Heritable phenotypic differences caused by epigenetic modifications, rather than DNA sequence mutations, pose a challenge to our understanding of natural. Epialleles can lead to variations at the phenotypic and molecular levels, epigenetic variations might be involved in plant adaptive evolution. In plants, silent epialleles segregating in Mendelian fashion can be stably inherited over many .. () Isolating mechanisms, evolution and temperature.
|Published (Last):||10 February 2013|
|PDF File Size:||9.84 Mb|
|ePub File Size:||20.4 Mb|
|Price:||Free* [*Free Regsitration Required]|
Tetsuji Kakutani; Epi-Alleles in Plants: Epigenetic modification of plant gene and transposon activity, which correlates with their methylation, is often heritable over many generations. Such heritable properties allow conventional genetic linkage analysis to identify the sequences affected in epigenetic im. Machinery controlling the establishment of the epigenetic state and role of the epigenetic controls in plant development are also discussed.
Nucleotide sequence is not the only heritable information on the chromosome. Epigenetic information, which is based on DNA methylation or chromatin states, is also heritable during cell propagation. In both plants and mammals, DNA methylation correlates with epigenetic suppression of transcription. Mammalian epigenetic phenomena, such as parental imprinting and X-chromosome inactivation, are developmentally regulated, and re-programming of the epigenetic states occur in each generation.
In contrast, the epigenetic states of plant genes are often inherited over generations. I review here the epigenetic phenomena in plants, with special emphasis on the epigenetic inheritance over generations.
The recent accumulation of genome sequence information revealed that transposable elements and their derivatives are a major constituent of the genome of vertebrates and higher plants SanMiguel et al.
Considering the abundance of transposons in their genome, it is surprising that only a low proportion of spontaneous mutations is caused by them. Mechanisms may exist that suppress uncontrolled transposition of these elements. Most of the methylated cytosine is found in transposons and repeats in the mammalian genome. Some eukaryotic species with less genomic DNA methylation, such as Drosophilasuffer from a high frequency of transposon-mediated mutations compared with vertebrates and higher plants.
These findings led Yoder et al. The active or inactive states are often heritable over generations. An interesting feature of these systems is that modification of the activity in the transposons or their derivatives affect the activity of the nearby host genes reviewed by Martienssen aFedoroff Again, methylation correlates with the epigenetic state of these systems. Paramutation, another type of interesting epigenetic silencing of endogenous genes in maize, may also be derived from control of transposons Martienssen b.
Interestingly, modifier of paramutation 1 mop1 mutation of maize, which prevents paramutation at b1r1 and pl1 loci, also reverts methylation and silencing of the Mutator transposon Lisch et al.
Arabidopsis mutants with reduced DNA methylation provide powerful systems to directly investigate the role of DNA methylation. In the hypomethylation background induced by the ddm1 mutant, epiallleles silent repeated sequences are reactivated transcriptionally Jeddeloh et al.
Epialleles via DNA methylation: consequences for plant evolution.
Evoljtion example, a sequence called TSI or transcriptionally silent informationwhich is a part of the repeated epiapleles sequences Athila, is silent in wild-type plants, epia,leles transcribed in the ddm1 mutant Steimer et al. In addition to the transcriptional activation, the ddm1 mutation induces high frequency transposition of at least two classes of the endogenous Arabidopsis elements, MULE Mutator-like, which is similar to maize Mutator elements; Singer et al.
Both of these elements are not mobile in wild-type Columbia background. These observations suggest that Epialoeles methylation effectively suppresses transposon activity. Independent of these studies on epigenetic control of transposons, epigenetic control of gene expression has been extensively studied using transgenic plant systems reviewed by Matzke and MatzkeVaucheret et al. In short, transgene-silencing in many systems has been categorized into two types: TGS is heritable over generations, whereas PTGS is reset after meiosis and recurs in every evoultion at some stage of plant development.
The meiotically heritable property of TGS would be advantageous for defense of the genome against transposon, as the silenced state is maintained throughout development. Some of meiotically heritable epigenetic changes affect plant development. Mutations in the SUP gene alter floral pattern formation by eepialleles the floral whorl boundary Sakai et al.
Although results of fine mapping and complementation by SUP transgene indicate that clk s are allelic to other sup mutations, they do not have any change in the nucleotide sequence of SUP gene Jacobsen and Meyerowitz Instead, in the clk plants, the SUP gene was heavily methylated and transcriptionally silenced Jacobsen and Meyerowitz Thus clk s are epigenetically suppressed alleles of the SUP gene.
Epialleles in plant evolution.
The epigenetic state is heritable over generations and behaves like a real mutation, except that they revert to the wild-type allele at low frequency. The factors causing initial epigenetic change of the SUP gene in these alleles are unknown. Counter-intuitively, similar epigenetic silencing of the SUP gene was observed in met1 mutants and MET1 antisense lines. In Arabidopsismore examples of such epigenetic alleles affecting plant development have been identified using the DNA hypomethylation mutants and the linkage analysis.
Both ddm1 mutants and antisense MET1 lines exhibit a variety of developmental abnormalities Finnegan et al. Insertion mutation spialleles by activation epalleles endogenous transposon under the ddm1 mutant background is one of the epualleles for the developmental abnormalities Miura et al. However, it does not explain the high frequency occurrence of similar phenotypes in independent ddm1 lines and met1 lines Finnegan et al.
For example, late flowering traits are frequently observed in ddm1 inbred lines and in lines expressing MET1 antisense RNA, suggesting the underlying mechanism is non-random and possibly epigenetic Ronemus et al. This late flowering phenotype is inherited as a monogenic dominant trait mapped to a chromosomal region containing FWA Plnt FWA is one evolutiom the flowering-time loci previously found by conventional mutagenesis and linkage analysis Koornneef et al.
