【作者】 文雁成；

【导师】 傅廷栋；

【作者基本信息】 华中农业大学 ， 作物遗传育种， 2012， 博士

【摘要】 在收获过程中油菜裂角落粒所造成的产量损失率通常在10%-30%。为了减少收获过程中的损失,培育抗裂角、适合机械化生产需要的油菜品种已经成为当务之急。甘蓝型油菜是我国生产上应用的主要油菜类型,育种实践要求首先进行甘蓝型油菜抗裂角的基础理论研究。本研究主要内容包括：建立抗裂角检测体系和量化指标、筛选抗裂角种质资源、研究影响抗裂角的因素、构建抗裂角作图DH群体和对该性状进行QTL定位。主要研究结果如下：1甘蓝型油菜抗裂角检测方法的建立和抗裂角品种(系)的筛选1.1对随机碰撞法进行了改进,采用抗裂角指数作为抗裂角性状的定量指标,为该性状的定量研究奠定了基础。研究发现抗裂角指数在甘蓝型油菜种质资源中存在广泛变异(0.0000-0.7675),变异系数高达114.4%。在229份品系中,极易裂角和易裂角品系分别占59.38%和32.75%。筛选出抗裂角品系(抗裂角指数在0.7以上)2个(H155和98009)。1.2研究表明,通过改进后的随机碰撞测试(RIT)检测的抗裂角指数是稳定可靠的。品种(系)间的抗裂角指数差异达到极显著水平,重复间、年份间以及品种(系)与年份的互作差异不显著。1.3相关分析结果表明,决定角果空间分布的性状(果柄与主花序轴朝上夹角、果柄与角果朝上夹角和果柄长度)与抗裂角指数间不存在相关性。角果密度与抗裂角指数间存在极显著的负相关。抗裂角指数与角果长度、角果宽度、角喙长度、角皮厚度和角粒数间均呈现极显著的正相关。1.4抗裂角品种与易裂角品种的角皮结构存在明显差异。抗裂角品种H155的壳状果瓣的内皮层细胞排列紧密、木质化程度高、多层薄壁细胞也发生了木质化,而且其壳状果瓣中维管束多、木质化程度高增强了壳状果瓣的机械强度。易裂角品种Qva的壳状果瓣比抗裂角品种的壳状果瓣的机械强度低,更易于破裂。2影响甘蓝型油菜角果抗裂特性的因素分析2.1遗传因素是决定抗裂角指数差异的决定因素。研究表明,相同品种(系)抗裂角指数在重复间差异不显著,品种间差异极显著,年份和地点间差异也不显著。2.2不同部位的角果抗裂角指数存在差异,变异系数在18.67%-93.57%之间。多数品种都是以下部第一分枝角果的抗裂指数最大。从第二分枝往上的分枝的角果的抗裂角指数差异不大。主花序中下部角果抗裂角指数相近,顶部角果的抗裂角指数分别比中部和下部角果小55.04%和60.43%。多数品种的主花序下部角果抗裂能力最强。2.3抗裂角品种(系)与易裂角品系在含水量降低的过程中,抗裂角指数的变化存在差异。在角果脱水过程中,易裂角品系Qva的抗裂角指数迅速下降,而抗裂角品系在脱水初期和后期抗裂角指数下降都较慢。研究发现角果相对含水量与抗裂角指数间呈一次直线回归关系(y=0.0149x-0.4779,R2=0.4284),即油菜的抗裂角指数随角果相对含水量的增加直线上升。3油菜游离小孢子培养技术的改进3.1将取样时间从初花后提前到初花前9天,产胚量达到15.60-34.08个/花蕾,与传统取样时间产胚量没有显著差异。采用在活体植株上摘取适合大小的花蕾替代在室内折断的花序上摘取花蕾,二者培养的出胚率差异不显著。但是该方法可以让较小的花蕾继续生长,大幅度减少供体植株数量,保证取样花蕾具有相同遗传背景。3.2利用13%蔗糖溶液替代传统的B5培养基作为小孢子提取液,二者提取的小孢子出胚率差异不显著。可以节省试剂和配置培养基所消耗的人力。改进后的培养程序直接将悬浮在NLN培养基中的小孢子置于32℃的培养箱中暗培养至有胚状体出现,使得出胚时间较传统方法提前4天,而胚状体产量几乎相同。4甘蓝型油菜遗传连锁图谱构建及抗裂角性状的QTL定位4.1H155(P1)、P2(Qva)及F1、F2、BC1F1、BC2F1的抗裂角指数呈连续分布。DH系的抗裂角指数的次数分布偏向易裂角亲本,呈现连续复合正态分布,抗裂角性状是受微效基因控制的数量性状。4.2利用JoinMap3.0进行连锁分析,构建了一张包含175个标记位点的遗传连锁图谱。这些标记分布在19个连锁群上,覆盖甘蓝型油菜1382.8cM,平均标记间距7.9cM。连锁群的遗传距离从33.5cM到107.7cM不等,连锁群上的标记从4个到20个4.3当LOD>2.0时,在郑州点检测到9个抗裂角性状有关的QTL,分别位于N1、N7、N8、N15和N18五个连锁群上,对表现型的总贡献率达到49.0%。这9个QTL间存在上位性互作,对表现型的贡献率达到45.9%。郑州点QTL及其互作效应对表现型的贡献率为94.9%。在武汉试点检测到4个与抗裂角指数有关的QTL,它们分别位于N1、N4、N7和N18四个连锁群上,对表现型的总贡献率为38.6%。这4个QTL间存在上位性互作,而且对表现型的互作达到12.8%。武汉点QTL及其互作效应对表现型的贡献率为51.4%。4.4在郑州检测出的3个QTL在武汉被重复检出。在郑州检测出的位于N1连锁群的qSSRI2在武汉重复检出(qSSRI10)。在郑州检测出的位于N7连锁群的qSSRI4也在武汉重复检出(qSSRI12)。在郑州检测出的位于N18连锁群的qSSRI9在武汉重新检出(qSSRI13)。4.5本研究定位的QTL结果与Raman等(2011)利用关联分析检测出的SSR标记结果一致性较好。而Raman等(2011)的结果与拟南芥和芸薹属作物的育角果开裂有关的转录组学分析结果一致(Jaradat et al.,2010)。更多还原

【Abstract】 During harvesting of rapeseed, the silique shatter loss proportion usually reaches up to10%to30%. In order to decrease the yield loss caused by silique shatter, it is necessary to breed varieties with silique-shatter resistance, which are suitable for combine harvesters. Since Brassica napus is the main rapeseed type used in production in China, breeding practice requires that theoretical researches of silique-shattrer resistance be carried out in advance.This study firstly improved random impact test (RIT) and set up silique-shatter resistance index (SSRI) as a measurement for this trait. Then a slilque-shatter resistant line H155and a slilque-shatter