【作者】 卢丙越；

【导师】 江玲；

【作者基本信息】 南京农业大学 ， 作物遗传育种， 2011， 博士

【摘要】 水稻种子休眠性是一个重要的农艺性状,与穗发芽抗性密切相关,关系到稻米的产量和品质。在水稻常规育种中,近年来培育的高产品种一旦在收获季节遇到高温多雨的天气就很容易发生穗发芽。休眠性强的品种可以抵抗穗发芽,但会导致田间出苗率低,出苗参差不齐,不利于水稻直播栽培方式的推广。因此,培育具有适度休眠性的优良水稻品种显得尤为重要。N22是强休眠的栽培品种,前人研究表明N22的强休眠性由1-2个主基因控制。本研究通过两种不同的方法对N22种子的强休眠性进行遗传研究。一是以无休眠的粳稻品种南粳35为轮回亲本,N22为供体分别构建了两个主效QTL,qSdn-1和qSdn-5的高代回交群体和近等基因系(NIL),利用高代回交群体分别对qSdn-1和qSdn-5进行了精细定位；二是对强休眠品种N22进行诱变处理,筛选与休眠相关的突变体,对筛选到的突变体进行遗传分析。两种方法相互结合,为揭示N22种子强休眠性的遗传机理奠定良好的基础。1.以无休眠粳稻品种南粳35为轮回亲本,强休眠的籼稻品种N22为供体,通过不断的回交和标记辅助选择分别构建了qSdn-1和qSdn-5的高代回交群体及近等基因系。2008年正季利用482株BC4F2(gSdn-1)和367株BC4F2(gSdn-5)分别对qSdn-1和qSdn-5进行了定位验证,将qSdn-1定位在SSR标记RM11669和RM1216之间,与标记RM11694共分离,qSdn-1可解释休眠表型变异的24.58%；将qSdn-5定位在标记RM480和RM3664之间,可解释休眠表型变异的17.58%。qSdn-1和qSdn-5的定位区间与之前定位的位置一致。并利用同时含有qSdn-1和qSdn-5位点的449株BC4F2(qSdn-1/qSdn-5)高代回交群体对两个位点间的互作进行了分析,结果表明qSdn-1和qSdn-5间不存在上位性,基因在休眠表型上的效应是可以累加的。为了进一步验证qSdn-1和qSdn-5对种子休眠的作用,我们还构建了含有qSdn-1或qSdn-5单个QTL,或含有qSdn-1和qSdn-5两个QTL位点及在这两个位点上都不含有N22片段的BC3F5高代群体。对这些高代回交家系的发芽情况进行统计,BC3F5(gSdn-1).BC3F5(qSdn-5)、BC3F5(qSdn-1/qSdn-5)和BC3F5(CK)的平均发芽率分别为7.9%、11.1%、6.1%和86.3%,这一结果进一步证明了qSdn-1和qSdn-5对N22种子的强休眠性起重要作用,是两个主效休眠位点,同时含有qSdn-1和qSdn-5两个位点的高代回交群体发芽率更低,这也证实了它们的作用是可以累加的。经7天50℃的干热处理,qSdn-1和qSdn-5控制的种子休眠便能彻底打破。2009和2010年正季利用遗传背景更加纯合的BC5F2和BC5F3高代回交群体对qSdn-1和qSdn-5分别进行了精细定位。2009年利用SSR标记RM128和RM11781从7300株BC5F2(qSdn-1)群体中筛选到95株极端表型的重组个体,对这些重组单株于2010年对表型进行后代(BC5F3)验证,通过进一步的加密标记将qSdn-1定位在标记L24和L34之间约655kb的范围内,与标记L27共分离；同样利用标记RM7452和RM3664从5888株BC5F2(qSdn-5)中筛选到111株重组个体,经交换单株验证后将qSdn-5定位在122kb的范围内,与Indel标记15-6共分离。qSdn-1和qSdn-5的精细定位一方面为进一步的图位克隆工作奠定了良好的基础,另一方面与休眠位点紧密连锁的标记可被用于分子标记辅助选择育种,培育具有适度休眠性的水稻优良品种对抗穗发芽。在精细定位的同时我们通过不断的回交及标记辅助选择构建了一套休眠QTL的NILs,即NIL (qSdn-1), NIL (qSdn-5)和NIL (CK)。2010年正季NIL (qSdn-1)、NIL (qSdn-5)和NIL(CK)的发芽率分别为23%、35%和98%,而NILs其它农艺性状与背景亲本南粳35没有差异,这也进一步验证了qSdn-1和qSdn-5对N22种子休眠性的作用。种子休眠通常与植物激素有密切的关系,通过外源激素及逆境处理实验表明NIL(qSdn-1)与NIL (qSdn-5)对ABA、GA和NaCl的敏感性存在差异,NIL (qSdn-5)对ABA、GA和NaCl表现的更加敏感,IAA处理实验NIL (qSdn-1)和NIL (qSdn-5)没有明显的差异,随IAA浓度的升高发芽率都有所上升。对外源激素及逆境处理响应的差异表明qSdn-1和qSdn-5潜在的基因调控种子休眠的机制有所不同。2.利用400Gy 60Co辐照N22种子,通过对突变表型的筛选获得两个弱休眠的突变体,暂时命名为Q4359和Q4646。Q4359和Q4646抽穗后35天收获的种子平均发芽率为43%和45%,较野生型N22发芽率高,而且在室温存放过程中突变体种子的休眠性较N22更容易破除,在种子萌发过程中突变体对ABA和NaCl的敏感性降低,且Q4359较Q4646对ABA和NaCl更加不敏感；N22种子发芽率随外源GA浓度的增加有所上升,而突变体Q4359和Q4646都表现出对GA不敏感；在对IAA的敏感性上突变体与N22没太大差异,发芽率都有所上升。突变体间的正反交实验表明Q4359和Q4646突变位点不等位,遗传分析实验表明两个突变性状都是由隐性单基因控制。之前的研究表明N22种子休眠性由1-2个主基因控制,通过定位分析检测到两个主效位点(qSdn-1和qSdn-5),是否是N22中的这两个主效基因发生突变导致种子休眠性减弱?于是我们利用Q4359和Q4646与无休眠的水稻品种南粳35构建分离群体对突变体中的休眠位点进行定位分析。在Q4359/南粳35 F2群体中共检测到3个控制种子休眠的QTL,其中第3染色体上检测的位点增强休眠性的等位基因来自南粳35；第5染色体检测的位置与前面定位的结果一致,为qSdn-5;另外还检测到一个控制种子休眠的新位点gSdn-9,LOD值5.5,可解释11.5%的表型变异。在Q4646/南粳35 F2群体中检测到2个控制种子休眠的QTL,位于第1染色体的QTL与之前定位的qSdn-1的位置一致,第3染色体检测的位点与Q4359/南粳35 F2群体的位置相同,增效基因同样来自南粳35。另外,像qSdn-2、qSdn-7和qSdn-11在本研究中也没被检测到。Q4359和Q4646的弱休眠表型可能是由于qSdn-1和qSdn-5位点突变引起的。更多还原

