Inheritance of DNA Methylation in Plant Genome

Genomic DNA contains not only information of DNA sequence, but also epigenetic information that is the direct DNA modification by methylation (the addition of methyl group to the 5th carbon of pyrimidine ring of cytosine) and histone modifications (acetylation, methylation, etc). Epigenetic information is closely related to regulation of gene expression. If a methyl group is dislocated to position 5 of the pyrimidine ring of cytosine, the hydrogen bond between complementary GC bases will not be inhibited, but this methyl group is positioned so as to be exposed in the major groove of the double-helix structure of DNA, and according to the genome region/sequence undergoing modification of methylation, gene expression is inhibited by the interaction between the genome and DNAbinding molecules.


The role of DNA methylation in mammals and plants
Genomic DNA contains not only information of DNA sequence, but also epigenetic information that is the direct DNA modification by methylation (the addition of methyl group to the 5th carbon of pyrimidine ring of cytosine) and histone modifications (acetylation, methylation, etc). Epigenetic information is closely related to regulation of gene expression. If a methyl group is dislocated to position 5 of the pyrimidine ring of cytosine, the hydrogen bond between complementary GC bases will not be inhibited, but this methyl group is positioned so as to be exposed in the major groove of the double-helix structure of DNA, and according to the genome region/sequence undergoing modification of methylation, gene expression is inhibited by the interaction between the genome and DNAbinding molecules.
Methylated cytosine is very common in plant and mammalian genomes, and plays an important role in the regulation of many cellular processes including X inactivation, chromosome stability, chromatin structure, embryonic development and transcription. DNA methylation in most mammals occurs at cytosine on the CG sequence, a 2-base sequence lined up in the order of cytosine-guanine. In plants, on the other hand, methylation of a non-CG sequence (CNG and CHH, where N is A, G, C or T, and H is A, C or T) frequently exists in addition to methylation of the CG sequence. Moreover, there is a large difference between mammals and plants in methylation dynamics throughout the life cycle [Law & Jacobsen 2010]. In mammals, methylation patterns change dramatically during gametogenesis and early development [Monk et al. 1987, andTada et al. 1997]. In mice, the methylation level of the genome decreases after fertilization to the lowest level at implantation, which is rapidly induced at the time of tissue differentiation after implantation. In reproductive cells, moreover, methylation of genomic DNA is eliminated once and then methylation occurs again, but the sex-specific methylation pattern is written according to the gene and is adjusted so that a specific uniparental allele may be expressed. The gene showing sexspecific expression is called the imprinted gene and plays an important role in development and differentiation [Obata et al. 1998, Ueda et al. 2000, and Davis et al. 2000].
In contrast, generational changes in methylation status and inheritance in plants have been unclear. The methylation statuses of some genes are stably inherited [Bender et al. 1995, Jacobsen et al. 1997, Kakutani et al. 1999, but, recent studies show that DNA methylation patterns are altered in F1 hybrids from interspecific or intraspecific crossing [Matzke et al. 1999, Wendel 2000, Shaked et al. 2001, Pikaard 2001, Madlung et al. 2002, Comai et al. 2003, Liu et al. 2004, Dong et al. 2006, and Akimoto et al. 2007]. These alterations might be caused by the interaction among alleles, and/or the change of epigenetic regulation. For example, imprinted genes that have sex-dependent methylation patterns in endosperm have been identified in plants, and they play an important role in the control of flowering or seed development [Grossniklaus et al. 1998, Kiyosue et al. 1999, Kinoshita et al. 1999, Choi et al. 2002, Xiao et al 2003, Kohler et al. 2003, Scott et al. 2004, Kohler et al. 2005, Baroux et al. 2006, and Gehring et al. 2006]. Another example, paramutation is an allele-dependent transfer of epigenetic information, which results in the heritable silencing of one allele by another [Brink 1956, andCoe 1959]. In recently, Shiba et al. (2006) suggested that tissue-specific monoallelic de novo methylation in F1 involved in determining the dominance interactions that determine the cruciferous self-incompatibility phenotype. These analyses of DNA methylation inheritance may help to identify new important genes, such as imprinted gene, and to further clarify the biological significance of DNA methylation.

