【作者】 刘头明；

【导师】 张启发； 邢永忠；

【作者基本信息】 华中农业大学 ， 生物化学与分子生物学， 2009， 博士

【摘要】 每穗颖花数是水稻产量性状的重要构成因子,因此,研究该性状具有重要理论和实际意义。然而,该性状是复杂的数量性状,同时受到多个微效和主效的QTL调控。利用分子标记连锁图和统计分析方法,可以实现QTL的定位。进一步的研究表明,水稻每穗颖花数同时受到QTL,上位性和环境的共同影响。通过构建目标区段QTL的近等基因系,可以消除遗传背景的干扰,使数量性状呈现质量性状的分离,从而实现QTL的精细定位和克隆。本研究利用两个重组自交系群体珍汕97/特青和明恢63/特青构建两张遗传图谱,分别收集两群体的两年度重复的表型数据,对包括每穗颖花数和每穗实粒数在内的一共9个农艺性状进行QTL定位,对两群体的6个每穗颖花数QTL和2个每穗实粒数构建近等基因系,进而近等基因系背景下重新分析了4个QTL的遗传效应,精细定位了3个QTL,利用图位克隆策略,分离确定了SPP7b的候选基因。具体结果如下:1、利用176个SSR标记构建珍汕97/特青的遗传图谱,总长1432.1 cM,标记间的平均距离为8.1 cM;利用133个标记构建了明恢63/特青群体,遗传连锁图总长1371.4 cM,标记间的平均距离为10.3 cM。两群体共有的标记为50对,标记的相对位置基本相同,且与已发表的图谱有较好的一致性。2、对两个群体分别收集了两个年度的表型数据,考察了包括每穗颖花数和每穗实粒数等的一共9个性状,并进行了QTL分析。珍汕97/特青群体在2004和2006年各检测到26个QTL,其中13个QTL两年共同检测到。而明恢63/特青群体分别在2005和2006年共检测到24和28个QTL,其中13个QTL两年都检测到。3、利用每穗实粒数和抽穗期数据,对两重组自交系群体的的每穗实粒数性状进行条件QTL分析。结果显示,在珍汕97/特青群体的8个QTL中的5个受到抽穗期的影响,而剩下的3个QTL以及所有明恢63/特青群体的5个QTL都不受抽穗期的影响。进而将每穗实粒数QTL分为两类,第一类仅控制每穗实粒数性状（typeⅠ）,第二类则通过延长抽穗期来提高每穗实粒数（typeⅡ）。4、对6个每穗颖花数QTL,2个每穗实粒数QTL构建近等基因系。分别得到了BC₄F₂的分离群体种子。5、对SPP3b在近等基因系背景下重新估计了其遗传效应,在F₂和F₃代中分别检测到一个LOD值是12.8和8.8的QTL,加性效应分别是11.89和7.85,分别解释表型变异的29.1%和20.2%。同时也在该群体中检测到一个千粒重的QTL,F₂和F₃代的LOD值分别是26.2和17.0,加性效应分别是-1.81和-0.89,各自解释表型变异的50.4%和34.5%。通过后代测验发现这两个性状共分离,两基因被当作基因标记准确的定位在染色体相同的位置,与标记RM15855和W3D16分别相距1.6 cM和1.0 cM。很有可能是一个多效性QTL同时控制每穗颖花数和千粒重。6、利用SPPl近等基因系分析该QTL的遗传效应,在F₂和F₃代中分别检测到一个LOD值是23.95和35.52的QTL,加性效应分别是22.05和10.81,分别解释表型变异的51.1%和63.4%。利用2200株的分离群体,进一步将QTL精细定位在一个107 kb的区域,生物信息学预测该区间一共有17个基因。7、利用SPP6的NILs分析QTL的效应,发现三个性状,抽穗期,株高,每穗颖花数都存在分离。在F₂代中,株高,每穗颖花数性状的QTL LOD值分别是72.58,29.96,加性效应分别是7.72,22.14,解释表型变异的81.2%,50.9%。通过后代测验发现抽穗期,株高,每穗颖花数3个性状共分离,进一步把这个基因作为一个基因标记定位在染色体相同的座位,与RM19746和RM19795的距离分别是3.1和3.4 cM。SPP6很有可能是一个多效性QTL同时控制这3个性状。利用3300株的NILs的分离大群体,进一步将该QTL定位在一个约1Mb的区域。生物信息学预测1Mb的区域包含一个已克隆的抽穗期基因Hdl,然而Hdl与SPP6的显隐性关系恰好相反,因此可以确定SPP6并不是Hdl基因。8、利用SPP7b的NILs重新分析了QTL的遗传效应,发现在该群体中,株高,抽穗期和每穗颖花数都存在分离。在F₂和F₃代分别检测到LOD值为19.22和29.68的主效QTL,加性效应分别为18.86和17.04,解释表型变异的45.3%和50.6%的每穗颖花数的QTL。就株高,在两代中分别检测到QTL的LOD值分别为24.73和24.37,加性效应3.13和3.22,各自解释表型变异的50.2%和45.0%。抽穗期的LOD值分别为59.33和30.83,加性效应3.91和3.52,各自解释表型变异的76.9%和51.5%。利用8018株的大分离群体,发现三个性状共分离,进一步将其定位在19 kb的区域。生物信息学预测里面包含两个完整的基因和一个部分的基因。比较测序确定其中的Loc_Os07g48989是SPP7b的候选基因。RT-PCR结果显示,该候选基因确实存在转录本,因此进一步确证了该基因的真实存在。通过筛选特青的BAC文库,得到一个包含目的基因的BAC克隆12D18。酶切该BAC克隆将带目的基因的约6.9 kb片段分离克隆到载体pCAMBIA 1301上。更多还原

【Abstract】 The number of spikelets per panicle or the number of grains per panicle is an important component of rice yield. It has a theoretical and practical significance to study spikelets per panicle. However, this trait is inherited in a quantitative manner and typically controlled by a number of major and minor quantitative trait loci （QTL） and affected by environment, which causes a challenge to characterize it. With the QTL analysis based on molecular makers, the spikelets per panicle QTL can be detected. Many reports have provided the evidences for very complicated genetic bases of yield traits, which are affected by QTL, epistasis and environments simultaneously in