【作者】 郝德荣；

【导师】 喻德跃；

【作者基本信息】 南京农业大学 ， 作物遗传育种， 2011， 博士

【摘要】 大豆[Glycine Max (L.) Merr.]是重要的粮食和经济作物,其产量的形成与许多重要的农艺性状和经济性状直接或间接相关,如开花期、成熟期、产量组分以及光合性能等。这些性状均属复杂的数量性状,应用传统育种方法进行遗传改良已经越来越难。随着生物技术的发展,利用分子育种技术培育高产大豆品种成为现代大豆育种计划的主要内容之一。通过鉴定大豆基因组中与大豆产量相关性状紧密关联的功能位点或单倍型,开发新的功能标记对目标性状进行分子标记辅助选择,或利用大豆产量相关功能基因进行转基因育种,可大大提高大豆高产育种效率,达到定向选育的目的。为了发掘大豆基因组中控制大豆产量相关性状的功能基因/QTL,本研究在大豆产量性状一致性QTL候选区域内利用SSR等标记进一步对产量性状QTL进行了关联定位,并对大豆产量、产量组分及光合相关性状进行了基于SNPs及单倍型的全基因组关联分析。在此基础上,结合大豆基因组序列信息,利用生物信息学方法筛选与大豆产量相关的候选基因,通过候选基因关联分析策略验证其功能、鉴定出优异单倍型,为大豆高产分子育种提供有效的依据。主要结果如下：1.利用关联分析方法在大豆产量性状QTL候选区域内对大豆产量性状QTL进行了关联定位。结果表明,产量相关性状在196份栽培豆组成的自然群体内存在广泛的表型变异；除了百粒重与花期、株高间相关不显著外,其余性状间均存在极显著相关；染色体6、7、19上共线性或非共线性位点组合均存在一定程度的LD,并且LD衰减速率较快；基于MLM模型,在5个环境中共检测到40个标记位点与产量相关性状显著关联,其中有26个标记位点在所有环境中均能被稳定检测到,7个标记位点在所有环境中同时与2个或2个以上产量性状共关联；稳定关联的标记位点等位变异数目差异较大,而且不同等位变异对表型的效应差异也较显著,在高产育种中可利用不同位点的优异等位变异进行优良性状聚合育种。位于染色体7上的Satt150在5个环境中均被稳定检测到与产量及其他相关性状存在显著关联,并且与国内外多次报道的主效QTL位点一致。基于连锁不平衡的原理,将大豆产量性状QTL进一步缩小到Satt150两侧标记518kb区间内(GMES2109-Satt316),为下一步进行产量相关候选基因的筛选和功能标记辅助育种奠定了基础。2.为了进一步发掘大豆基因组内与产量及产量组分存在关联的功能区域,本研究以191份栽培豆为研究群体,通过2年5环境试验,利用1536个SNPs和209个单倍型,对大豆产量及产量组分进行了全基因组关联分析。研究结果表明,191份栽培豆组成的自然群体内存在广泛的表型变异和遗传多样性；利用1142个SNPs(最小等位基因频率大于10%)将该群体聚为两个亚群,个体间不存在或仅存在较弱的亲缘关系,群体内位点间LD衰减距离大约为500kb。全基因组关联分析共检测到19个SNPs和5个单倍型在3个或3个以上环境中与单株荚数、单株粒数、百粒重和产量存在显著关联,部分显著关联的位点位于前人利用连锁分析检测到的QTL内或与其紧密相邻。在显著关联的位点中,检测到9个标记位点与2个以上不同的产量性状同时关联；稳定检测到4个SNPs和1个单倍型位点在所有环境中均与百粒重存在显著关联,并且各位点等位变异所对应的百粒重均值差异极显著,优良等位变异与较差等位变异间的百粒重均值差异达4g。这些位点均有助于更好地揭示大豆产量及产量组分的遗传机制,为将来通过分子育种技术改良大豆产量,实现高产稳产奠定了基础。3.根据大豆产量性状QTL候选区域关联定位及全基因组关联分析结果,结合大豆基因组序列信息,利用生物信息学手段,在染色体7上筛选到大豆产量相关的候选基因GmGA3ox.通过对191份栽培豆中的GmGA3ox序列进行多态性分析,共检测到的16个SNPs和4个Indel(最小等位基因频率大于10%),并且存在5个明显的单倍型域。关联分析共检测到5个SNP/Indel在3个以上环境与产量间存在显著关联,其中有2个多态性位点在所有环境中均被检测到与产量显著关联。在3个以上环境中,共检测到1个单倍型(GmGA3ox_H4)与百粒重存在显著关联,2个单倍型(GmGA3ox_H4、GmGA3ox_H5)与产量显著关联。其中,GmGA3ox_H4在所有环境中均被检测到与百粒重存在显著关联,对百粒重的表型变异解释率达9.67%,最优单倍型与最差单倍型间百粒重差异达4g；GmGA3ox_H5则在所有环境中均与产量显著关联,对产量的表型变异解释率达20.19%,最优单倍型与最差单倍型产量差异达1倍；序列结构分析发现GmGA3ox_H4和GmGA3ox_H5均位于3’非编码区内,可能参与mRNA的转录后调控,引起GA代谢途径及相关代谢通路的变化,从而引起大豆百粒重及产量的改变。4.对叶绿素含量(CCI)、叶绿素荧光参数(JlP-test参数)及光合速率(PN)进行的全基因组关联分析结果表明,2009年和2010年研究(光合速率为2008年)中,共检测到47个SNPs、11个单倍型与CCI显著关联,30个SNPs、8个单倍型与Fv/Fm显著关联,33个SNPs、5个单倍型与ABS/RC显著关联,25个SNPs、5个单倍型与ETo/TRo存在显著关联,20个SNPs、7个单倍型与PIABS存在显著关联,15个SNPs、2个单倍型与PN显著关联。在所有检测到的位点中,共检测到26个SNPs位点同时与2个或2个以上的性状存在显著关联。其中,4个与JIP-test参数关联的位点同时与PN共关联；12个SNPs和2个单倍型与大豆产量或产量组分共关联。共稳定检测到8个SNPs和2个单倍型在两年中均与相应的光合性状存在显著关联,稳定关联的SNP及单倍型的等位变异对光合性状的表型效应差异均达极显著水平,在大豆分子育种中,利用这些稳定关联的位点的优异等位变异进行MAS,可能会达到同步改良大豆光合性能和提高产量的效果。更多还原

【Abstract】 Soybean [Glycine max (L.) Merr.] is an important food and cash crop. Seed yield of soybean is directly or indirectly determined by many agronomic and economical traits, such as plant height, flowering time, maturity, yield components and photosynthetic performance. These traits are controlled in a complex manner by quantitative trait loci (QTLs), and environmental variations can trigger and modify the actions of related genes. In soybean, it is becoming more difficult to improve yield using traditional breeding methods. The development of genomics has provided alternative tools to improve breeding efficiency in plant breeding programs. Use of molecular breeding techniques to breed high-yielding soybean varieties has become one of