Inheritance and alteration of genome methylation in F1 hybrid rice

Research Article

Inheritance and alteration of genome methylation in F1hybrid rice

We analyzed the inheritance of DNA methylation in the ?rst ?lial generation(F1)hybrid of Oryza sativa L.(‘‘Nipponbare’’?‘‘Kasalath’’)by restriction landmark genome scan-ning (RLGS).Most parental RLGS spots were found in the F1,but eight spots (4%)showed abnormal inheritance:seven of the eight spots were missing in the F1,and one was newly detected in the F1.Here we show demethylation at restriction enzyme sites in the F1.We also found a candidate site of stable heterozygous methylation in the genome.These results show the applicability of the RLGS method for analysis of the inheritance and alteration of methylation in F1hybrid plants.

Keywords:

DNA methylation /F1hybrid /Inheritance /Oryza sativa /Restriction landmark genome scanning DOI 10.1002/elps.200700784

1Introduction

Genomic DNA contains not only genetic information but also epigenetic information such as DNA methylation (cytosine methylation)and histone modi?cations.Methy-lated cytosine is very common in plant and mammalian genomes and plays an important role in the regulation of gene expression.In mammals,methylation patterns change dramatically during gametogenesis and early development [1,2].In contrast,generational changes in methylation status and inheritance in plants have been unclear.The methylation statuses of some genes are stably inherited [3–5],but,recent studies show that methylation patterns are altered in ?rst ?lial generation (F1)hybrids,by introgres-sion,and in allopolyploids [6–14].Imprinted genes that have

sex-dependent methylation patterns in endosperm have been identi?ed,and they play an important role in the control of seed formation or ?owering [15–24].The precise analysis of methylation inheritance may help to identify new imprinted genes and to further clarify the biological signi?cance of methylation.

Tiling microarrays [25],restriction landmark genome scanning (RLGS)[26,27],and methylation-sensitive ampli-?cation polymorphism [28,29]are three major methods for genome-wide methylation analysis.Tiling microarrays are the most comprehensive tools,but are not suitable for exploratory studies on account of their high cost and long set-up time.Methylation-sensitive ampli?cation poly-morphism is the most easily applied of the three methods,but because it uses some PCR ampli?cation steps,the methylation level is not re?ected directly in its results.Although RLGS analyzes a more limited region of the genome than tiling microarrays,it can be done at much lower cost and yields results in only 3days.Additionally,the RLGS spot intensity provides direct quantitation of methy-lation level and can also detect partial methylation such as imprinting.Consequently,RLGS is appropriate for methy-lation surveys that broadly sample a genome.The utility of RLGS has been demonstrated in the construction of genetic

Tomoko Takamiya 1,2,3Saeko Hosobuchi 1,2Tomotsugu Noguchi 1,2Kenji Asai 4

Eiji Nakamura 4Yoshiki Habu 5

Andrew H.Paterson 6Hiroshi Iijima 3

Yasufumi Murakami 2Hisato Okuizumi 1

1

Division of Genome and

Biodiversity Research,National Institute of Agrobiological Sciences (NIAS),Tsukuba,Ibaraki,Japan 2

Department of Biological Science and Technology,

Faculty of Industrial Science and Technology,Tokyo University of Science,Noda,Chiba,Japan 3

College of Pharmacy,Nihon University,Funabashi,Chiba,Japan 4

Department of Electrical and Electronics Engineering,Faculty of Engineering,Aichi Institute of Technology,Toyota,Aichi,Japan 5

Division of Plant Sciences ,NIAS,Tsukuba,Ibaraki,Japan 6

Plant Genome Mapping Laboratory,University of Georgia,Athens,GA,USA Received October 26,2007Revised March 14,2008Accepted April 23,

2008

Abbreviations:F1,?rst ?lial generation ;RLGS,restriction landmark genome scanning

Correspondence:Dr.Hisato Okuizumi,National Institute of Agrobiological Sciences-Division of Genome and Biodiversity Research,2-1-2Kannondai,Tsukuba,Ibaraki 305-8602,Japan E-mail:okuizumi@nias.affrc.go.jp Fax:181-29-838-7408

Electrophoresis 2008,29,4088–4095

4088

linkage maps[30,31],methylation analysis in tumor tissue [32,33],and identi?cation of imprinted genes in mammals [34–36].RLGS in rice bene?ts from the availability of the whole genome sequence[37],enabling in silico RLGS analysis[38].

