Eukaryotic genomes consist to a significant extent of retrotransposons that are suppressed by host epigenetic mechanisms, preventing their uncontrolled propagation. However, it is not clear how this is achieved. Here we show that in Arabidopsis seedlings subjected to heat stress, a copia-type retrotransposon named ONSEN (Japanese 'hot spring') not only became transcriptionally active but also synthesized extrachromosomal DNA copies. Heat-induced ONSEN accumulation was stimulated in mutants impaired in the biogenesis of small interfering RNAs (siRNAs); however, there was no evidence of transposition occurring in vegetative tissues. After stress, both ONSEN transcripts and extrachromosomal DNA gradually decayed and were no longer detected after 20-30 days. Surprisingly, a high frequency of new ONSEN insertions was observed in the progeny of stressed plants deficient in siRNAs. Insertion patterns revealed that this transgenerational retrotransposition occurred during flower development and before gametogenesis. Therefore in plants with compromised siRNA biogenesis, memory of stress was maintained throughout development, priming ONSEN to transpose during differentiation of generative organs. Retrotransposition was not observed in the progeny of wild-type plants subjected to stress or in non-stressed mutant controls, pointing to a crucial role of the siRNA pathway in restricting retrotransposition triggered by environmental stress. Finally, we found that natural and experimentally induced variants in ONSEN insertions confer heat responsiveness to nearby genes, and therefore mobility bursts may generate novel, stress-responsive regulatory gene networks.

In mammals, the DNA methyltransferase 1 (Dnmt1) faithfully copies the pattern of cytosine methylation at CpG sites to the newly synthesized strand, and this is essential for epigenetic inheritance. In Arabidopsis thaliana, several DNA methyltransferases or chromatin modifiers coupled to methylation changes have been characterized, and mutations that cause loss of their function are recessive. This is surprising because plant gametogenesis includes postmeiotic DNA replication in haploid nuclei before fertilization. Therefore, the recessive character of the mutations excludes the affected components from a regulatory role in postmeiotic maintenance or modification of epigenetic states. Here we show, however, that depletion of A. thaliana MET1, a homolog of mammalian Dnmt1 (ref. 8), results in immense epigenetic diversification of gametes. This diversity seems to be a consequence of passive postmeiotic demethylation, leading to gametes with fully demethylated and hemidemethylated DNA, followed by remethylation of hemimethylated templates once MET1 is again supplied in a zygote.

Transgenerational epigenetic inheritance has been defined by the study of relatively few loci. We examined a population of recombinant inbred lines with epigenetically mosaic chromosomes consisting of wild-type and CG methylation-depleted segments (epiRILs). Surprisingly, transposons that were immobile in the parental lines displayed stochastic movement in 28% of the epiRILs. Although analysis after eight generations of inbreeding, supported by genome-wide DNA methylation profiling, identified recombined parental chromosomal segments, these were interspersed with unexpectedly high frequencies of nonparental methylation polymorphism. Hence, epigenetic inheritance in hybrids derived from parents with divergent epigenomes permits long-lasting epi-allelic interactions that violate Mendelian expectations. Such persistently ''unstable'' epigenetic states complicate linkage-based epigenomic mapping. Thus, future epigenomic analyses should consider possible genetic instabilities and alternative mapping strategies. The term ''epigenome'' refers to the genome-wide distribution of epigenetic marks such as DNA methylation, histone modifications, and the presence of histone variants (Jenuwein 2002). Specific combinations of these marks are thought to determine the local chromatin structure that affects transcription and genome stability (Jenuwein 2002). In plants and mammals, DNA methylation is the beststudied epigenetic modification. Its faithful propagation is not only critical for proper development but also silences transposable elements (Finnegan 1996;Miura et al. 2001;Singer et al. 2001;Kato et al. 2003;Chan et al. 2005). Thus, apart from a role in development, DNA methylation protects genome integrity.Methylation modifies cytosines preceding guanines ( m CG) and in plants METHYLTRANSFERASE 1 (MET1), a homolog of the mammalian Dnmt1, is required for propagating m CG patterns during DNA replication (Finnegan 1996). Loss of MET1 leads to almost a complete erasure of CG methylation and indirect losses of plantspecific non-CG methylation . Loss of MET1 also results in the suppression of DNA demethylation activities, altered RNA directed de novo methylation, and the redistribution of other repressive marks such as histone H3 dimethylation in Lys 9 and trimethylation in Lys 27 (Soppe et al. 2002;Tariq et al. 2003;Mathieu et al. 2005Mathieu et al. , 2007, creating further epigenetic variation. Thus, transgenerational inheritance of the epigenome in plants is coordinated by the faithful replication of m CG patterns (Mathieu et al. 2007). Notably, loss of m CG results in hypomethylated epi-alleles that, at some loci, may be stably inherited over many generations (Kakutani et al. 1996;Mathieu et al. 2007;Vaughn et al. 2007). This is similar to inheritance of epigenetic states of ribosomal DNA following intercrossed Arabidopsis accessions Richards 2002, 2005;Woo and Richards 2008

Maintenance of CG methylation ((m)CG) patterns is essential for chromatin-mediated epigenetic regulation of transcription in plants and mammals. However, functional links between (m)CG and other epigenetic mechanisms in vivo remain obscure. Using successive generations of an Arabidopsis thaliana mutant deficient in maintaining (m)CG, we find that (m)CG loss triggers genome-wide activation of alternative epigenetic mechanisms. However, these mechanisms, which involve RNA-directed DNA methylation, inhibiting expression of DNA demethylases, and retargeting of histone H3K9 methylation, act in a stochastic and uncoordinated fashion. As a result, new and aberrant epigenetic patterns are progressively formed over several plant generations in the absence of (m)CG. Interestingly, the unconventional redistribution of epigenetic marks is necessary to "rescue" the loss of (m)CG, since mutant plants impaired in rescue activities are severely dwarfed and sterile. Our results provide evidence that (m)CG is a central coordinator of epigenetic memory that secures stable transgenerational inheritance in plants.

SummaryA large number of recent studies have demonstrated that many important aspects of plant development are regulated by heritable changes in gene expression that do not involve changes in DNA sequence. Rather, these regulatory mechanisms involve modifications of chromatin structure that affect the accessibility of target genes to regulatory factors that can control their expression. The central component of chromatin is the nucleosome, containing the highly conserved histone proteins that are known to be subject to a wide range of post-translational modifications, which act as recognition codes for the binding of chromatin-associated factors. In addition to these histone modifications, DNA methylation can also have a dramatic influence on gene expression. To accommodate the burgeoning interest of the plant science community in the epigenetic control of plant development, a series of methods used routinely in our laboratories have been compiled that can facilitate the characterization of putative chromatin-binding factors at the biochemical, molecular and cellular levels.