susceptible line Qva were selected from the Brassica germplasm. Silique-shatter resistant trait and their influencing factors were also studied.During construction of DH mapping population, microspore cultural techniques were improved for higher efficiency and lower costs. A linkage map with19linkage groups was constructed and13QTLs which were related to silique-shatter resistance were mapped. These results can be helpful in molecular assistant selection in rapeseed silique-shatter resistant breeding.The main results in this study are listed as followings.1Setting up a test system for silique-shatter resistance and screening silique shatter resistant lines1.1Random impact test (RIT) was improved and silique-shatter resistance index (SSRI) was also firstly set up for evaluating silique-shatter resistant ability in rapeseed. Results indicated that there were extensive variance of SSRI in Brassica napus germplasm, which SSRI ranged from0.000to0.7675and the coefficient of variance (CV) was as high as114.4%. In the screened germplasm, about59.38%were silique-shatter sensitive accessions,32.75%were silique-shatter susceptible accessions. Only4.01%accessions with more than0.5of SSRI. Two silique-shatter resistant accessions (H155and98009) with more than0.7of SSRI were obtained.1.2Results showed that SSRI detected by RIT were stable and credible.The differences of SSRI among accessions reached1%significant level. However the differences among years and replications did not reached significant level.1.3Results of simple correlation analysis revealed that SSRI had no correlation with the angle between stalks and inflorescences, angles between stalks and silique and stalk length.All these parameters decide silique distribution in space and are directly related to scraping and silique shatter loss during harvesting. Silique number per cm along inflorescences and branches had significant (1%) negative correlation with SSRI. Silique length, silique width, beak length, valve thickness and seed number per silique had significantly (1%) positive correlation with SSRI.1.4There were differences in the valve structures between silique-shatter resistant accession H155and silique-shatter susceptible accession Qva.Compared with Qva, the cells in endopericarp of H155arranged more tightly and the cell wall lignified more seriously, and even the thin-walled cells in mesocarp were lignified and formed thick lignified layer.There were more vascular boundles in the valves of H155, which enhanced the mechanical strength.2Influencing factors analysis of silique-shatter resistance in Brassica napus2.1Results revealed that genetic differences were the crucial influencing factors of silique-shatter resistance.The SSRI of the same accessions among different replications did not reached to signifficant level. The SSRI of different accessions got to signifficant difference. The SSRI CV of silique from different branches ranged from18.67%to93.57%.Except for Ningyou No.10, siliques from the