@misc{oai:ir.soken.ac.jp:00001694, author = {佐々木, 卓 and ササキ, タク and SASAKI, Taku}, month = {2016-02-17, 2016-02-17}, note = {DNA methylation is an epigenetic mark important for heterochromatin silencing and regulation of gene expression. In plants, DNA methylation is observed in cytosines in all of contexts, CG, CHG, and CHH. DNA methylation in all contexts of cytosines depends on a SWI2/SNF2 chromatin remodeling factor called DDM1 (decrease in DNA methylation 1). In the ddm1 mutant, DNA methylation was reduced globally, and silent repetitive elements were de-repressed. Although ddm1 mutant initially grows normally, various developmental abnormalities were induced during repeated self-pollinations. Most of the abnormalities were induced by de-repression of silent element caused by the DNA hypomethylation. However, one abnormal phenotype named bonsai (bns) was associated with DNA hyper-methylation and silencing of the responsible gene, BNS, which encodes a protein similar to Anaphase Promoting Complex 13 (APC13). BNS locus is hyper-methylated locally in the background of global hypomethylation. To understand this enigmatic phenomenon, here I examined the factors involved in the BNS hyper-methylation. In addition, I analyzed effect of the ddm1-induced DNA hyper-methylation in genome-wide level.
   To understand the factors required for this phenomenon, I first generated double mutants between ddm1 and various factors involved in the epigenetic modifications. After repeated self-pollinations, I examined DNA methylation states in BNS locus. Previously-characterized components of RdDM pathway were dispensable for ddm1-induced BNS methylation. On the other hand, mutation in plant-specific DNA methyltransferase, CHROMOMETHYLASE 3 (CMT3) suppressed the induction of DNA hyper-methylation. CMT3 was considered as a DNA methyltransferase for maintenance of non-CG methylation. The mutation in KRYPTONITE (KYP), a histone H3K9 methyltransferase which directs non-CG methylation by CMT3, also suppressed the induction of BNS methylation. These results indicate that ddm1-induced BNS methylation was mediated by KYP and CMT3, rather than RdDM pathway.
   Next, to understand the genome-wide effect of ddm1-induced DNA hyper-methylation, I examined global states of DNA methylation in ddm1mutant by MeDIP-chip analysis. Consistent with previous reports, drastic reduction of DNA methylation in transposable elements was observed as a rapid effect in ddm1 mutant. During self-pollinations, DNA methylation states changed in many loci. Some loci kept decreasing DNA methylation during self-pollinations. I also found numerous loci which behave in BNS-like manner. Induction of DNA methylation was observed in both genes and transposable elements. The slow DNA hyper-methylation was induced in all the three contexts, but it was generally most frequent at CHG sites. Compared to genic regions, induction of DNA hyper-methylation in transposable elements tend to be moderate. In most loci, as was the case for the BNS locus, induction of DNA hyper-methylation was suppressed in the ddm1 kyp double mutant.
   In this study, I showed KYP and CMT3 mediated induction of DNA methylation in ddm1 mutant. Interestingly, although KYP and CMT3 were considered as components required for the maintenance of DNA methylation in non-CG contexts, this study revealed that they are also involved in induction of DNA methylation including CG context in ddm1 background. KYP has a SRA domain, which can bind to methylated DNA, and CMT3 has a chromo-domain, which can bind to methylated histone, so they could be recruited to epigenetic modifications catalyzed by the other factor. This self-reinforcement mechanism may be involved in the spreading of DNA methylation in ddm1 mutant., application/pdf, 総研大甲第1348号}, title = {Genetic dissection of DNA hypermethylation induced by ddm1 mutation.}, year = {} }