primary mapping populations. By developing near isogenic lines （NILs） which can avoid genetic background noise, the QTL can be visualized as a Mendelian factor, and make QTL fine mapping and cloning become easier. In this study, utilizing two sets recombinant inbred lines （RILs） Zhenshan 97/Teqing and Minghui 63/Teqing, their genetic linkage map were constructed. Collecting two years’ phenotype data of two sets RILs, nine traits QTLs were detected. Eight QTL containing six spikelets per panicle QTL and two. grains per panicle QTL were developed for NILs. Genetic effects for Four QTLs’ were re-estimated in NIL background and three QTLs were fine mapped. On the basis of map Based cloning strategy, the SPP7b candidate gene was cloned.1. The genetic linkage map of Zhenshan 97/Teqing population was constructed based on 176 loci, which covered a total of 1432.1 cM with an average interval of 8.1 cM between adjacent loci; the genetic linkage map of Minghui 63/Teqing population was constructed based on 133 loci, which covered a total of 1371.4 cM with an average interval of 10.3 cM between adjacent loci. There were 50 common SSR makers between two RILs with maker orders on chromosomes well according to that published in previous reports.2. Collecting two years’ phenotype data of two sets RILs, all nine traits’ QTL containing spikelets per panicle, grains per panicle et al were detected on basis of composite interval mapping. Totally, 26 QTLs were detected in Zhenshan 97/Teqing in both the years 2004 and 2006, of which 13 QTLs were commonly detected in two years. 24 and 28 QTLs were detected in Minghui 63/Teqing in 2005 and 2006 where 13 QTLs are commonly detected in two years.3. Combining the grains per panicle and heading date phenotype data, the conditional QTL analysis was executed in two RILs. The result showed that 5 QTL of 8 GPP QTL in Zhenshan 97/Teqing were affected by heading date of rice. The other 3 QTL in Zhenshan 97/Teqing and all 5 GPP QTL in Minghui 63/Teqing were not influenced by rice flowering time. Hence, the panicle size QTL can be differentiated two types: the type I only controls grain per panicle, and type II is a flowering time affecting QTL which increase grains per panicle by elongation of life cycle.4. Eight NILs of panicle size QTL containing 6 spikelets per panicle QTL and 2 grains per panicle QTL were developed. And all NILs BC4F2 material was attained.5. The genetic effect of SPP3b was re-estimated in NIL background. In the BC₃F₂ population and BC₃F₃ progeny test population, a main effect QTL were detected, which have a LOD value of 12.8 and 8.8, additive effect of 11.89 and 7.85, and contribution of 29.1% and 20.2% of phenotype variance, respectively. Also, a main effect QTL controlling 1000-grain weight were detected in this NILs, which have a LOD value of 26.2 and 17.0, additive effect of -1.81 and -0.89, and contribution of 50.4% and 34.5% of phenotype variance, respectively. The co-segregation