Here,we assayed for methylation alteration in F1hybrid (Nipponbare?Kasalath)rice.We detected altered inheri-tance and demethylation of speci?c RLGS spots in F1 plants,indicating the utility of RLGS for the analysis of methylation in F1populations.

2Materials and methods

2.1Plant materials

The seeds of Oryza sativa L.subsp.japonica‘Nipponbare’and subsp.indica‘Kasalath’were sown and grown in the ?eld.F1hybrids were derived from crossing of Nipponbare as the seed parent and Kasalath as the pollen parent.

2.2Preparation of genomic DNA and RLGS

procedure

We collected leaf blades and sheaths of Nipponbare, Kasalath,and F1plants grown for8wk.Total genomic DNA was isolated by the CTAB extraction method,with modi?cations as described previously[39–41].We used 0.4m g of genomic DNA and performed RLGS with the restriction enzyme combination Not I–Msp I–Bam HI or Not I–Hpa II–Bam HI(NEB,Beverly,MA,USA),as described previously[38].Msp I and Hpa II are restriction enzymes that recognize the same sequence,but have different methylation sensitivities.In the case where the second C of CCGG is methylated(C m CGG),Msp I cleaves the site but Hpa II does not.

2.3Identi?cation of RLGS spots by in silico RLGS

and spot cloning

To identify RLGS spots,we developed a computer software named‘‘in silico RLGS’’,which simulates RLGS analysis of genome sequence data[38].Not I sites in the whole rice genome sequence(http://rgp.dna.affrc.go.jp/J/IRGSP/ Build3/build3.html)were searched and given id numbers. The nearest Msp I sites to the Not I sites were also located. The software calculated the length and mobility of each DNA fragment(from Not I end to Msp I end or to next Not I end)in the1-D electrophoresis gel.We also surveyed Bam HI sites,calculated the DNA fragment length(from Not I end to Bam HI end)in the2-D gel,and drew a2-D pattern(in silico RLGS pattern).We compared autoradio-graphic RLGS patterns with corresponding in silico RLGS patterns and identi?ed each RLGS spot.

Spots unidenti?ed by in silico RLGS analysis were cloned and sequenced as described previously[38]with speci?c cloning linkers:Not I linker(5’-GGCCGCAT-GAATGGCGCGCCAAAGA-3’,3’-CGTACTTACCGCGCG GTTTCT-Biotin-5’)and Bam HI linker(5’-GATCCTG-TACTGCACCAGCAAATCC-3’,3’-GACATGACGTGGTCG TTTAGG-5’).

2.4Con?rmation of restriction enzyme sites

To compare methylation status among Nipponbare,Kasa-lath,and F1s,we con?rmed the presence of restriction enzyme sites in the parents.We designed?anking primers for the Not I and Msp I/Hpa II sites of each RLGS spot.By using1ng of Nipponbare or Kasalath genomic DNA as a template,PCR was carried out with0.4U of KOD plus polymerase(Toyobo,Tokyo,Japan), 1.5m L of?anking primers(10pmol/m L),1mM MgSO4,0.2mM dNTPs,and KOD buffer(total volume:20m L).PCR conditions were941C for5min followed by30cycles of941C for15s,601C for30s and681C for1min.Part of each PCR product was treated with Not I or Msp I.Then untreated and treated products were electrophoresed in an agarose gel(0.8–3.0%),and the band sizes were compared to con?rm that the sites were present and did not differ by DNA size polymorphism.Then we analyzed the methylation status of these sites and their inheritance in the F1as described in Section2.5.

2.5Con?rmation of methylation status

2.5.1PCR-based method

We checked the methylation status of Not I and Msp I/Hpa II sites of RLGS spots.Genomic DNA(1ng)of Nipponbare, Kasalath,or F1s was digested with30U of Not I,Msp I,or Hpa II,and used as a PCR template.Undigested genomic DNA was used as a positive control.PCR was performed as described in Section2.4.