Concept of RLGS as a tool for DNA methylation analysis
Restriction Landmark Genome Scanning (RLGS) [Hayashizaki et al. 1993] is a unique quantitative approach well suited for simultaneous assay of methylation status [Costello et al., 2002]. The other genome wide methylation analyze methods viz., Tiling microarrays ] and methylation-sensitive amplification polymorphism (MSAP) [Reyna-Lopez et al. 1997, Xiong et al. 1999 are the comprehensive or easily applied method, respectively, but RLGS excels these methods because of its reduced cost and short span of experimental time (3 days). In RLGS, intensity of the RLGS spots directly reveals the methylation level and partial methylation such as imprinting, whereas in other methods the methylation levels are not inferred directly from the results. These important advantages of RLGS rank it as an appropriate method for genome wide methylation survey. This method had been used for development of genetic linkage maps [Okuizumi et al. 1995A, Okuizumi et al. 1995B], methylation analysis in tumor tissue , Miwa et al. 1995, Wang et al. 2009], and identification of imprinted genes in mammals [Hayashizaki et al. 1994A, Shibata et al. 1995, Plass et al. 1996] and based on which several interesting research in epigenetics and genetics such as alteration of genomic DNA methylation [Takamiya et al. 2008B, Takamiya et al. 2009] and genetic diversity study [Okuizumi et al. 2010] had been carried out. The "in silico RLGS", is a software originally developed and named by us, and now the name and the concept are spread to other researchers [Matsuyama 2008]. This software can be utilized for the RLGS analysis of the organisms for those the whole genome sequence is available. One among such organism is Rice which enables the utilization of "in silico RLGS" analysis [Takamiya et al. 2006] because of the availability of its whole genome sequence [International Rice Genome Sequencing Project, 2005]. 2. Methodology of RLGS to detect DNA methylated sites efficiently in the plant genome 2.1 Method development to use isoschizomer restriction enzyme such as MspI and HpaII RLGS is a high-speed genome scanning system. It employs direct end-labeling of the genomic DNA digested with rare-cutting restriction enzyme such as NotI, followed by highresolution two-dimensional (2-D) electrophoresis. Thousands of loci with high reproducibility can be detected as spots on the 2-D pattern in this method. A lot of methylated sites can be analyzed using the conventional RLGS method because the recognition site (GCGGCCGC) of the first cutter NotI [Hayashizaki et al. 1993, Watanabe et al. 1995, which is a methylation sensitive restriction enzyme (Fig. 1A), is often located in the CpG islands [Bird 1992]. But, this conventional method has bottle-neck in distinguishing methylation polymorphism from sequence polymorphism.
This bottle-neck of the conventional RLGS urges for the development of improved RLGS method (Figs. 1A and1B). In this improved method, isoschizomer restriction enzymes (MspI and HpaII) were used for the direct detection of methylated sites, that we produce 2 patterns; the "[MspI] pattern" employs the restriction enzyme combination of NotI-MspI-BamHI as the 1 st -2 nd -3 rd cutter, and another "[HpaII] pattern" uses NotI-HpaII-BamHI as the 1 st -2 nd -3 rd cutter. Rationale to utilize the isoschizomer is, they recognize the same sequence (CCGG) but with difference in methylation sensitivity. For example, the MspI cleaves the CCGG at 2 nd C which is methylated (C 5m CGG) but HpaII doesn't cleave because of its differential sensitivity for the methylation. The different methylation sensitivity between MspI and HpaII is reflected in difference of RLGS spot patterns ( Fig 1B). Briefly, the spot which is only detected in either [MspI] or [HpaII] pattern show the MspI/HpaII site of the RLGS spot is methylated. We called this spot as "methylated spot." The additional qualities of this improved RLGS method were (1) easy identification of methylated sites and their location in genome, (2) methylation of coding regions are surveyed efficiently and (3) its ability to scan methylation sites in an individual or a tissue. Furthermore, in an unsequenced species and even in a cloned organism, this improved method distinguishes DNA mutation from the methylation changes.