between the spikelets per panicle and 1000-grain weight was found by progeny test analysis, and this two QTLs considering as a maker were exactly map in the locus 1.6 and 1.0 cM away from markers RM15855 and W3D16, respectively. These results suggested that it is possible that the SPP3b was pleiotropic QTL controlling panicle size and 1000-grain weight.6. Analysing the genetic effect of SPP1 in NILs of F₂ and F₃ progeny population showed that a main effect QTL was detected, which have a LOD value of 23.95 and 35.52, additive effect of 22.05 and 10.81, and contribution of 51.1% and 63.4% of phenotype variance in F₂ and F₃ respectively. Utilizing a segregating population which has 2200 individuals, the SPP1 was fine mapped to a 107 kb region which contains 17 putative genes by bioinformatics predicting.7. In SPP6 NILs F₂ pouplation, three traits containing heading date, plant height, spikelets per panicle were co-segregated. The QTL for plant height, spikelets per panicle explained 81.2% and 50.9%of phenotype variance with a LOD value of 72.58, 29.96 and an additive effect of 7.72, 22.14 in F₂, respectively. The co-segregation among the three traits including heanding date, plant height and spikelet number per panicle was found by progeny test analysis, and this four QTLs considering as a maker were exactly map in the locus 3.1 and 3.4 cM away from markers RM19746 and RM19795, respectively. Theses results suggested that it is possible that the SPP6 was pleiotropic QTL controlling these three traits. Utilizing a segregating population which has 3300 individuals, the SPP6 was fine mapped to a 1 Mb region which contains Hd1, a cloned gene for heading date of rice. Due to the dominance of flowering in Hd1, which was counter to SPP6, it was a novel gene which is different from Hd1.8. There is a segregation of heading date, plant height and spikelets per panicle in NILs of SPP7b. By re-estimating the genetic effects of QTLs for these three traits, the main effect QTL were detected which explained 76.9%, 50.2%, 45.3% of phenotype variance with a LOD value of 59.33, 24.73, 19.22 and an additive effect of 3.91, 3.13, and 18.86 in F₂, respectively. Also in F3 progeny test population, main effect QTLs were detected which explained 51.5%, 45.0%, 50.6% of phenotype variance with a LOD value of 30.83, 24.37, 29.68 and an additive effect of 3.52, 3.22 and 17.04, respectively. Utilizing a segregating population which has 8018 individuals, the SPP7b was delimited to a 19-kb region which contains 3 putative genes by bioinformatics predicting. Sequencing comparison of 19-kb region between Zhenshan 97 and Teqing, the candidate gene of SPP7b was ascertained as Loc_Os07g48989. By RT-PCR, the result showed that Loc_Os07g48989 was expressed in Teqing genome as a real gene. A BAC clone 12D18 which contained the 19-kb region was screened to the Teqing BAC clone library. By utilizing the vector pCAMBIA 1301, the candidate gene of SPP7b was cloned.更多还原