2.5.2Sequence-based method

We used a sequence-based method to analyze some methylated sites in the F1to determine which parent the methylation was inherited from.First,we obtained PCR products by ampli?cation of genomic DNA of Nipponbare and Kasalath with?anking primers.Nucleotide sequences of the products were determined directly with an Applied Biosystems3130Genetic Analyzer(Applied Biosystems, Foster City,CA,USA)and a BigDye Terminator v3.1Cycle Sequencing Kit(Applied Biosystems)to reveal single nucleotide polymorphisms(SNPs)between Nipponbare and Kasalath.Then,genomic DNA of F1s was digested with Not I or Hpa II to create https://www.360docs.net/doc/fb10509920.html,ing the SNPs to determine the parental origin(s),we could then sequence the fragments with methylated Not I or Hpa II sites.

Electrophoresis2008,29,4088–4095Nucleic acids4089

3Results

3.1Analysis of RLGS patterns and spot

identi?cation

Figure1A shows the RLGS pattern of Nipponbare with the Not I–Msp I–Bam HI combination([Msp I]pattern).There were85spots in this pattern.To identify these spots,we analyzed the genome sequence data of Nipponbare with the in silico RLGS software.The resultant pattern showed117 spots in the same range(Fig.1B).Fifty-six spots were identi?ed by comparison of the relative spot positions of two patterns.For example,spot97was detected in both patterns, with a locus on chromosome(Chr.)9.The same genomic DNA was used for RLGS with Not I–Hpa II–Bam HI([Hpa II] pattern),and78spots were detected(Fig.1C).

Comparing

Figure 1.RLGS patterns for detection of

methylated sites.(A)RLGS pattern of Nippon-

bare genomic DNA with restriction enzyme

combination Not I–Msp I–Bam HI([Msp I]

pattern).(B)In silico RLGS pattern predicted

from rice genome sequence data.(C)Nippon-

bare Not I–Hpa II–Bam HI([Hpa II])pattern.In

comparison with[Msp I]pattern,25spots were

speci?c to[Msp I]and18to[Hpa II].(D)Kasalath

[Msp I]pattern.(E)Kasalath[Hpa II]pattern from

the genomic DNA in(D):19spots were speci?c

to[Msp I],and13to[Hpa II].(F)F1hybrid[Msp I]

pattern.In comparison with its parents,spots

200and235were absent,and spot f2was new.

(G)F1hybrid[Hpa II]pattern from the genomic

DNA in(F):29spots were speci?c to[Msp I],

and26to[Hpa II].Compared with the parental

patterns(C and E),spots23,65,200,323,501,

and525were absent,and F1-speci?c spot f2

was new(G).

Electrophoresis2008,29,4088–4095 4090T.Takamiya et al.

the [Hpa II]pattern (Fig.1C)with the pattern resulting from the methylation-insensitive [Msp I](Fig.1A)revealed 43spots that differed in methylation.Similarly,RLGS analysis of Kasalath showed 77spots in [Msp I](Fig.1D)and 71in [Hpa II](Fig.1E).Comparison of both patterns revealed 32spots that differed.In [Msp I]patterns of Nipponbare and Kasalath,35and 27spots,respectively,were speci?c,and 50spots (45%)were common.In [Hpa II]patterns,42and 35spots,respectively,were speci?c,and 36spots (32%)were common.

Samples of the spots that were not identi?ed by in silico RLGS were cloned from the 2-D polyacrylamide gel.For example,spot 326,which was speci?cally detected in [Hpa II]

patterns of Nipponbare and Kasalath (Fig.1C and E),proved to be identical to spot 97(Fig.1A and see above).This shows that the Msp I/Hpa II site (1354bp downstream of Not I site)of spot 97was methylated (C m CGG)and shifted to a higher molecular weight as spot 326.

RLGS analysis of the F1hybrid revealed a total of 111spots in [Msp I]pattern (Fig.1F)and 108in [Hpa II]pattern.Seven parental spots (23,65,200,235,323,501,and 525)disappeared,and one spot (f2)was newly detected in the F1.These unexpected spots,summarized in Table 1,seemed to indicate abnormal inheritance.