Genomic DNA was isolated from 8-week old leaf blades and leaf sheath of Nipponbare, Kasalath, and F1 plants (Crossing subsp. japonica cv. Nipponbare as the seed parent with subsp. indica cv. Kasalath as the pollen parent gave F 1 hybrids designated NKF 1 ) grown for 8 weeks. Isolated genomic DNA (0.2 µg) was treated with 2 U of DNA polymerase I (NIPPON GENE, Tokyo, Japan) in 10 µL of blocking buffer (10 mM Tris-HCl, pH 7.4, 10 mM MgCl 2 , 1 mM dithiothreitol (DTT), 0.4 µM dGTP, 0.2 µM dCTP, 0.4 µM ddATP and 0.4 µM ddTTP) at 37°C for 20 min. Thereafter, the polymerase I was inactivated by incubating at 65 ˚C for 30 min. Then the DNA was digested with 20 U of NotI (NEB, Beverly, MA, USA) in volume of 20 µL. The cleavage ends were filled in and labeled with 32 P in the presence of 1.3 U of Sequenase ver. 2.0 TM (USB, Cleveland, OH, USA), 0.33 µM [α-32 P] dGTP (3,000 Ci/mmol), 0.33 µM [α-32 P] dCTP (6,000 Ci/mmol) and 1.3 mM DTT at 37 ˚C for 30 min in 22.5 µL. Next, www.intechopen.com DNA Methylation -From Genomics to Technology 72 this reaction mixture was incubated at 65 ˚C for 30 min to inactivate the enzyme. The processed sample was divided into two tubes. One was digested with 25 U of MspI (Toyobo, Tokyo, Japan) and the other was treated with 25 U of HpaII (Toyobo) and incubated at 37 o C for 1 h. Each sample was fractionated on an agarose disc gel (0.8% Seakem GTG TM agarose, FMC Bioproducts, Rockland, Maine, USA) in the 2.4 mm diameter  63 cm long tube, and then electrophoresed in the 1 st -dimensional (1-D) buffer (0.1 M Tris-acetate, pH 8.0, 40 mM sodium acetate, 3 mM EDTA, pH 8.0, 36 mM NaCl) at 100 V for 1 h followed by 230 V for 23 h. The size fractionated genomic DNA was carefully extruded from the tube and soaked for 30 min in the reaction buffer for BamHI. Thereafter, DNA was digested in the gel with 1500 U of BamHI for 2 h. The gel was fused into the top edge of a 50 cm (W)  50 cm (H)  0.1 cm (thickness) 5% polyacrylamide vertical gel by adding melted agarose (0.8% at 60-65˚C) to connect each gel. The 2nd-dimensional (2-D) electrophoresis was carried out in Tris borate EDTA (TBE) buffer (50 mM Tris, 62 mM boric acid, 1 mM EDTA), at 100 V for 1 h followed by 150 V for 23 h. An area 35 cm  41 cm of the original gel was excised and dried. Autoradiography was performed for 3-10 days on a film (XAR-5; Kodak, Rochester, NY, USA) at -80 ˚C using an intensifying screen (Quanta III; Sigma-Aldrich, St. Louis, MO USA), or for 1-3 days on an imaging plate (Fuji Photo Film, Tokyo, Japan). Finally, the imaging plate was analyzed by the BAS-2000 TM (Fuji Photo Film). The conventional RLGS method is a tool that uses landmarks to directly label restriction enzyme sites scattered on genomic DNA, expand them on a single image by high-resolution two dimensional electrophoresis, and detect several thousands of spots at once. In the conventional method, the differences of methylation status among samples are surveyed using NotI methylation sensitivity. However, it has been difficult to distinguish methylation polymorphism from sequence polymorphism, and can't adopt for methylation surveillance of one sample. In improved method (left flow chart), isoschizomers (MspI and HpaII) are used as second cutter. We are possible to detect methylated sites directly in even if an individual or a tissue by comparison between [MspI]

In silico RLGS analysis oriented by whole genome sequence
