|
内容記述 |
DNA cytosine methylation found in several eukaryotic genomes has been implicated in ensuring proper gene expression and stabilizing genome structure through the silencing transposon activity In order to examine these possibilities, Arabidopsis provides a genetically tractable system because viable mutants affecting genomic DNA methylation are available. MET1 (Methyltransferase 1 ) is required for maintenance of DNA methylation at CG site. Arabidopsis additionally methylates cytosines at non-CG site, which depends on CMT3 (Chromomethylase 3). The third gene necessary for DNA methylation is DDM1 (decrease in DNA methylation), which encodes a chromatin remodeling factor. The mutation in DDM1 gene reduces DNA methylation at both CG and non-CG sites, and causes several morphological abnormalities.<br /> Arabidopsis endogenous transposon CACTA1 was originally isolated as an inserted gene from a mutant allele induced by ddm1 mutation. This transposon is silent and does not transpose in wild type background but it is mobilized in ddm1-induced DNA hypomethylation background. Although this observation is consistent with the interpretation that DNA methylation is important for silencing transposons, it remains unclear whether the loss of DNA methylation is sufficient for mobilization of transposons, because the primary function of DDM1 is likely to be chromatin remodeling.<br /> In this study I examined the behavior of transposon CACTA1 in mutants defective in DNA methyltransferases, MET1 and CMT3. The met1-cmt3 double mutation induced high frequency transposition of CACTA1 element. However, the single mutation, either met1 or cmt3, was not sufficient for mobilization of CACTA1. These observations suggest that both CG and non-CG methylation redundantly function as genome defense against transposon movement. Chromomethylase-mediated non-CG methylation has only been reported in plant kingdom. It might have evolved as an additional epigenetic mark for silencing transposons.<br /> Epigenetic alteration of gene expression not based on a change in DNA sequence has been reported in plants. Notably, these epigenetic mutations are often stably inherited over generations. However, the biological significance of epigenetic inheritance remains understood. I propose here that the maintenance of silent state over generations can function as genome defense against deleterious effect of transposon movement. Transposon CACTA1 activated by ddm1 mutation remained mobile in wild type background. In addition, the CACTA1 hypomethylated by ddm1 mutation remained hypomethylated in the presence of wild-type-derived functional DDM1 copy. This is in contrast to the wild type plants, in which the CACTA1 is heavily methylated and kept silent. These results suggest that de novo silencing is not efficient for immobilization, at least or this class of transposon. Silencing of transposons depends on the maintenance of silent state over generations with the stable inheritance of DNA methylation.<br /> This system enables us to identify the integration specificity of transposon in wild type background. Transposons and their derivatives are not randomly distributed in the genome. They tend to be localized near centromeric heterochromatin in several eukaryotes including plants. Although it remains unknown how the transposons accumulate in such region, one possible explanation is that the transposon preferentially integrates into heterochromatin. In order to test this possibility, I examined the integration sites of transposon CACTA1 induced in wild type DDM1 background. The CACTA1 mobilized in wild type did not show preferential integration into wild-type-derived heterochromatin or transposon-rich regions, despite the accumulation of CACTA-like sequences near centromeric and transposon-rich regions in wild type Arabidopsis strains. These results suggest that the centromere-biased distribution of CACTA elements is not directly caused by the targeted integration into heterochromatin. |
要旨・審査要旨 (232.7 kB)
