【作者】 沈新莲；

【导师】 张天真；

【作者基本信息】 南京农业大学 ， 作物遗传育种， 2004， 博士

【摘要】 棉花是重要的纺织工业原料。棉花工作者长期以来面临的挑战是同步提高产量和纤维品质以满足棉花生产者和纺织工业的要求。随着纺织技术的不断发展，对棉纤维品质提出了更高的要求，从而更多地要求更强，更细和更整齐的棉纤维。然而，由于纤维品质性状的复杂性，加上高的费用和低的选择效率，棉纤维品质的改良虽经过几十多年的艰苦努力，纤维品质与产量的负相关依然存在，传统育种在进一步改良棉纤维品质上显然存在相当的困难。随着分子生物学研究的进展，发展与棉纤维品质性状连锁的DNA标记使得育种者在棉花生长的早期阶段或早期分离世代就能追踪这个重要性状，从而提高纤维品质的选择效率。 然而目前有关棉纤维品质QTLs的筛选存在以下问题：（1）使用的标记多为操作复杂、成本高的RFLP标记，直接利用这些标记进行辅助选择比较困难；（2）使用的群体多为海陆种间群体，用于陆地棉品质改良难度大；（3）使用的群体都为F₂暂时性群体，无法进行多年多点重复试验，难以获得大量稳定的QTLs，进行QTL×QTL、QTL×环境互作研究难度大。针对这些问题，本研究设计了如下总体思路：用三个不同来源的陆地棉高强纤维种质系构建三个F₂及F_2：3群体，用基于PCR的SSR标记筛选纤维品质QTL，获得不同遗传背景下、不同世代稳定表达的QTLs；构建一个RIL群体，并开展多年多点的重复试验，筛选不同环境中稳定表达的土效QTLs，研究QTL×QTL、QTL×环境互作效应；用所获得的与纤维比强度QTL连锁的分子标记开展比强度的分子标记辅助选择，探讨MAS在棉纤维品质改良中的作用。结果如下： 1 不同遗传背景下纤维品质QTLs的筛选 用三个陆地棉高强纤维种质系构建了三个F₂种內连锁图，分别覆盖了666.7cM，557.8cM和588cM的遗传距离，相当于棉花基因组的14.82％，12.4％和13.07％。用复合区间作图法共筛选到38个与纤维品质有关的QTLs，其中纤维长度11个，纤维比强度10个，麦克隆值9个，纤维伸长率8个。15个稳定的QTLs（39.47％）能同时在F₂和F_2：3检测到，至少3个QTLs能在二个群体中表达，不同遗传背景下共同QTLs的发现表明现有的优质纤维基因可能有相同的来源。三个群体中都发现增效基因的方向与期望的表型不一致的现象，但大部分棉纤维品质的增效基因来自优质亲本，显示了优质纤維基因资源在棉纤维品质改良中的重要作用。另外，发现了3对可能的部分同源QTLs，分别是：Chr7-Chr16上的纤维长度QTLs，LGA02-LGD03上的纤维比陆地棉纤维品质QTL的筛选、定位及其应用强度和纤维伸长率QTLs。推测种质渐渗过程中，染色体交换可能在同源区域随机发生。2陆地棉重组自交系的构建及纤维品质、产量性状QTL的筛选 以7235和TM一1为亲本，采用F:衍生家系法，构建了一套陆地棉重组自交系 （Reeombined inbred lines）群体，并进行了二年二点的重复试验，RIL群体的纤维品质及产量性状在每个环境中都表现极强的超亲分离现象，显示了亲本间的遗传多样性和重组自交系的特点。110个SSR多态性位点产生了27个连锁群，共覆盖810.07cM的遗传距离，大约相当于棉花基因组的18.02%。 用复合区间作图法对班L群体的所有性状按单环境分析，共筛选到35个纤维品质QTLs，25个产量性状QTLs，38（63.33%）个QTLs至少能在二环境中检测到，纤维比强度、麦克隆值、铃重、衣分QTLS显示了最好的稳定性，这些性状的所有QTLs至少能在二环境中检测到，，而且不同环境中QTLS的位置也较一致。用四环境或二环境性状的平均数进一步进行QTLs作图，所获得的QTLs的稳定性进一步提高，说明采用多环境平均数进行QTLs作图是获得稳定QTLs的可靠方式。 多数产量性状QTLs与品质性状QTLs位于相同或相邻的区间，从而造成了产量与品质的强烈负相关。最明显的例子是连锁群D03的中、下部区间，中部7cM区间里含有纤维比强度、麦克隆值、成熟度、铃重、籽指、单株结铃性、籽棉产量和皮棉产量QTL各一个，下部7.6cM的区间里含有纤维比强度、麦克隆值、成熟度、黄度、铃重、籽指、单株结铃性、籽棉产量和皮棉产量QTL各一个，7235等位基因在这些位点对纤维品质性状有增效作用，对产量性状有减效作用。表明产量与品质的负相关是连锁或连锁和一因多效共同作用的结果。MAS在棉纤维改良中最大的作用可能是鉴别重组、降低连锁累赘。3纤维品质与产量性状的上位性及QE互作分析 用基于混合线性模型的复合区间作图（QTLMaPerl.6）进行多环境联合分析，共获得加性效应QTLsZI个，互作效应QTLsZI对，三类上位性效应（主效QTLs间的互作，主效QTL与背景位点间的互作和两个互补性位点间的互作）都存在;有12个性状检测到了上位性效应QTL，说明上位性在纤维品质与产量性状中是普遍存在的，大部分上位性位点本身无加性效应;短纤维指数和纤维整齐度的上位性效应比较显著，纤维比强度上位性效应较小，纤维长度加性效应和上位性效应可能同样重要。 用基于混合线性模型的复合区间作图（QTLM即erl .6）进行Q”E互作分析，只有7个性状表现Q“E互作，其中纤维品质性状4个，产量性状3个。Q xE互作主要有三种类型:a.不同环境中表现不稳定的QTL表现Q/E互作，如纤?更多还原