A software named "in silico RLGS" was developed that simulates RLGS spots based on known sequence data to identify each spot in actual RLGS. The software searches for restriction enzyme sites in the entire genome sequence that were used in an actual RLGS experiment, and calculates fragment length between the restriction sites and mobility for simulating a 2-D spot pattern. In our experiment the whole rice genome sequence data was obtained from http://rgp.dna.affrc.go.jp/J/IRGSP/Build3/build3.html. NotI sites and MspI sites that were near to the NotI sites were searched through the whole sequence and given ID numbers. Length of identified fragments between restriction sites (from NotI end to the nearest MspI end or from NotI end to next NotI end) and mobility of each DNA fragment in the 1-D electrophoresis were calculated by the software according to Southern E. (1979). The exact mobility of RLGS spots were confirmed from electrophoresis of λ DNA fragments with known sequences. In addition, the BamHI sites were also surveyed and DNA fragment (from NotI end to the nearest BamHI end) length in 2-D was estimated. Based on this, a 2-D graph (in silico RLGS pattern) was drawn. The in silico RLGS pattern was compared to its corresponding autoradiographic (actual) RLGS pattern with relative spot positions. This comparison leads to the identification each RLGS spot immediatelyand precisely.

Detection of methylated spots in F1 hybrid and the parents
To detect the DNA methylation in plant genome, improved RLGS and in silico RLGS were used [Takamiya et al. 2008A]. Two rice individuals, Oryza sativa L. var Nipponbare and O. sativa L. var Kasalath as parent and its F1s were used as the experimental material to analyze the pattern of the DNA methylation and its inheritance in the F1 hybrid. F1 were obtained from crossing Nipponbare as the seed parent with Kasalath as the pollen parent. In F1, though most of the spots followed the Mendelian law of inheritance, still 8 parental spots (spot numbers: 23, 65, 200, 231, 235, 323, 501 and 525) disappeared and one new spot (f2) was detected. These 9 spots (8 disappeared spots and 1 new spot) indicate the abnormal inheritance and the details were summarized in Table 1. The pattern of spot 323 in parents, F1 and selfed progenies were shown Figure 3c.

Mapping of methylation status of rice genome using RLGS
The mapping of methylation status of rice genome using RLGS was carried out as following steps.
(1) Methylated spot was detected by comparison [MspI] and [HpaII] patterns, and its locus was identified by in silico RLGS. (2) The presence of restriction site of the NotI and the MspI/HpaII of methylated spot was confirmed using PCR analysis. (3)The methylation status of restriction sites of methylated spot was confirmed by another PCR analysis. (4) The parental origin of the methylated alleles in F1 was determined by using the sequence analysis. These steps are explained in detail as follows.
The presence of the restriction sites of NotI and MspI/HpaII was confirmed by PCR analysis by designing the flanking primers for the restriction sites of RLGS spots that were identified by in silico RLGS or spot cloning. The presence of restriction sites was confirmed in the RLGS spot 97 of Nipponbare and Kasalath (Fig. 4) as an example. The Figure 4A shows  www.intechopen.com

79
To confirm the methylation status of spots with restriction enzyme site in Nipponbare and Kasalath, different PCR products were obtained (Fig. 4B). The PCR product loaded in lane 1 genomic DNA of Nipponbare and the lane 2 and 3 were loaded with the product obtained by using MspI or HpaII treated genomic DNA of Nipponbare as template, respectively. The lanes 4 to 6 are loaded with the PCR products of Kasalath obtained in similar manner to that of Nipponbare. The lane 1 and 4 are used as the positive control for Nipponbare and Kasalath, respectively. The lane 2 and 5 do not have any bands where as the lane 3 and 6 had the band sizes similar to that of their positive controls. This result showed that the MspI/HpaII site was methylated (C 5m CGG), and it was correspondence with RLGS result. Similar to the parents, the MspI/HpaII site was methylated in F1 (Fig. 4C).