3.2Con?rmation of restriction enzyme sites by PCR

We designed ?anking primers for the restriction enzyme sites of RLGS spots that were identi?ed by in silico RLGS or spot cloning,and we con?rmed the existence of sites and methylation status by PCR.First,we con?rmed the presence of the restriction enzyme sites Not I and Msp I/Hpa II for RLGS spots in the genomic DNA of Nipponbare and Kasalath.The band in lanes 1and 2(Fig.2A)is the PCR product (579bp)of spot 97,ampli?ed from genomic DNA of each parent with ?anking primers for Msp I/Hpa II sites.The band in lanes 3and 4is the PCR product ampli?ed from each parent followed by treatment with Msp I.The DNA fragments of lanes 3and 4are smaller than those of lanes 1and 2because the Msp I/Hpa II sites were digested,and divided into 456bp (detected bands)and 123bp (estimated size).The result shows the existence of the Msp I/Hpa II site in both Nipponbare and Kasalath.Next we analyzed the

Table 1.Altered inheritance of RLGS spots Spot no.

In silico

Nipponbare Kasalath F1Msp I

Hpa II Msp I Hpa II Msp I Hpa II 23x o o o o o x 65x o o o o o x 200x o o x x x x 235x o o x x x o 323x x o x x x x 501x x x x o x x 525x x x x o x x f2

x

x

x

x

x

o

o

o,spot present;x,spot absent.One spot (f2)was newly detected in F1,and seven spots were absent in the

F1.

Figure 2.Con?rmation of restriction enzyme sites of RLGS spots and their methylation status.(A)Con?rmation of restriction enzyme sites with ?anking primers for the Msp I/Hpa II site of spot https://www.360docs.net/doc/fb10509920.html,nes 1and 2are PCR products ampli?ed from genomic DNA of Nipponbare (N)and Kasalath (K),respec-tively.The PCR products were puri?ed,treated with Msp I,and then loaded into lanes 3(N/M)and 4(K/M),respectively.(B)PCR-based DNA methylation analysis of spot 97of the parents (Nipponbare and Kasalath).Lanes 1,2,and 3for Nipponbare or 4,5,and 6for Kasalath are the PCR products for the templates (un-,Msp I-,and Hpa II-digested genomic DNA,respectively).(C)PCR-based DNA methylation analysis of spot 97in the https://www.360docs.net/doc/fb10509920.html,ne 1is the uncut positive control (U).Lanes 2and 3are PCR products from F1genomic DNA that was treated with Msp I (M)and Hpa II (H),respectively.(D)Part of the nucleotide sequence of the DNA fragment in lane 1in (B)(Nipponbare).(E)Part of the nucleotide sequence of the DNA fragment in lane 4in (B)(Kasalath).One SNP (shown by arrow-head)was detected in this region.(F)Part of the nucleotide sequence of the DNA fragment in lane 3in (C).This DNA fragment had C/T heterozygously.

Electrophoresis 2008,29,4088–4095

Nucleic acids 4091

methylation status of the Msp I/Hpa II site of spot97as described in Section3.3.

3.3Con?rmation of methylation status by PCR

We checked the methylation status of spots with restriction enzyme sites in Nipponbare and https://www.360docs.net/doc/fb10509920.html,nes1and4 (Fig.2B)were PCR products ampli?ed from genomic DNA of Nipponbare and Kasalath,respectively,as positive controls for spot97.Similarly,Msp I-treated genomic DNAs of each were used as templates,and the products were loaded in lanes2and5,respectively.No band was detected in these lanes.On the other hand,using Hpa II-treated genomic DNA as a template detected the same bands as in the positive control in lanes3and6,respectively. The results show that the Msp I/Hpa II site of spot97 was methylated(C m CGG)in Nipponbare and Kasalath and corresponded to the RLGS result.We checked the methylation status of90Not I and92Msp I/Hpa II sites(96 actual or in silico spots in total)in Nipponbare by a PCR-based method(summarized in Supporting Information Table1).Using the same?anking primer sets,we checked the methylation status of82Not I and84Msp I/Hpa II sites (in total93spots)in Kasalath.In total,60out of182sites (33%)were methylated in Nipponbare,and59out of166 sites(36%)in Kasalath.We found a total of four Nipponbare-speci?c methylated sites and ten Kasalath-speci?c sites.

