements from Arabidopsis thaliana

We report results of a comprehensive computer-assisted analysis of new transposable elements (TEs) from Arabidopsis thaliana. Our analysis revealed several previously unknown pogo- and En/Spm-like families and two novel superfamilies of DNA transposons, Arnold and Harbinger. One of the En/Spm-like families (Atenspm) was found to be involved in generating satellite arrays in paracentromeric regions. Of the two superfamilies reported, Harbinger is distantly related to bacterial IS5-like insertion elements, and Arnold contains DNA transposons without terminal inverted repeats (TIRs), which were never reported in eukaryotes before. Furthermore, we report a large number of young and diverse copia-like autonomous and nonautonomous retroelements and discuss their potential evolutionary relationship with mammalian retroviruses. The A. thaliana genome harbors copia-like retroelements which encode a putative env-like protein reported previously in the SIRE-1 retrotransposon from soybean. Finally, we demonstrate a nonrandom chromosomal distribution of the most abundant A. thaliana TEs clustered in the first half of chromosome II, which includes the centromeric region. The families of TEs from A. thaliana are relatively young, extremely diverse and much smaller than those from mammalian genomes. We discuss the potential factors determining similarities and differences in the evolution of TEs in mammals and A. thaliana.