the main objectives in soybean modern breeding programs. Using the functional markers located in or closely linked to the gene or QTL underlying the target trait in marker-assisted selection (MAS), or use of functional genes in transgenic breeding, can greatly increase breeding efficiency and directionally improve soybean yield.In present study, in order to identify the causal polymorphisms underlying the soybean yield-related traits, association analysis between SSR markers and soybean yield-related traits was conducted in candidate regions of major QTLs of soybean yield-related traits. Then genome-wide association analysis was performed in different environments to identify SNPs and haplotypes underlying soybean yield, yield components, and photosynthesis-related traits,. On this basis, combined soybean genomic information with bioinformatics, yield-related candidate gene was screened, and thus, candidate-gene association analysis was conducted to identified the polymorphisms and haplotypes in the yield-related candidate gene. The main results of this study are as follows:Using80SSR/EST-SSR/DT1markers, we conducted fine mapping of soybean yield-related traits based on the association analysis across five different environments in the candidate regions of major QTLs of soybean yield-related traits. The results showed that there was extensive phenotypic variation in196soybean landraces. Except for seed weight, days to flowering and plant height, the other traits were significantly related. The analysis indicated that there was extensive linkage disequilibrium intra-chromosome or inter-chromosome among chromosome6,7and19, and that the LD decayed rapidly in our studied population. Based MLM model, we detected40markers significantly associated with yield-related traits across five environments. Of which,26loci can be detected in all environments.7loci were co-associated with two or more different traits in all environments. The locus of Satt150on chromosome7was detected in all environments associated with soybean yield and other related traits. Moreover, this locus was detected repeatedly associated with soybean yield and related traits in previous reports. Based on the principle of linkage disequilibrium, the QTL underlying the soybean yield traits was fine mapped in the interval of518kb between the flanking markers on both sides of Satt150(GMES2109-Satt316).In order to identify the causal polymorphisms underlying soybean yield and yield components, in this study, we further evaluated a group of191soybean landraces in five different environments using1536single-nucleotide polymorphisms and209haplotypes. The analysis revealed that there was abundant phenotypic and genetic diversity in the studied population. The191soybean landraces could be divided into two subpopulations using1142SNPs with minor allele frequency of≥10%. No or weak relatedness was detected between pair-wise accessions within this population. The average decay distance of linkage disequilibrium intra-chromosome was estimated at approximately500kb. Genome-wide association analysis based on a unified mixed model approach identified19SNPs and5haplotypes associated with number of pods per plant, number of seeds per plant,100-seed weight, and seed yield in three or