FWA gene was subsequently cloned by a map-based approach Soppe et al.
Both fwa-1 evolutikn fwa-2 mutants are semi-dominant Koornneef et al. Intragenic revertants with normal flowering time have been recovered from fwa-1 after mutagenization, again suggesting that the original fwa-1 mutation is a gain-function mutation Soppe et al. All the three revertant alleles of fwa-1 have putative loss of function mutation in the FWA gene, indicating that it is the responsible gene Soppe et al.
Plqnt, both fwa-1 and fwa-2 alleles do not show change in the nucleotide sequence compared with the wild-type allele. Instead, FWA gene is ectopically expressed in the fwa-1 and fwa-2 mutants Soppe et al. Thus, they are gain-of-function epigenetic alleles, analogous to the loss-of-function epigenetic alleles of the SUP gene. The ectopic expression epiallees the FWA gene is associated with hypomethylation of direct repeat around the transcriptional starting site Soppe et al.
The ectopic expression and hypomethylation of the repeats were also observed in the late flowering ddm1 and ddm2 met1 lines, suggesting that the hypomethylation can generate the meiotically heritable epigenetic mark responsible plaant the late-flowering phenotype Soppe et al.
Another interesting example of evolutjon abnormalities induced under the ddm1 background is a dwarf phenotype called ball bal. This phenotype is also inherited as a monogenic Mendelian trait Kakutani et al. Over-expression of R gene seems to be responsible for the bal phenotype, as constitutive expression of the R gene At4g in transgenics mimics the phenotype Stokes et al.
Interestingly, the epigenetic over-expression state of the BAL locus is heritable epiallelee metastable: In short, epi-alleles of both FWA and BAL genes result from over-expression with loss of silencing in the repeated sequences. Is such inheritance of the epigenetic state over generations unique to plants?
Stable chromosomal inheritance of the epigenetic state during mitosis and meiosis has also been found in fission yeast Grewal and Klaralthough genomic DNA methylation has not been found in this organism.
In the fission yeast systems, association of heterochromatin protein, rather than DNA modification, may be responsible for the epigenetic epialeles Nakayama et al.
Moreover, inheritance of the epigenetic state was even been found in mouse: The stable inheritance of the epigenetically suppressed state over generations might be an evolutionary prototype of epigenetic control.
If so, mammals may modify it to control development through control of the DNA methylation pattern. Consistent with the idea epalleles developmental control of DNA methylation evolved relatively late in evolution, a change in the DNA methylation pattern has not been found in zebra fish Macleod et al.
Methylation in the majority of Arabidopsis genomic sequences seems to be controlled in a similar way. All these sequences remained hypomethylated even after out-crossing to the DDM1 wild-type background Kakutani et al.
The stable inheritance of the DNA methylation pattern over generations raises the question how methylation patterns are initially established.
In other words, why are some sequences sites methylated while others not?
Very intriguingly, Mette et al. A similar phenomenon has previously been found in viroid-infected transgenic plants Wassenegger et al. These phenomena, called RNA-directed DNA methylation RdDMcould be induced in various sequences and generally occurs at all cytosines including non-symmetrical sites Wassenegger et al. Although global de novo methylation comparable to that during mammalian development has not been found in Arabidopsisits genome has copies of genes structurally similar to mammalian de novo DNA methyltransferase DNMT3s Okano et al.
In addition, after introduction of the inverted repeat copy by transformation or genetic hybridization between ecotypes, the unmethylated copy was methylated de novo Luff et al. Chromomethylases are structurally related to other DNA methyltransferases but contain additional chromodomain motifs Henikoff and ComaiPapa et al.
Biochemically, it is not known whether CMT3 has maintenance or de novo methylation activity. However, the CMT3 product may be a component of machinery causing de novo methylation at non CpG sites, because transformation of the cmt3 mutant with the wild-type CMT3 gene results in re-methylation of the PAI sequence Bartee et al.
In addition to the CMT3 gene, Jackson et al. The methylation of histone H3 is associated with heterochromatin formation in many eukaryotic systems.
The involvement of histone methylation on DNA methylation has also been found in Neurospora. The methylation of histone H3 is necessary for stable heterochromatin formation in many eukaryotic systems, and DNA methylation may function to further stabilize the silent state.
InFire et al. Since then, genetic and biochemical studies in C. Interestingly, major targets of methylation by CMT3 are transposons and RNAi-based process may be involved in target recognition Tompa et al.
In mammalian epigenetic systems, establishment of gene silencing is often accompanied by production of non-coding RNA, such as Xist and Air Avner and HeardReik and WalterSleutels et al. PTGS machinery may be involved in plant development.
Identification of direct targets of PTGS involved in these developmental processes would be an important breakthrough. By direct cloning procedures, many additional members of microRNA have been cloned in C. Application of such an approach in plants might lead to identification of targets of PTGS machinery controlling plant development. TGS in plants is often so stable that it is inherited over generations.
In addition, separation of somatic and germ cells is not as clear in plants as it is in animals: Provided that the DNA methylation pattern is inherited by the next generation, irreversible change of the methylation in the apical meristem may not be a good strategy for controlling development.
Instead, control of development by DNA methylation might be possible in terminally differentiated tissues which do not contribute to the next generation; an obvious example is endosperm.
This review does not cover another type of transcriptional gene silencing mediated by polycomb proteins. This type of chromatin silencing, which is reset in each generation, is important for many developmental processes including flowering and endosperm formation reviewed by Preussrecent advances in Gendall et al.
The possible connection between silencing mediated by polycomb proteins and DNA methylation is another challenging field Finnegan et al. Important papers have recently appeared on plant small RNAs Llave et al.