3.4Analyses of methylation inheritance

3.4.1Inherited spot

Using PCR and sequence-based methods,we analyzed methylation inheritance in the F1at the sites that were methylated in Nipponbare or Kasalath.Figure2C shows the methylation status of the Msp I/Hpa II site of spot97in the https://www.360docs.net/doc/fb10509920.html,ne1is the positive https://www.360docs.net/doc/fb10509920.html,nes2and3are PCR products ampli?ed from genomic DNA of the F1treated with Msp I and Hpa II,respectively.A PCR product of the Hpa II-digested template was detected(lane3);it showed the same methylation as in the parents.To determine the parental origin of the methylated allele,we puri?ed the bands in lanes1and4of Fig.2B and in lane3of Fig.2C for sequence analysis.We found an SNP for spot97showing C in Nipponbare(Fig.2D)and T in Kasalath(Fig.2E).Next, we sequenced the DNA fragment of lane3(Fig.2C),which was ampli?ed from Hpa II-digested genomic DNA of the F1. It was heterozygous,with both C and T(Fig.2F).This result shows that methylation prevented digestion of the Msp I/Hpa II sites of both alleles.This con?rms that the methylation of the Msp I/Hpa II site was inherited from both parents.Similarly,we examined the other25methy-lated spots that showed the same appearance or absence between the parent and F1and con?rmed that all methylation statuses were inherited by the F1(summarized in Supporting Information Table1).

3.4.2Altered spots

Eight spots(23,65,200,235,323,501,525,and f2)were altered in the F1(Table1).One spot,f2,was newly detected in the F1,in both[MspI]and[Hpa II]patterns.On the other hand,seven parental spots(23,65,200,235,323,501,and 525)disappeared in the F1.Three spots(200,235,and323) were speci?c to Nipponbare,and two(501and525)to Kasalath([Hpa II]).The remaining two spots(23and65) were detected in both parents.

As an example,we cloned spot323.This spot is[Hpa II]-speci?c in Nipponbare but was not detected in Kasalath or the F1(Table1).The possibility of heterozygous methylation of an Msp I/Hpa II site in the parental Nippon-bare was low,because all parental selfed progeny had full (not half)spot intensity(data not shown).In the parental Nipponbare,the M1and M2sites of spot323(Fig.3A)were methylated,and the DNA fragments that were digested at N (Not I)and M3(Msp I/Hpa II)were size-fractionated by1-D electrophoresis in RLGS,followed by a third digestion with Bam HI.The fragments between sites N and B(Bam HI) were fractionated in2-D gel as spot323(Fig.3A). The disappearance of spot323in F1patterns(Fig.1G)is due to demethylation at the Msp I/Hpa II sites.We analyzed methylation at Msp I/Hpa II sites in Nipponbare and eight F1plants(F1-1–8)(Fig.3B).Genomic DNA was treated with Msp I(lanes marked M)or Hpa II(lanes marked H)and used as templates.A major band in Nipponbare was detected in lanes2(Msp I)and3(Hpa II).This result shows that the M1site of Nipponbare was methylated (m CCGG or m C m CGG).On the other hand,no band

was Figure3.Analysis of an abnormally inherited RLGS spot(323).

(A)DNA of spot323was located on Chr.8.Arrows P1and R1 indicate?anking primers for PCR-based methylation analysis.

(B)PCR-based methylation analysis of M1site of spot323in the parents and eight F1hybrids(F1-1–8).Lanes1,4,7,10,13,16,19, 22,25,and28are the PCR products from genomic DNA of each line as positive controls(U5uncut).M and H indicate Msp I and Hpa II digests of each line.The methylation status of M2and M3 sites were also checked(data not shown).m:size marker,1.0kb band.

Electrophoresis2008,29,4088–4095

4092T.Takamiya et al.

detected in F1-1,-2,-3,-4,-6,-7,or-8treated with either Msp I or Hpa II.These results show that M1sites were demethylated in these F1s.F1–5did not show demethyla-tion at M1,but M2was demethylated(data not shown). Methylation of the Not I site of spot323was not detected in the parents or F1s(data not shown).Additionally,we detected demethylation at the Not I of spot52(Chr.11)and the Msp I/Hpa II site of spot105(Chr.3)in F1.These spots were speci?cally detected in the in silico pattern,but not in the actual parental patterns,because Not I sites were methylated.