more different environments. Nine markers were co-associated with two or more different traits. Many markers were located in or close to quantitative trait loci mapped by linkage analysis in previous reports. The SNPs and haplotypes identified in this study will help to further understand the genetic basis of soybean yield and its components. This information lays the foundation for high-yield molecular breeding of soybean.According to the results of fine mapping for QTLs underlying soybean yield-related traits in the candidate regions of major QTLs of soybean yield-related traits and the genome-wide association analysis, using the soybean genomic sequence information and bioinformatics tools, we screened a candidate gene GmGA3ox on chromosome7, which may be associated with soybean yield-related traits. The sequence polymorphism analysis of GmGA3ox in191soybean landraces identified a total of16SNPs and4Indels (with minor allele frequency of≥10%), and these loci consisted five distinct haplotype blocks. Candidate-gene association analysis showed that5SNP/Indels were associated with seed yield in three or more environments. Of which,2loci were associated with seed yield in all environments. In present study, one and two haplotypes were significantly associated with100-seed weight and seed yield in three or more environments, respectively. Of which, GmGA3ox_H4was identified in all environments significantly associated with100-seed weight, with contribution rate of9.67%, the difference of seed weight between the optimal haplotype and worst haplotype was about4g. GmGA3ox_H5was significantly associated with the seed yield in all environments with contribution rate of20.19%. The yields of the optimal haplotype of GmGA3ox_H5were more than1fold to the worst haplotype. Sequence analysis suggested that GmGA3ox_H4and GmGA3ox_H5were located in the the3’non-coding region, and might be involved in mRNA post-transcriptional regulation. This will lead to a series of changes in metabolic pathways of GA, and results in differences in seed weight and seed yield within the studied population.Based on MLM, genome-wide association analysis was conducted to detect the genetic polymorphisms for chlorophyll content (CCI), chlorophyll fluorescence parameters (JIP-test parameter) and the photosynthetic rate (PN).In present study,123SNPs and22haplotypes were detected significantly associated with photosynthesis-related traits. Of which,47SNPs and11haplotypes were associated with CCI,30SNPs and8haplotypes were associated with Fv/Fm,33SNPs and5haplotypes were associated with ABS/RC,25SNPs and5haplotypes were associated with ETo/TRo,20SNPs and7haplotypes were associated with PIABS.15SNPs and2haplotypes were associated with PN in2008. Among all detected loci,26SNPs were co-associated with two or more traits or parameters. Of which,4loci associated with the JIP-test parameters were identified co-associated with PN,12SNPs and2haplotypes were co-associated with soybean yield or yield components.8 SNPs and2haplotype were detected stably associated with two or more different traits or parameters in two years. Use of these stablely associated loci in soybean molecule breeding, will improve photosynthetic performance and increase soybean yield synchronously.更多还原