In Nipponbare, the methylation status of 90 NotI and 92 MspI/HpaII sites were checked. Out of 182 sites, 60 sites (33%) were methylated and 4 Nipponbare specific sites were identified. Similarly in Kasalath, 82 NotI and 84 MspI/HpaII sites were tested for methylation status resulted in identification of 59 methylation sites (36%) and 10 Kasalath specific sites.
Finally, to determine the parental origin of the methylated allele, the bands were subjected for sequence analysis. The bands of Nipponbare in lane 1 and bands of Kasalath in lane 4 ( Fig 4B) were purified and sequence analysis was done. This analysis detects an SNP for spot 97 showing C in Nipponbare ( Fig 4D) and T in Kasalath (Fig 4E). Similarly the band in lane 3 (Fig. 4C), which was amplified from HpaII digested genomic DNA of F1, was sequenced and it was found to be heterozygous with both C and T (Fig. 4F). The sequence analysis shows that methylation prevented digestion of the MspI/HpaII sites of both the alleles, and the methylation status of the parents was inherited to F1 following Mendelian law. Similarly, we examined the other 25 methylated spots that showed the same appearance or absence between the parents and F1, and confirmed that all methylation status were inherited to F1.
Some of the NotI and MspI/HpaII sites that were specifically found in the in silico RLGS were also checked based on the PCR and sequence analysis. This analysis identified some new altered methylations in F1. The identified spots were demethylations at NotI of spot 52 (Chr.11) and the MspI/HpaII site of spot 105 (Chr.3) in F1, and their details are given in the Table 1 and Figure 5. The specific occurrence of the spots only in in silico RLGS is due to the methylation in NotI sites.
In the entire process of analysis, a total of 103 RLGS spots were identified and most of these spots were analyzed for methylation status by RLGS and PCR. The result of this analysis was summarized in methylation map (Fig 5). In the map, the numbered spot depict that those spots were methylated at one or more NotI or MspI/HpaII sites of Nipponbare, Kasalath or F1. In total, seven altered spots were mapped (shown as 'AI' in Fig 5). The other spots viz., 23, 501, 525 and f2 are not yet to cloned. Out of this 103 NotI sites, 17 and 14 sites were located within 2.0 kb upstream and downstream of a gene respectively, and 63 sites located within the gene. The remaining 9 sites were in the intergenic regions. Similarly, 25 MspI/HpaII sites were in 5' upstream region, 48 within gene and 15 sites in 3' downstream region. Thus, 182 sites (88%) out of 206 were located between 2.0 kb upstream and downstream of genes.

Non-Mendelian inheritance of DNA methylation
To prove non-Mendelian inheritance of DNA methylation in altered spot loci, we analyzed reciprocal F 1 hybrids (subsp. japonica cv. Nipponbare × subsp. indica cv. Kasalath) of rice (Oryza sativa L.). Reciprocal hybrids were produced by crossing the same individual of each cultivar as the female parent on one culm and as the male parent on another culm. Crossing Nipponbare as the seed parent with Kasalath as the pollen parent gave F 1 hybrids designated NKF 1 . The inverse cross gave KNF 1 hybrids. The seeds of Nipponbare, Kasalath, NKF 1 (nine individuals from the same parents), and KNF 1 (nine individuals from the same parents), and the selfed progeny of the parents were grown for 2 months, and the genomic DNA was isolated from the leaf blade and sheath of each individual, and the RLGS analysis was performed.  4, 7, 10, 13, 16, 19, 22, 25, 28, 31, 34, 37, 40, 43, 46, 49, and  Consider spot 323 of Nipponbare as an example which is a HpaII specific spot but not detected in the F1. In most of the F1, the demethylation of the M1/M2 site of spot 323 was detected based on the PCR analysis (Fig 6 shown the result of M1 site). The PCR product of Kasalath genomic DNA was not amplified by the primer P1 and R1. This non-amplification of these regions by the primers may be due to some difference in base sequence between Nipponbare and Kasalath. The spots 200, 231, 235, 501 and 525 were specific to Nipponbare or Kasalath and these spots do not appeared in F1. Therefore, the regions of these spots may have some differences in DNA sequence or methylation status between Nipponbare and Kasalath. Alternatively, these polymorphic regions might have altered methylation status in the F1 hybrid [Takamiya et al. 2009]. The altered methylation of spot 323 suggests possible sequencedependent demethylation, for example as a result of paramutation induced by allelic exclusion (Chandler et al., 2000). For better understanding of the methylation behavior of spot 323, detailed analyses like expression analysis of the gene near to the spot may be required.