3.4.3Mapping of methylation status

We identi?ed103spots and analyzed the methylation status of most by RLGS and PCR(Supporting Information Table1)as summarized in Fig.4.Seventeen Not I sites were located within2.0kb upstream of a gene(50region),63 sites within a gene,and14within 2.0kb downstream (3’region).The remaining nine sites were in intergenic regions.Twenty-?ve Msp I/Hpa II sites were located in a 5’upstream region,48within a gene,and15in a3’downstream region.Thus,182sites(88%)out of206were located between 2.0kb upstream and downstream of genes.Numbered spots in the map(Fig.4)were methylated at one or more Not I or Msp I/Hpa II sites of Nipponbare, Kasalath,or F1.We mapped three altered spots(shown as‘‘AI’’);the rest(23,65,200,235,501,525,and f2)are waiting to be cloned.4Discussion

4.1Analysis of altered inheritance

Before this research,we had found that selfed progeny of both Nipponbare and Kasalath showed methylation poly-morphism(data not shown).RLGS analysis of nine selfed progeny of Nipponbare revealed some heterozygous methy-lation in the parent.Therefore,in this research it was important to designate a speci?c individual as a crossing parent in order to analyze the inheritance of methylation. That way we could eliminate the noise of methylation polymorphism and?nd cases of abnormal inheritance.

The detection of different spots among Nipponbare, Kasalath,and F1s contributes to the detection of different methylation statuses among them.Methylation statuses may have important roles in speci?c gene expression.Addition-ally,revealing how speci?c methylation status is inherited or altered in the next generation helps to explain the role of methylation in heterosis,reproduction,and genome barriers. For example,spot323is[Hpa II]-speci?c in Nipponbare but was not detected in the F1(Table1).Analysis by PCR detected demethylation of the M1or M2site of spot323in all F1hybrids.This region has some differences in base sequence between Nipponbare and Kasalath,because primers P1and R1did not amplify PCR products from genomic DNA of Kasalath(lane4of Fig.3B).Four other spots(200,235,501,525)that disappeared in the F1were speci?c to Nipponbare or Kasalath(Table1).Therefore,

the Figure4.Map of methylated sites.Numbered loci had at least one methylation in Not I or Msp I/Hpa II digests of Nipponbare,Kasalath,or F1s.‘‘AI’’indicates three loci with altered inheritance.Inheritance at the other loci appeared to be consistent.Centromeres(CEN)are indicated by ellipses.

Electrophoresis2008,29,4088–4095Nucleic acids4093

regions of these spots may have some differences in base sequence or methylation status between Nipponbare and Kasalath.Alternatively,these polymorphic regions might have altered methylation status in the F1hybrid.The methylation of spot323suggests possible sequence-depen-dent demethylation,for example,as a result of paramutation induced by allelic exclusion[42].It is important to examine which allele is expressed in F1hybrids in order to reveal the role of methylation in spot323in allelic exclusion.Spot200 was detected in[Msp I]and[Hpa II]patterns of Nipponbare but disappeared in the F1,and its spot intensity was half that of other spots.From these results,we predicted that the locus of spot200was methylated heterozygously in the parent (Nipponbare),and the presence and absence of the spot segregated1:1in the F1s.However,RLGS analysis of nine selfed progeny detected spot200at half intensity in all nine selfed progeny of Nipponbare.Therefore,it is unlikely that there are multiple copies in the genome of Nipponbare,with one homozygous for methylation and another homozygous for non-methylation.These results suggest that the region of spot200is always methylated heterozygously as seen in imprinted genes and possibly regulates the monoallelic expression of downstream genes.

