Studies on control of mobile retrotransposons in Arabidopsis シロイヌナズナ属における転移可能なレトロトランスポゾンの制御に関する研究
Studies on control of mobile retrotransposons in Arabidopsis
Retrotransposons are major components of the eukaryotic genome. They have proliferated in the genome during evolution through their copy-and-past type transposition. However, in the current genome, most of them are inactivated by degeneration or silencing mechanism of hosts. In the model plant Arabidopsis thaliana, most of retrotransposons are located in pericentromeric region where epigenetic marks such as DNA methylation and histone H3 lysine 9 methylation (H3K9me) are enriched, that are associated with inactive state of chromatin, heterochromatin. In Part 1, I identified a mobile endogenous LTR (long terminal repeat) retrotransposon, ATGP3 in ddm1 (decrease in DNA methylation 1) mutant of A. thaliana, where DNA methylation is globally decreased. ATGP3 is transcribed and it retrotransposed in ddm1, while it is inactive in wild-type. ATGP3 is also mobilized in the double mutant of DNA methyltransferase for CG sites, MET1 (MTHYLTRANSFERASE 1) and for non-CG sites, CMT3 (CHROMOMETHYLASE 3), suggesting that ATGP3 is repressed by both CG and non-CG methylation. ATGP3 was firstly identified as a mobile endogenous LTR retrotransposon in A. thaliana.Distribution of the retrotransposons in the genome is not uniform. Most of them are located in pericentromeric region. However, it is not known how retrotransposons are accumulated in pericentromeric region. There are two hypotheses for the biased distribution. One is that retrotransposon insertions in genic region have been eliminated by natural selection and they have remained in gene-poor heterochromatic region. The other is that retrotransposons preferentially integrate into gene-poor region. However, mobile retrotransposon with targeted integration into such region has not been identified in plants.In Part 2, I investigated the second hypothesis using another retrotransposon family, COPIA93. COPIA93 family is one of the retrotransposon families mobilized in ddm1 mutant of A. thaliana. COPIA93 related elements are abundant in A. lyrata, which is closely related species to A. thaliana, and their sequences are very similar to each other, suggesting their recent amplification in A. lyrata. In addition, most of the copies of COPIA93 in A. lyrata are located in centromeric repeats, implying that they specifically integrate into centromeric repeats. In order to examine insertion specificity of their de novo integrations, I cloned and introduced one of the COPIA93 family retrotransposon in A. lyrata, which we named Tal1, into A. thaliana. Introduced Tal1 was transcribed and retrotransposed in A. thaliana. Surprisingly, Tal1 exclusively integrated into centromeric repeats, suggesting that Tal1 has integration specificity into centromeric region. Tal1 provides us useful experimental system to investigate evolutionary dynamics of retrotransposon.