【正文】 ing the strong plant height reduction ability. Thus, Epidf may be used for rice intersubspecific heterosis breeding in the future.(2) The mutative gene was fine mapped within a 49kb region, whereas there was no nucleotide mutation. However, we found ORF5 (one of the seven ORFs within the mapping region) was ectopically expressed in Epidf, but silenced in WT. Considered the revertants emerged from Epidf population, we speculated that Epidf was an epigenetic mutant. By using bisulfite sequencing, we found DNA hypomethylation was occurred at 5’ region of FIE1. We also analyzed the methylation patterns of six revertants and found even FIE1 was silenced in them, DNA methylation was not recovered to the WT level, the recovered sits were enriched around the transcriptional starting site and within the third exon. We also found there were reduced H3K9me2 and increased H3K4me3 at the 5’ region of FIE1. Thus, Epidf is an unexpected epigenetic mutant, which would like to provide an intriguing opportunity to unravel epigenetic modifications for development regulation in important crop plants.(3) We discovered that FIE1 was a maternalspecific expressing gene in endosperm, which was coincided with the methylation pattern of FIE1 in tissues. The methylation level is higher in leaf, culm and young panicle than that in endosperm 6, 9 and 12 days after pollination. We also found the methylation of maternal FIE1 was much lower than that of paternal. If Epidf was used as pollen donator, the imprinting pattern was disturbed, and the methylation levels of both parental were lower. Unlike Arabidopsis, which contains only one ubiquitously expressed FIE gene, there are two FIE gene in rice, the other gene FIE2 is expressed in all the tissues. We conclude that during the genome duplication and the latter evolution, epigenetic marks may play an important role in the differentiation of the two FIE gene in rice.(4) Yeast twohybrid assay showed FIE1 interacted with rice E(z) homologs, suggesting FIE1 participates in PRC2 repression which catalyzes H3K27me3 at targets. Microarray analysis showed 305 genes were misregulated in Epidf acpanied with changed H3K27me3 levels. Thus, ectopic expression of FIE1 resulted in the mutant phenotype via abnormal distribution of H3K27me3. (5) H3K9me2 