4.2How is DNA methylation altered in F1hybrids? DNA methylation is often altered in F1hybrids,introgres-sive lines,and allopolyploids[6–14].Dong et al.[14] proposed three possible mechanisms for introgression:(i) intrinsic epigenetic chromatin states are disturbed by insertion of alien chromatin;(ii)an exogenous trans-acting ‘‘methylation-modifying’’factor is introduced from the donor;or(iii)altered cytosine methylation patterns are directed by genomic rearrangements that lead to the formation of aberrant structures.These proposals are relevant to explaining our?ndings that abnormal inherited spots had some differences in base sequence or methylation status between Nipponbare and Kasalath(spots200,235, 323,501,and525).Those differences might re?ect aberrant chromatin structure and methylation alteration in the F1 hybrid.Additionally,the methylation alteration in the maturation of pollen[43]indicates that methylation pattern changes in gametophytes in a sex-dependent manner. Although this phenomenon has not been recon?rmed yet, it might provide clues to methylation alteration in F1 hybrids.

In mammals,many imprinted genes are methylated in a sex-dependent manner,and the methylation pattern is maintained in somatic cells[44–46].On the other hand, evidence of sex-dependent methylation(genomic imprint-ing)in plants is only in endosperm.Whether mammalian-type imprinted genes occur in plants is unclear.Some imprinted genes of mammals have been detected by RLGS analysis of reciprocal crosses[34–36].RLGS analysis of reciprocal hybrids in plants might provide new observations of the role of DNA methylation in reproduction.

We thank Drs.Kawase M.,Tomioka K.,Okamoto H.,Ueda T.,Takahashi S.,and Shinozuka H.for their invaluable advice or technical support.This work was supported by the Ministry of Agriculture,Forestry and Fisheries to H.O.

The authors have declared no con?ict of interest.

5References

[1]Monk,M.,Boubelik,M.,Lehnert,S.,Development1987,

99,371–382.

[2]Tada,M.,Tada,T.,Lefebvre,L.,Barton,S. C.et al.,

EMBO J.1997,16,6510–6520.

[3]Bender,J.,Fink,G.R.,Cell1995,83,725–734.

[4]Jacobsen,S.E.,Meyerowitz,E.M.,Science1997,277,

1100–1103.

[5]Kakutani,T.,Munakata,K.,Richards,E.J.,Hirochika,H.,

Genetics1999,151,831–838.

[6]Matzke,M. A.,Mittelsten Scheid,O.,Matzke, A.J.,

Bioessays1999,21,761–767.

[7]Wendel,J.F.,Plant Mol.Biol.2000,42,225–249.

[8]Shaked,H.,Kashkush,K.,Ozkan,H.,Feldman,M.et al.,

Plant Cell2001,13,1749–1759.

[9]Madlung,A.,Masuelli,R.W.,Watson,B.,Reynolds,S.

H.et al.,Plant Physiol.2002,129,733–746.

[10]Pikaard,C.S.,Trends Genet.2001,17,675–677.

[11]Comai,L.,Madlung, A.,Josefsson, C.,Tyagi, A.,

Philos.Trans.R.Soc.Lond.B Biol.Sci.2003,358, 1149–1155.

[12]Akimoto,K.,Katakami,H.,Kim,H.J.,Ogawa,E.et al.,

Ann.Bot.(Lond.)2007,100,205–217.

[13]Liu,Z.,Wang,Y.,Shen,Y.,Guo,W.et al.,Plant Mol.

Biol.2004,54,571–582.

[14]Dong,Z.Y.,Wang,Y.M.,Zhang,Z.J.,Shen,Y.et al.

Theor.Appl.Genet.2006,113,196–205.

[15]Grossniklaus,U.,Vielle-Calzada,J.P.,Hoeppner,M.A.,

Gagliano,W.B.,Science1998,280,446–450.

[16]Kiyosue,T.,Ohad,N.,Yadegari,R.,Hannon,M.et al.,

https://www.360docs.net/doc/fb10509920.html,A1999,96,4186–4191.

[17]Kinoshita,T.,Yadegari,R.,Harada,J.J.,Goldberg,R.B.

et al.,Plant Cell1999,11,1945–1952.

[18]Choi,Y.,Gehring,M.,Johnson,L.,Hannon,M.et al.,

Cell2002,110,33–42.

[19]Xiao,W.,Gehring,M.,Choi,Y.,Margossian,L.et al.,

Dev.Cell2003,5,891–901.

