【作者】 金亮；

【导师】 包劲松；

【作者基本信息】 浙江大学 ， 生物物理学， 2009， 博士

【摘要】 水稻的许多重要农艺性状如产量、品质、抗逆性等多表现为数量性状，是由多个基因和环境共同作用的结果，对其遗传基础的研究比较困难。水稻品种的改良有赖于所掌握资源的数量和对其农艺、品质等性状遗传基础的了解与掌握程度。近年来，随着基因组学的发展和生物统计软件的完善，特别是分子标记技术在遗传育种领域的广泛应用，以连锁不平衡（linkage disequilibrium，LD）为基础的关联分析（association analysis）方法的出现为水稻数量性状遗传的研究提供了新途径，也为作物的分子设计育种提供了新的思路，它与基于连锁作图的数量性状位点（quantitative trait locus，QTL）定位结果可相互验证、相互补充。本研究利用关联分析和QTL定位两种方法来研究稻米淀粉品质和营养品质性状的分子遗传基础，并利用分子标记辅助选择（marker assisted selection，MAS）进行稻米品质改良。其主要结果如下：1．利用全基因组上的100个微卫星标记（simple sequence repeat，SSR），对广泛收集的416份水稻材料（包括地方品种、栽培品种和育种系）进行全基因组扫描，分析群体遗传多样性、LD状况和群体结构，以构建关联定位群体。结果表明这个群体具有广泛的遗传变异。基于模型的群体SSR数据遗传结构分析发现，该群体可划分为7个亚群体。整个群体中，62．8％的SSR位点之间存在显著的LD（P＜0．05），而在各亚群中存在显著LD的位点占5．9～22．9％。测定了9个农艺性状，发现群体结构可解释6．2％～46．6％的变异，平均值为27．7％。这个群体可用于全基因范围上的SSR标记、候选基因与表型性状的关联分析。2．对淀粉品质性状与淀粉合成有关基因标记进行了基于群体结构的关联分析，结果表明在结构化的水稻群体中Wx SNP和Wx SSR与表观直链淀粉含量（AAC）、淀粉粘滞性（HPV、CPV、BD和SB）、凝胶硬度（HD）显著相关，是控制淀粉品质性状重要的位点。SSⅡa SNP与成糊温度（PT）显著相关，是控制PT的重要位点。Wx SNP、Wx SSR和SSⅡa SNP位点对于改良稻米淀粉品质有十分重要的意义。3．对淀粉品质性状与全基因组上100个SSR标记进行关联分析。结果表明，与5个淀粉品质性状显著（P＜0．001）关联的标记位点累计达到了21个。有5个SSR标记与AAC显著关联，不过这些位点的变异解释率都小于5％。其中，RM346与这5个性状都显著关联。位于第6染色体上的RM276对于PT变异的解释超过了10％。2个SSR位点（RM253、RM484）已有相应的QTL报道。这些SSR位点可能也是淀粉品质相关的基因。4．以籼粳互交加育293/Lemont的重组自交系（RIL）群体为材料，研究淀粉合成相关基因的SSR标记与淀粉品质性状的相关性。结果表明Wx和SSl（淀粉合酶1）基因的确与淀粉特性相关。而淀粉特性与SBEl（淀粉分枝酶1）等位基因没有相关性。5．通过两次回交和两次自交，并利用功能标记的辅助选择（MAS），将来自品质优良保持系宜香B（有香味，低AAC、GT）中Wx-（CT）₁₇、SSⅡa-TT和fgr（8bp缺失）等位基因导入到品质较差Ⅱ-32B（高AAC，高GT，没有香味）中，以改良其品质。利用功能标记来选择具有Wx-（CT）₁₇、SSⅡa-TT和fgr等位基因型的植株。最终选出17株具有Wx-（CT）₁₇、SSⅡa-TT和fgr纯合基因型的株系。改良后的Ⅱ-32B具有香味和较低的AAC和GT。6．对稻米颜色参数和营养品质性状与控制种皮颜色的Rc、Ra，香味基因fgr、淀粉主效基因位点以及100个SSR标记位点进行基于群体结构的关联分析。结果表明，对于颜色参数性状，Ra、Rc确实是主效基因，而RM316可能是一个影响较大的位点。与3个营养品质性状显著关联的标记位点累计达到15个。对于多酚含量、类黄酮含量和抗氧化活性，Ra对表型变异解释率最大，都超过44％。而对于多酚含量和抗氧化活性Rc对表型变异解释率仅次于Ra，也分别达到了27．39％和23．09％。而RM316对于3个性状的表型变异解释率都超过10．0％，RM316可能是一个重要的未知的QTL。7．对本群体中361份白米材料的颜色参数及营养品质性状与100个SSR标记，以及fgr和淀粉主效基因位点进行关联分析。发现累计34个位点与5个颜色性状显著关联（P＜0．001）。不过各个位点的贡献率都比较小，表明在白米材料群体中，其颜色参数性状受微效多基因（位点）控制，其遗传方式比较复杂。对白米材料的营养品质性状的关联分析表明，累计6个位点与多酚含量和抗氧化活性性状显著关联（P＜0．001），共解释超过30％的表型变异。没有检测到与类黄酮含量显著关联的位点。对于多酚含量RM336对表型变异解释率最大，为7．16％；对于抗氧化活性，只检测到一个显著关联的位点RM251，其表型变异解释率为6．10％。对于白米材料颜色参数以及营养品质性状的关联分析表明，颜色参数性状受多微效基因控制，而影响多酚含量和抗氧化活性位点的效应也较小。8．以典型的籼粳交窄叶青8号/京系17的加倍单倍体（DH）群体及其分子图谱，对稻米颜色参数、多酚、类黄酮含量以及抗氧化能力性状进行了QTL定位分析。在这个群体中，这些性状均呈连续性分布，并存在一定数量的超亲分离。对于5个颜色参数性状，总共检测出12个主效QTL，其中qL-2、qB-2和qC-2都位于第2染色体上的同一区域。对于多酚含量性状检测到3个主效QTL，分布在第2、4、12染色体上，总共解释超过43％的表型变异。对于类黄酮含量检测到3个QTL，分布在第2、11染色体上，总共解释超过26％的表型变异。对于抗氧化能力性状检测到3个QTL，分布在第1、7、11染色体上，总共解释超过32％的表型变异。更多还原

【Abstract】 Most agronomic traits such as yield, quality and stress-tolerance of rice are quantitativetraits. Dissection of the genetic basis for these traits is difficult, because they are controlledby multi-genes that are affected by environmental factors, Rice improvement and dissectionof complex traits require germplasm diversity and understanding the genetic basis of thesetraits. With the development of genomics and biostatistics software and applications ofmolecular markers in crop breeding, association analysis based on the linkage disequilibrium（LD） offers a new method for identifying of loci controlling those traits. Combination ofquantitative trait loci （QTL） mapping and association analysis facilitates dissection ofcomplex traits, which results in important information and markers for molecular breeding bydesign in