【Abstract】 Grain dormancy is an important trait for breeders in many cereals because of its association with preharvest sprouting, which can damage end-use quality such as seed quality and yield in rice. In conventional rice breeding, high-yield varieties are liable to initiate germination before harvest given suitable environmental conditions around the time of crop maturity in recent years, strongly influence quality of rice. Strong levels of seed dormancy are correlated with a low probability of PHS and vice versa. Excessive dormancy of course can also be problematical, because it leads to uneven seedling establishment. Therefore, breeding for an intermediate level of dormancy in rice is highly desirable. The indica cultivar N22 has very strong level of dormancy. Here, two different strategies were used to genetic analysis of N22. The first, the advanced backcross (AB) populations and near-isogenic lines (NILs) contained qSdn-1 or qSdn-5 that were two major effect dormancy QTL in N22 was developed respectivily. qSdn-1 and qSdn-5 was fine mapping in a narrow region using AB-population and the effect of these two locus were also verfied using NILs. The second, the seeds of N22 were treated with 400Gy 60Co gamma-radiation, the mutants associated with seed dormancy were screened, and the simple genetic analyses were done for these mutants. These two strategies will provide some useful information to reveal the genetic mechanism for seed dormancy of N22.1. The intrachromosomal positions of the two grain dormancy quantitative trait locus (QTL) qSdn-1 (chromosome 1) and qSdn-5 (chromosome 5) were obtained from the segregation analysis of the advanced backcross populations derived from the cross between rice cultivars N22 and Nanjing35. Marker-assisted selection (MAS) was applied to select derivatives carrying one or both of qSdn-1 and qSdn-5 in a genetic background which was nearly isogenic to Nanjing35.In 2008, an analysis of dormancy in the BC4F2 population allowed qSdn-1 to be located between the simple sequence repeat (SSR) markers RM11669 and RM1216; the QTL explained 24.58% of the overall phenotypic variation and the most closely linked marker was RM11694. qSdn-5 was mapped between RM480 and RM3664, and explained 17.58% of the overall phenotypic variation. The SSR locus RM19080 mapped within 0.4 cM of qSdn-5. No epistasis was observed between qSdn-1 and qSdn-5. The mean germination rates of lines containing qSdn-1, qSdn-5 and both qSdn-1 and qSdn-5 was 7.9, 11.1 and 6.1%, respectively, whereas that of the check line lacking both QTL was 86.3%. The dormancy of both qSdn-1 and qSdn-5 could be readily broken by a 7-day post-harvest treatment at 50℃.Later, qSdn-1 and qSdn-5 were fine mapped using advanced backcross populations BC5F2 and BC5F3 in 2009 and 2010. In 2009,95 extreme recombinant plants were identfied using the SSR markers RM128 and RM11781 from 7300 BC5F2 (qSdn-1), the phenotypes of these recombinant plants were verified by the progenies (BC5F3) in 2010, qSdn-1 was mapped between SSR marker L24 and L34 with about 655kb, co-segregating with L27; using the same method,111 extreme recombinant plants were identfied using RM7452 and RM3664 from 5888 BC5F2 (qSdn-5), finally, qSdn-5 was mapped between Indel marker 15-2 and SSR marker RM19080 with 122kb, co-segregating with Indel marker 15-6. Fine mapping of qSdn-1 and qSdn-5 established good base for map-based cloning of these two QTL. The SSR loci linked most tightly to qSdn-1 and qSdn-5 are suitable