[20]Kohler,C.,Hennig,L.,Spillane,C.,Pien,S.et al.,Genes

Dev.2003,17,1540–1553.

[21]Scott,R.J.,Spielman,M.,Curr.Biol.2004,14,

R201–R203.

[22]Kohler,C.,Page,D.R.,Gagliardini,V.,Grossniklaus,U.,

Nat.Genet.2005,37,28–30.

[23]Baroux,C.,Gagliardini,V.,Page,D.R.,Grossniklaus,U.,

Genes Dev.2006,20,1081–1086.

[24]Gehring,M.,Huh,J.H.,Hsieh,T.F.,Penterman,J.et al.,

Cell2006,124,495–506.

Electrophoresis2008,29,4088–4095

4094T.Takamiya et al.

[25]Zhang,X.,Yazaki,J.,Sundaresan,A.,Cokus,S.et al.,

Cell2006,126,1189–1201.

[26]Hayashizaki,Y.,Hirotsune,S.,Okazaki,Y.,Hatada,I.

et al.,Electrophoresis1993,14,251–258.

[27]Okazaki,Y.,Okuizumi,H.,Sasaki,N.,Ohsumi,T.et al.,

Electrophoresis1995,16,197–202.

[28]Reyna-Lopez,G.E.,Simpson,J.,Ruiz-Herrera,J.,Mol.

Gen.Genet.1997,253,703–710.

[29]Xiong,L.Z.,Xu,C.G.,Saghai Maroof,M.A.,Zhang,Q.,

Mol.Gen.Genet.1999,261,439–446.

[30]Okuizumi,H.,Okazaki,Y.,Ohsumi,T.,Hayashizaki,Y.

et al.,Electrophoresis1995,16,233–240.

[31]Okuizumi,H.,Okazaki,Y.,Ohsumi,T.,Hanami,T.et al.,

Electrophoresis1995,16,253–260.

[32]Ohsumi,T.,Okazaki,Y.,Okuizumi,H.,Shibata,K.et al.,

https://www.360docs.net/doc/fb10509920.html,mun.1995,212,632–639.

[33]Miwa,W.,Yashima,K.,Sekine,T.,Sekiya,T.,Electro-

phoresis1995,16,227–232.

[34]Hayashizaki,Y.,Shibata,H.,Hirotsune,S.,Sugino,H.

et al.,Nat.Genet.1994,6,33–40.

[35]Shibata,H.,Yoshino,K.,Muramatsu,M.,Plass,C.et al.,

Electrophoresis1995,16,210–217.[36]Plass,C.,Shibata,H.,Kalcheva,I.,Mullins,L.et al.,Nat.

Genet.1996,14,106–109.

[37]International Rice Genome Sequencing Project,Nature

2005,436,793–800.

[38]Takamiya T,Hosobuchi S,Asai K,Nakamura E.et al.,

Electrophoresis2006,27,2846–2856.

[39]Kawase,M.,Theor.Appl.Genet.1994,89,861–864.

[40]Okamoto,H.,Takamiya,T.,Saito,A.,Domon,E.et al.,

Jpn.Agric.Res.Q.2006,40,65–69.

[41]Takamiya,T.,Ohtake,Y.,Hosobuchi,S.,Noguchi,T.

et al.,Jpn.Agric.Res.Q.2008,42,151–155.

[42]Chandler,V.L.,Eggleston,W.B.,Dorweiler,J.E.,Plant

Mol.Biol.2000,43,121–145.

[43]Oakeley,E.J.,Podesta,A.,Jost,J.P.,Proc.Natl.Acad.

https://www.360docs.net/doc/fb10509920.html,A1997,94,11721–11725.

[44]Brenton,J. D.,Drewell,R. A.,Viville,S.,Hilton,

K.J.et https://www.360docs.net/doc/fb10509920.html,A1999,96, 9242–9247.

[45]Drewell,R.A.,Brenton,J.D.,Ainscough,J.F.,Barton,

S.C.et al.,Development2000,127,3419–3428.

[46]Lau,J.C.,Hanel,M.L.,Wevrick,R.,Nucleic Acids Res.

2004,32,3376–3382.

Electrophoresis2008,29,4088–4095Nucleic acids4095