crops. In the present study, association analysis （mapping） and QTL mapping wereused to study genetic basis for starch quality and nutritional quality of rice grain, facilitatingmolecular improvement of rice quality. The main results are summarized as follows:1. One hundred genome-wide simple sequence repeat （SSR） markers were used toassess the genetic diversity, population structure, and LD of a set of 416 rice accessionsincluding landraces, cultivars and breeding lines. Results showed that the population includeda diverse genetic variation. A model-based population structure analysis subdivided thepopulation into 7 clusters. LD was significant at a 0.05 level among 62.8% of the SSR pairsin the entire sample and with a range of 5.9～22.9% in each cluster. On average, theestimates of population structure account for 27.7% of phenotypic variation, ranging from6.2% to 46.6%. The results suggested that the population could be used for detection ofgenome-wide SSR marker-phenotype association mapping.2. Association analysis was carried out between the starch gene markers and starchquality traits based on population structure. Results confirmed that Wx SSR and singlenucleotide polymorphism （SNP） were strongly associated with apparent amylose content（AAC）, pasting viscosities （HPV, CPV, BD and SB）, gel hardness （HD）, whereas the SSⅡaGC/TT SNPs were strongly associated with the pasting temperature （PT）. These markers areuseful in molecular breeding for improvement of rice eating and cooking qualities.3. In addition, a genome-wide scan was performed using 100 SSR markers to identifyQTL associated with starch quality in rice. A total of 21 SSR loci were found to besignificant associated （P＜0.001 ） with 5 starch quality traits. Five SSR loci were found to besignificant associated with AAC, but all the five SSR loci explained less than 5% of the totalphenotypic variations. The RM346 on chromosome 7 was associated simultaneously with the 5 traits. RM276 on chromosome 6 explained more than 5% of the total phenotypic of PT.Two associated markers （RM253, RM484） were located in regions where QTL hadpreviously been identified. These results suggested that the association analysis approachcould be a useful alternative to linkage mapping for the identification of unreported regionsof the rice genome containing putative QTL of starch quality.4. A recombinant inbred line （RIL） population derived from the reciprocal cross ofLemont （a premium high-quality tropical japonica rice）and Jiayu 293 （a high-yield butlow-quality indica rice） was used to test the association of microsatellite markers ofstarch-synthesizing genes with starch quality parameters. The results confirmed theassociation of Wx and starch synthase 1 （SSl） alleles with various starch properties measuredin rice flour. However, the starch properties were not associated with the starch branchingenzyme 1 （SBE l） gene alleles.5.