【Abstract】 Cotton is an important cash crop in the world. A long-term challenge facing cotton breeder is the simultaneous improvement of yield and fiber quality to meet the demands of the cotton producer as well as the textile industry. In the recent years, improvement of cotton fiber quality has been extremely important because of changes in spinning technology. However, a negative association between lint yield and fiber quality are still presented after over dozens years of exhausting breeding for improved fiber properties due to the genetic complexity of fiber quality properties. Conventional breeding procedures exist difficulty in further improving fiber quality because of its high costs, long duration, and low selective efficiency. The development of DNA markers linked to the fiber quality QTL would allow cotton breeders to trace this very important trait in early plant-growing stage or early segregating generations. The use of these DNA markers is increasing the prospect for streamlining the cotton breeding programs for improving fiber quality while maintaining fiber yield.However, QTL analysis for fiber properties is problematic for several reasons: (1) RFLP with complicated procedure and high cost were used in majority of previous reports. It is very hard to use it directly in MAS programm; (2) QTLs obtained from interspecies population of Sealand cotton and Upland cotton showed less valuable in improving fiber quality of Upland cotton; (3) In previous reports, population constructed were all F2 population. Replicated experiment couldn’t be carried out. Analysis of epistasis and QE interaction effect remains to be difficulty. Considering of these problems above, the objective of this research is to determine the presence of common QTL associated with a given trait across different background; to develop a RIL population and identify stable QTL over environments; to analysis AA and AE interaction effect; to preliminary study on MAS on fiber strength.1. QTLs mapping for fiber properties under different back groundingUsing three elite fiber strains of upland cotton (Gossypium hirsutum L.) as parents, three F2 linkage maps were constructed. They covered 666.7cM, 557.8cM and 588cM, approximately 14.82%, 12.4% and 13.07% of the total recombinational length of the cotton genome, respectively. There are 38 QTLs for fiber traits in which 11 QTLs were for fiber length, 10 for fiber strength, 9 for micronaire, and 8 for fiber elongation detected by composite interval mapping. Among them, 15 stable QTLs (39.47%) could be found in F2 and F2:3 simultaneously. At least 3 QTLs could be identified at two populations. Characterization of identical QTL at different populations indicates exiting elite fiber gene have possibly the same origin. In addition, we found three pairs of putative homoeologousQTLs, qFL-7-1c and qFL-16-lc, qFS-D03-la and qFS-A02-lb, qFS-A02-lc, and qFE-D03-la and qFE-A02-lc. It was assumed recombinational event occurred randomly in A or D subgenome homoeologous region during the process of introgression.2. Development a RIL population and QTLs mapping for fiber quality and yield traits under different environmentsBy means of bulk-selfing method, a RIL population was developed from individuals of (7235 TM-1) F2 in Upland cotton. Repeated experiment was conducted at two locations in 2002 and 2003. Wide variation occurred among RI lines and transgressive segregation was observed across all environments for all traits. The total of 110 SSR loci was grouped into 27 linkage group and covered 810.07 cM, approximately 18.02% of the total recombinational length of the cotton genome.A total of 35 QTLs for fiber quality and 25 QTLs for yield traits were detected in two or four environments independently. Thirty QTLs were found in the joint analysis based on means over two or four environments. QTLs obtained based on means over several environments seems to provide the best strategy for obtaining stable QTL.Most QTLs for fiber quality and yield traits were located at the same interval or neighbor interval, especially at LGD03.更多还原