Role of CG and non-CG methylation in immobilization of transposons in arabidopsis

Methylation of cytosine residues in eukaryotic genomes is often associated with repeated sequences including transposons and their derivatives [1, 2]. Methylation has been implicated in control of two potential deleterious effects of these repeats: (1) uncontrolled transcription [2-4], which often disturbs proper expression of nearby host genes [5, 6], and (2) changes in genome structure by transposition and ectopic recombination [2, 7]. Arabidopsis thaliana provides a genetically tractable system to examine these possibilities, since viable mutants in DNA methyltransferases are available. Arabidopsis MET1 (METHYLTRANSFERASE 1, ortholog of mammalian DNA methyltransferase Dnmt1) is necessary for maintaining genomic cytosine methylation at 5'-CG-3' sites [8, 9]. Arabidopsis additionally methylates non-CG sites using CHROMO-METHYLASE3 (CMT3) [10, 11]. We examined the mobility of endogenous CACTA transposons in met1, cmt3, and cmt3-met1 mutants. High-frequency transposition of CACTA elements was detected in cmt3-met1 double mutants. Single mutants in either met1 or cmt3 were much less effective in mobilization, despite significant induction of CACTA transcript accumulation. These results lead us to conclude that CG and non-CG methylation systems redundantly function for immobilization of transposons. Non-CG methylation in plants may have evolved as an additional epigenetic tag dedicated to transposon control. This view is consistent with the recent finding that CMT3 preferentially methylates transposon-related sequences [12].