for MAS for reduced pre-harvest sprouting in rice.When we fine mapped qSdn-1 and qSdn-5, three NILs, NIL (qSdn-1), NIL (qSdn-5) and NIL (CK) were developed using phenotype and marker-assisted selection. The germination rates of NIL (qSdn-1), NIL (qSdn-5) and NIL (CK) were 23%,35% and 98%, respectively in 2010, while the major agronomic traits of the NILs were same with Nanjing35. This also veried qSdn-1 and qSdn-5 act very important effect for seed dormancy of N22.Seed dormancy have a tight connection with plant hormones, the seed germinations of NIL (qSdn-1) and NIL (qSdn-5) were treated with hormone and adversity stress, the results showed different sensitivity to ABA, GA and NaCl between NIL (qSdn-1) and NIL (qSdn-5), NIL (qSdn-5) displayed more sensitivity than NIL (qSdn-1), there were no sensibly different between NIL(qSdn-1) and NIL (qSdn-5) when treated with IAA, the different sensitivity indicated the genetic mechanism of the gene underlying qSdn-1 and qSdn-5 is different.2. Two weak dormancy mutants, designated Q4359 and Q4646, were obtained from the rice cultivar N22 after treatment with 400Gy 60C0 gamma-radiation. The germination rates of Q4359 and Q4646 were 43% and 45%, respectively after 35d heading, higher than <2% of wide-type N22. The dormancy of the mutant seeds was more readily broken when exposed to period of room temperature storage. The mutants also showed a reduced level of sensitivity to ABA and NaCl compared to the N22 cultivar, although Q4359 was more insensitive than Q4646. The germination rate of N22 increased with accruement of GA concentration, while Q4359 and Q4646 displayed insensitive to GA. The germination rates of N22, Q4359 and Q4646 all increased with accruement of IAA concentration, no obvious different.A genetic analysis indicated that in both mutants, the reduced dormancy trait was caused by a single recessive allele of a nuclear gene, but that the mutated locus was different in each case. There is one or two major gene(s) in N22, in previous study, qSdn-1 and qSdn-5 were detected as major effect QTL. Whether or not the reduced dormancy accociated with these two QTL? So we detected the dormancy QTL of Q4359 and Q4646 using two segregation populations, Q4359/Nanjing35 F2 and Q4646/Nanjing35 F2. In Q4359/Nanjing35 F2,3 QTL, qSdNj-3, qSdn-5, and qSdn-9 were detected on chromosome 3,5 and 9, respectively. The QTL qSdn-9 was determined to be a novel dormancy locus, and it was mapped between SSR markers RM7038 and RM105 with a LOD score of 5.5, explaining 11.5% of the overall trait variation. The major dormancy QTL qSdn-1 was not detected in Q4359/Nanjing35 F2. Two QTL, qSdn-1 and qSdNj-3, were detected in Q4646/Nanjing35 F2, the position of qSdNj-3 was accorded with the QTL in Q4359/Nanjing35 F2, and qSdn-5 was not detected in Q4646/Nanjing35 F2. The following QTL:qSdn-2, qSdn-7 and qSdn-11 were not detected in the two populations. Therefore, qSdn-1 and qSdn-5 could be inherited as the major dormancy locus, but qSdn-5 in Q4359 and qSdn-1 in Q4646 were not detected. Whether the reduced dormancy phenotype caused by qSdn-1 and qSdn-5 mutated or not will require further investigation.更多还原