The spot 200 was detected in both the [MspI] and [HpaII] patterns of Nipponbare at a diminished spot intensity but disappeared in Kasalath (Fig 7). The spot is another non-Mendelian example [Takamiya et al., 2009]. In RLGS analysis halved/diminished spot intensity indicates the heterozygote which was confirmed theoretically and practically in earlier studies (Hayashizaki et al. 1994B., Okuizumi et al. 1997. The DNA fragment of spot 200 was cloned and sequence analysis was carried out, which place it in the 5' region of a non-protein coding transcript (Os11g0417300) (Fig 8A). The spot position of autoradiographic RLGS pattern of Nipponbare was compared with in silico RLGS pattern. Then, the DNA fragments digested at NotI (N) and MspI (M) was separated by 1-D electrophoresis, and the N and BamHI (B) sites were fractionate by 2-D electrophoresis ( Fig  8A). Restriction enzyme digestion and sequencing was employed to confirm the existence of N, M and B in the parental Nipponbare, whereas in Kasalath the N and M site exist but the B site was absent. The presence or absence of spot 200 obtained based on the RLGS analysis of NKF 1 and KNF 1 shows that this spot segregated as 1:1 in both the population (Fig 7 and Table 2). The diminished spot intensity and its segregation in F1 hybrids show that the MspI/HpaII site of spot 200 is methylated heterozygously. Similarly, it was assumed that the detection and absence of spot 200 in F1 was due to the non-methylated and methylated M site, respectively. Moreover in all selfed progenies (9 individuals) of Nipponbare, the spot 200 was detected with halved intensity (Fig 7, Table 2). From this, it was concluded that in the selfed progeny and in parental Nipponbare, the M site was methylated heterozygous because of non-Mendelian inheritance of methylation.  Expression analysis of the non-protein coding transcript (Os11g0417300), which is the nearest gene to the MspI/HpaII site of spot 200 (Fig 8A) [Takamiya et al., 2009], was done to clarify whether the methylation status in correlated with expression of nearest gene. The cDNA (Genebank accessions No: AK 109537) of the non protein coding transcript expressed in flower, leaf and panicle (http://www.ncbi.nlm.nih.gov/sites/entrez?db=unigene& cmd=search&term=AK109537). Total RNA was isolated from the leaf blade and sheath of the parental Nipponbare, Kasalath, two individuals each in NKF 1 (NK5 and NK7) and KNF 1 (KN5 and KN10) hybrids. The pattern of NK5 and KN5 alone detected the spot 200 whereas in the pattern NK7 and KN10 the spot was not detected. The cDNAs' of both NK5 and KN5 were PCR amplified and separated by agarose gel electrophoresis (Fig 8B). The gene was expressed in the parents, NK5, and KN5 (Fig. 8B). The single nucleotide polymorphism (C/T) between Nipponbare and Kasalath was found and the RT-PCR products were sequence to reveal the parental origin of the expressed sequence in F 1 hybrid ( Fig 8C). The RT-PCR products of NK5 and KN5 which had spot 200 in their RLGS pattern were subjected to sequence analysis. Allelic expression bias for the Nipponbare allele was found in NK5 and KN5 (Fig 8C). Similarly, the bias for Nipponbare allele was found in KN7 and NK10 which do not have spot 200 (data not shown). Strong allele bias was found in the reciprocal hybrids and this shows the monoallelic expression of the Nipponbare allele. In addition, a splicing variant (smaller transcript with lower expression than Nipponbare allele) specific to Kasalath was detected (Fig 8B), but this was absent in NKF 1 and KNF 