and H3K27me3 are two conserved repressive epigenetic marks in both animal and higher plants. H3K9me2 is mainly enriched in heterochromatin and functions in suppressing transposons, while H3K27me3 is mainly localized in euchromatin and provides a cellular memory to maintain the repressive state of target genes. We found there was high level of H3K9me2 at 5’ region of FIE1, whereas in Epidf, H3K9me2 was reduced and resulted in ectopic expression of FIE1 and abnormal distribution of H3K27me3. We conclude that silencing of FIE1 via H3K9me2 is essential for normal function of H3K27me3 in rice. KEY WORDS: Rice。與擬南芥僅有一個(gè)FIE基因不同，水稻含有兩個(gè)FIE基因，另一個(gè)FIE基因FIE2是組成性表達(dá)的，因此我們推測(cè)水稻基因組復(fù)制以及隨后的進(jìn)化過(guò)程中，表觀遺傳修飾對(duì)兩個(gè)FIE基因的功能分化起了重要作用。（2）利用圖位克隆方法將突變基因精細(xì)定位在49kb區(qū)域內(nèi)，但并未發(fā)現(xiàn)DNA序列突變，卻發(fā)現(xiàn)其中的ORF5在突變體中表達(dá)強(qiáng)烈，而在野生型中沉默。目前正在進(jìn)行的超級(jí)稻育種主要利用理想株型和秈粳亞種間雜種優(yōu)勢(shì)以期水稻單產(chǎn)得到進(jìn)一步提高。對(duì)本文的研究做出重要貢獻(xiàn)的個(gè)人和集體，均已在文中以明確方式標(biāo)明。本人授權(quán)南京農(nóng)業(yè)大學(xué)可以將本學(xué)位論文的全部或部分內(nèi)容編入有關(guān)數(shù)據(jù)庫(kù)進(jìn)行檢索，可以采用影印、縮印或掃描等復(fù)制手段保存和匯編學(xué)位論文。因此發(fā)掘新的水稻株高控制基因特別是顯性矮稈基因，對(duì)進(jìn)一步提高水稻單產(chǎn)具有重要意義。盡管FIE1在回復(fù)突變株中保持沉默，但DNA甲基化并未恢復(fù)到野生型水平，發(fā)生恢復(fù)的位點(diǎn)主要集中在轉(zhuǎn)錄起始點(diǎn)附近和第3個(gè)外顯子。利用染色質(zhì)免疫共沉淀技術(shù)發(fā)現(xiàn)這些表達(dá)量改變的基因同時(shí)伴隨著H3K27me3修飾的改變。 Epigenetic mutation。然而秈粳亞種間雜種表現(xiàn)株高偏高，生育期超親晚熟以及結(jié)實(shí)率偏低限制了秈粳雜種優(yōu)勢(shì)的利用(楊守仁等,1982)。 Holliday, 2006)。Takedea(1977)以節(jié)間長(zhǎng)度占株高的比例為指數(shù)將矮稈分為dn、dm、dg、nl和sh五種類型，將節(jié)間比例正常的品種定為N型。盧永根等（1979）和顧銘洪等（1988）對(duì)我國(guó)秈稻品種的矮生性遺傳作了系統(tǒng)性研究，發(fā)現(xiàn)我國(guó)秈稻中存在的矮稈基因有四個(gè)：sdsdg、sdn(t)和sdt(t)。植物激素都是簡(jiǎn)單的小分子有機(jī)化合物，但它們的生理效應(yīng)卻非常復(fù)雜，可以影響植物細(xì)胞的分裂、伸長(zhǎng)以及分化，從而影響植物株高以及其它生理過(guò)程。赤霉素是廣泛存在的一大類植物激素的總稱，其化學(xué)結(jié)構(gòu)都屬于二萜類酸，由四環(huán)骨架衍生而來(lái)。GA20ox2對(duì)應(yīng)于sd1位點(diǎn)，它在葉片、節(jié)間和已開放的花中表達(dá)強(qiáng)烈，而GA20ox1主要在未開放的花中表達(dá)。Sakamoto等（2004）利用其它物種中已知的赤霉素合成酶基因，從水稻基因組數(shù)據(jù)庫(kù)中分離了水稻赤霉素合成途徑中的29個(gè)候選同源基因。擬南芥中的GAI(Gibberllininsensitive)是赤霉素響應(yīng)的負(fù)調(diào)控因子，突變后使擬南芥矮化(Peng等, 1997)。 Hartweck等, 2006)。施加外源BR后能恢復(fù)表型，因此推斷brd1突變體中BR合成受阻。BRD2編碼擬南芥的同源蛋白DIM/DWF1，催化從24亞甲基膽固醇（24methylenecholesterol）到油菜甾醇（campesterol）這一過(guò)程(Hong等, 2005)。