Ⅱ-32B is a key Chinese maintainer line used for hybrid rice breeding. However, it isof poor quality because of a high AAC, a high gelatinization temperature （GT）, andnon-fragrant. It is known that the AAC, GT and fragrance traits are controlled by the Wx,starch synthaseⅡa （SSⅡa） and fragrance （fgr） genes. The quality ofⅡ-32B was improved byintrogressing the Wx, SSⅡa and fgr genes from YixiangB, another maintainer line, that has alow AAC, a low GT and fragrant sensory, with the assistance of the functional markers. TheWx microsatellite [（CT）₁₇ allele] SSⅡa single nucleotide polymorphism （TT allele） and fgrinsertion/deletion allele （8 bp）, were transferred toⅡ-32B by two backcrosses and twoselfings. Molecular marker-assisted selection was applied in the series to select forindividuals carrying Wx-（CT）₁₇, SSⅡa-TT and fgr alleles. According to the marker genotypes,17 homozygous lines for Wx, SSⅡa and fgr gene were finally selected. The improvedⅡ-32Blines were fragrant with reduced AAC and GT.6. Association analyses were performed between the Rc, Ra, fgr and starch gene markersto grain color and nutritional quality traits in the 416 rice accessions. The result showed thatRc and Ra gene markers were strongly associated with grain color traits, whereas the RM316might be a putative QTL of grain color. A total of 15 SSR loci were found to be significantassociated （P＜0.001） with 3 nutritional traits i.e. total phenolics content, flavonoids contentand antioxidant capacity. Ra explained more than 44% of the total phenotypic variations forthe three nutritional traits. For total phenolics contents and antioxidant capacity, Rc explained27.39% and 23.09% of the total phenotypic variations, respectively. RM316 explained morethan 10% of the total phenotypic variations for the three nutritional traits. The resultconfirmed that Ra and Rc were main-effect loci for rice grain color and nutritional quality traits, whereas RM316 might be an important putative QTL.7. Association analysis between the fgr, starch gene markers and SSRs and grain colorand nutritional quality traits in the 361 white rice accessions were also performed. A total of34 markers were found to be significant associated （P＜0.001 ） with the five grain color traits,which showed these traits were controlled by polygenes with minor effects. Five SSR lociwere found to be significant associated （P＜0.001 ） with total phenolics contents, whichexplained more than 30% of the total phenotypic variations. Among them, RM336 explained7.16% of the total phenotypic variations of total phenolics content. RM251 was the onlylocus associated with antioxidant capacity, which explained 7.16% of the total phenotypicvariations. No SSR locus was found to be significant associated with flavonoids content.8. Mapping of quantitative trait loci （QTL） for grain color, Total phenolics, flavonoids,antioxidant capacity was carried out using a doubled haploid （DH） population derived from across between indica variety Zaiyeqing 8 （ZYQ8） and japonica variety Jingxi 17 （JX17）. Theresults indicated that all three parameters were continuously distributed among the DH lines,but many DH lines showed transgressive segregation for all traits. A total of 12 QTLs wereidentified for five color parameters, three of which, qL-2, qB-2 and qC-2 shared the sameregion on chromosome 2. A total of three QTLs on chromosome 2, 4, 12 were identified fortotal phenolic content, which explained more than 43% of the total phenotypic variations.Three QTLs on chromosome 2, 11 were identified for total flavonoid content, whichexplained more than 26% of the total phenotypic variations. Three QTLs on chromosome 1, 7,11 were identified for antioxidant capacity, which explained more than 32% of the totalphenotypic variations.更多还原