1 . This transcript was sequenced and a splicing variant that leads to a 76 bp deletion at the 3' end of exon 2 was revealed. The reason is unknown, but, it implies that some effect of methylation is influencing the variant expression. The spot 231, which behaved similar to that of non-Mendelian spot 200, had diminished spot intensity and 1:1 segregation in NKF 1 and KNF 1 (Table 2), and also it was detected in all the selfed progeny of Nipponbare. For this spot 231, the expression of the nearest gene (DUF 295 family protein Os01g0327900) was analyzed in two NKF 1 (NK5 and NK7) and two KNF 1 (KN5 and KN10) individuals. Sequence analysis of the RT-PCR products showed that only the Kasalath allele was expressed in NK5, NK7, KN5, and KN10 ( Fig 8D shown   . Besides, 17 out of 30 genes analyzed showed >1.5-fold expression bias for one of two alleles, with monoallelic expression of one gene in Populus interspecific hybrids [Zhuang et al. 2007]. Therefore, it is considered that allelic expression bias was caused by epigenetic status of DNA methylation and/or histone modification. In plants, identification of more RLGS spots exhibiting non-Mendelian inheritance and simultaneously studying their methylation status of the corresponding DNA sequence and their expressed allele will explain the importance and better understanding of gene regulatory mechanisms such as monoallelic expression. Extensive and detailed expression analysis of genes in F1s with different genetic backgrounds is very much essential because the findings can be applied to other genes as well. The mechanism of allelic exclusion inducing heterosis, hybrid weakness and genome barriers might be better understood by revealing the regulation and function of the splicing variant of Kasalath (Fig. 8B) in F 1 hybrids.

Various aspects on DNA methylation roles
The RLGS method is very powerful for methylation analysis other than that for genetic analysis of DNA methylation. In the Takamiya et al. 2006, the methylated status was compared among 3 ecotypes of Arabidopsis using the RLGS method. Methylation at a total of 17 sites (NotI: 9 sites and HpaII: 8 sites) was detected in the 3 ecotypes. Among them, there were 8 common methylation sites among the 3 ecotypes, and the 9 residual sites (53%) showed methylation polymorphism. Among all restriction enzyme sites analyzed (37 nonmethylated sites and 17 methylated sites), the sites showing a different methylated status among the 3 ecotypes accounted for 17% of the total sites (9/54). In the studies so far, it has been reported that methylation is involved in inactivation and metastasis inhibition of transposon and retrotransposon [Hirochika et al. 2000]. In the pseudogene and gypsy-like retrotransposon family-like region in centromeres, both the NotI site and MspI/HpaII site were methylated in the 3 ecotypes, which may suggest the relation between the pseudogene in the centromere region and inactivation of the movable element.
Moreover, the RLGS method can be applied to the detection of genomic variation in plant tissue culture [Takamiya et al. 2008B]. The genome DNA of 2 ramets obtained from one seed of rice was extracted and analyzed for RLGS with the combination of NotI-HpaII-BamHI and with NotI-MspI-BamHI to compare their pattern. As a result, 10 different spots (6%) were detected between ramets. One spot among the 10 different spot was cloned to confirm the methylated status by PCR, it was found that the methylated status at the restriction enzyme (HpaII) site was different between the 2 ramets.