圖位克隆后發(fā)現(xiàn)D61與擬南芥BRI1同源，編碼BR受體激酶OsBRI1(Brassinosteriod insensitive1)(Yamamuro等, 2000)。Tong等（2009）發(fā)現(xiàn)一個(gè)矮化少蘗突變體dlt(dwarf and low tillering)，DLT編碼一個(gè)植物特有的GRAS家族蛋白，并且OsBZR1蛋白能結(jié)合到DLT的啟動(dòng)子，這與DLT對(duì)BR處理的負(fù)反饋調(diào)控機(jī)制相一致。而顯性矮稈基因的利用可以解決這個(gè)問(wèn)題,同時(shí)在一定程度上也彌補(bǔ)了隱性矮稈基因單一化帶來(lái)的品種遺傳背景狹窄的問(wèn)題。Asano等（2009）從N甲基N亞硝基脲（MNU）誘變的突變體庫(kù)中篩選到三個(gè)半顯性矮稈突變體Slr1d，圖位克隆后發(fā)現(xiàn)均來(lái)自DELLA蛋白SLR1的功能獲得性突變。因此表觀遺傳學(xué)研究的內(nèi)容從廣義上來(lái)講包括:DNA甲基化、組蛋白修飾、基因沉默、RNAi、基因組印記、染色質(zhì)重塑、RNA剪接與編輯、蛋白質(zhì)剪接與翻譯后修飾、X染色體失活以及副突變等。 從頭合成DNA甲基化從頭合成DNA甲基化是指RNA介導(dǎo)的DNA甲基化（圖11）。 Zaratiegui等, 2007。與常規(guī)的RNA聚合酶II（Pol II）類似，Pol IV和Pol V也是由多個(gè)亞基組成的蛋白復(fù)合體，它們各自包含特有的亞基，同時(shí)也共享一些亞基，或與Pol II共享一些亞基(Huang等, 2009。 Zheng等, 2007)，并通過(guò)堿基配對(duì)與Pol V合成的非編碼RNA結(jié)合(Wierzbicki等, 2008)。RDM1(RNAdircted DNA methylaton 1)是在篩選擬南芥轉(zhuǎn)錄后基因沉默（Transcriptional gene silencing, TGS）突變體中發(fā)現(xiàn)的，在RdDM效應(yīng)復(fù)合體中與AGO4和DRM2互作。 Pontier等, 2005。 Zilberman等, 2007)。無(wú)論是CMT3(Chromomethylase3)（負(fù)責(zé)維持CHG甲基化的DNA甲基轉(zhuǎn)移酶）功能的缺失(Lindroth等, 2001)，還是SUVH4/KYP(Kryptonite)（組蛋白H3K9甲基轉(zhuǎn)移酶）的突變，都會(huì)導(dǎo)致DNA甲基化水平的嚴(yán)重降低(Jackson等, 2002。另外一些位點(diǎn)的CHH甲基化是由CMT3和DRM2共同決定的(Cao等, 2003)。 OrtegaGalisteo等, 2008)。DME主要在配子發(fā)生時(shí)起建立印記的功能(Huh等, 2008)，而其它該家族成員主要在營(yíng)養(yǎng)器官中起作用，可能用于修復(fù)RdDM途徑產(chǎn)生的過(guò)度甲基化(Kapoor等, 2005。通過(guò)比較胚和胚乳的DNA甲基化水平鑒別出一些DMRs（DNA demethylation regions,控制印記的DNA甲基化區(qū)域）,并確定了5個(gè)親本特異性表達(dá)的基因和大約40個(gè)可能的印記基因(Gehring等, 2009)，這表明植物和動(dòng)物可能擁有相同數(shù)目的印記基因。圖12 植物中的DNA主動(dòng)去甲基化及其生物學(xué)功能(He等，2011)。 McCabe等, 2005。最后，MET1的下調(diào)可以保證由DME產(chǎn)生的半甲基化位點(diǎn)不會(huì)被再次全甲基化。這些修飾通過(guò)改變組蛋白DNA和組蛋白組蛋白的相互作用而引起染色質(zhì)的結(jié)構(gòu)和功能變化，影響基因表達(dá)，并且這種影響與氨基末端被修飾的位點(diǎn)和修飾的程度相關(guān)(Strahl等, 2000)。組蛋白賴氨酸甲基化是一種非常重要也是非常復(fù)雜的表觀遺傳修飾，不同位點(diǎn)的修飾和修飾的程度分別與轉(zhuǎn)錄激活或沉默相關(guān)。 組蛋白精氨酸甲基化組蛋白精氨酸甲基化主要發(fā)生在H3的第2位(H3R2)、第8位(H3R8)、第17(H3R17)位和第26位(H3R26)精氨酸，以及H4的第3位精氨酸(H4R3),這些位點(diǎn)的甲基化是由一個(gè)小的精氨酸甲基轉(zhuǎn)移酶(PRMTs)家族控制的。H4R3能夠被AtPRMT5/SKB1對(duì)稱性的二甲基化(Pei等, 2007。與prmt1敲除小鼠胚胎致死不同，擬南芥atprmt1a atprmt1b雙突變體并沒(méi)有明顯的突變表型。擬南芥GNAT家族包括AtGCN5/HAGHAT1/HAG2以及AtELP3/ELO3/HAG3,其中AtGCN5研究的最充分，它在體外能乙?；疕3(Earley等, 2007)，并且Atg5突變體中整體水平上的H3乙?；瘻p少(Bertrand等, 2003)，特別是特定基因的H3K14和H3K27位點(diǎn)(Benhamed等, 2006)。AtHDA6/RPD3B能夠使H3和H4多個(gè)賴氨酸位