In our recent RLGS analysis, we tried to detect the tissue-specific methylated status by conducting RLGS analysis of endosperm, embryo, leaf blade and leaf sheath and comparing the methylated status of each. As a result, there were 35 shared methylated spots in 3 defferent tissues among 58 methylated spots in total (Fig. 9). Fifty-six methylated spots detected in endosperm and embryo, including 19 tissue-specific spots. That is, a 34% difference was observed between endosperm and embryo. Next, 56 methylated spots were detected in endosperm and the leaf blade/sheath, including 20 tissue-spcific spots which showed a 36% difference between endosperm and leaf blade/sheath. Next, 49 methylated spots were detected in embryo and leaf blade/sheath, including 7 tissue-specific spots. That is, a 14% difference was observed between embryo and leaf blade/sheath. The result of comparison between endosperm and embryo (34%) showed a 2.4-fold larger difference with the comparison between embryo and leaf blade/sheath (14%), and the result of comparison between endosperm and blade/sheath (36%) showed a 2.6-fold larger difference than the comparison between embryo and leaf blade/sheath (14%). The leaf blade/sheath is differentiated from the embryo, but because the endosperm and embryo are independent tissues and have different functions and gene expression, the difference in methylated status is also considered large. Two recent studies show that endosperm DNA methylation is reduced genome-wide, and this reduction is likely to originate from demethylation in the central cell nucleus of the female gametophyte [Hsieh et al. 2009, andGehring et al. 2009]. Among the spots showing tissue-specific methylated status, expression analysis of spot 226 was conducted by RT-PCR. The MspI/HpaII site of spot 226 was positioned in the 5' region of the Chr. 5 gene Zn-finger, C-x8-C-x5-C-x3-H type domain-containing protein (Os05g0497500) (Fig. 10A). When RT-PCR analysis (primer sets: rtpcr226-F, 5'-CTGGTGGAGATATGAAGAACAA-3'; rtpcr226-R: 5'-TATGTTTAACAACGGGATGTGT-3') This gene was not expressed in endosperm before water absorption, but it was expressed in embryos and leaf blades/sheaths, and weakly in endosperm after water absorption. 18S rRNA is a control. (C) In regard to the MspI/HpaII site of spot 226, we analyzed the methylation status in the endosperm before and after water absorption and in embryos and leaf blades/sheaths by the PCR method. We used MboI-treated DNA as a positive control. The MspI/HpaII site of spot 226 was methylated in embryos and leaf blades/sheaths. These figures were cited from Hosobuchi, S. 2007.
of this gene was conducted, this gene was expressed strongly in the embryo and leaf blade /sheath in which methylation was detected at the MspI/HpaII site (Fig. 10B) by RLGS analysis. Moreover, this gene never expressed in the endosperm before absorption of water, but weak expression was confirmed in the endosperm after absorption of water for 10 minutes (Fig. 10B). Next, when the methylated status at the MspI/HpaII site was analyzed by the PCR method, methylation was confirmed in the embryo and leaf blade/sheath (Fig.  10C). When comparing the endosperm before and after absorption of water, more PCR products were amplified in the endosperm after absorption of water ( Fig. 10C. Comparison between Lanes 4 and 7). In the future, quantitative analysis will be required, but from the present results, it is assumed that the partial methylation rate of the MspI/HpaII site is higher in the endosperm after absorption of water than in the endosperm before the absorption of water. That is, there is the possibility that induction of DNA methylation and gene expression may begin within as very a short time as 10 minutes. When the 5' region is methylated, gene expression is usually inhibited, but our results are inconsistent with the general role. However, because the methylation analysis was conducted only at the MspI/HpaII site of spot 226, it is necessary to analyze the DNA methylated status in the promoter region of this gene widely and quantitatively.

Conclusion
RLGS is useful for genome wide surveillance of epigenetic alterations which effects to gene regulation and unknown phenomena with DNA methylation, because RLGS is suitable for exploratory studies on account of its low cost, and short set-up time. It can analyze any unsequenced living things. As mentioned above, RLGS analysis of reciprocal hybrids in rice provided new interesting observations, and this strategy will apply to study of mammalian epigenetics.