first branch had the biggest SSRI for other4accessions.The SSRI of siliques from the second branch and other above branches had not significant difference.2.2SSRI differed because of siliques from different portions in plants in different accessions, which CV of SSRI ranged from18.67%to93.57%. For most accessions, the siliques from the first branch had the largest SSRI. Siliques from the second branche and upper branches hand similr SSRI.Siliques from middle portion and lower portion had similar SSRI. SSRI of siliques from top portion were less than those from middle and lower portion in inflorescence by55.04%and60.43%, respectively. Siliques from the lowest portion had the largest SSRI.2.3There were difference of SSRI changes during dehydration bentwen silique-shatter resistant accessions and silique-shatter susceptible accessions.Compared with silique-shatter resistant accessions, SSRI of silique-shatter susceptible accessions decreased more sharpy. Results implied that water content in silique had linear regression relationship with SSRI (y=0.0149x-0.4779).The coefficient of determination (R2) was0.4284. Consequently SSRI enhanced while water content in silique increased.3Technique improvements in isolated microspore culture3.1Sampling time was put forward from traditional time (3days after anthesis) to9days before anthesis. Results showed the embryo yields were not significantly different from improved sampling time to that by traditional sampling time.The improved sampling time could increase total sampling time by12days and the total suitable sampling time was doubled.3.2In improved sampling method, suitable buds were harvested from donor plants in field and small buds were left to grow up and one donor plant could be used for several times. In addition, buds from the same donor plants had the same genetic background.3.313%sugar solution substituted B5medium as extraction solution. Results confirmed that embryo yield with improved extraction solution was not significantly different from that of B5medium. Sugar solution could save reagents and labour for preparation B5medium.3.4In the improved microspore incubation procedure, the microspores were directly incubated at32℃in dark till the embryos were visible. Compared with the traditional incubation procedure, the embryos were visible on the tenth day, which decreased the time for embryo visibility by four days.4Construction of linkage map and QTLs mapping of silique shatter resistance4.1The frequency distributions of SSRI in P1(H155), P2(Qva), F1,F2, BC1F1and BC2F1were continuous.The proportion of silique-shatter susceptible plants in populations of F1, F2, BC1F1and BC2F1were high and no plants had higher nor lower SSRI than those of both parents. The SSRI frequency distributions of ZZ DH lines in Zhengzhou and Wuhan were similiar and they distorted to silique-shatter susceptible parent Qva. The SSRI distribution of ZZ DH lines belong to muliti-normal distrbution and it was continuous. Therefore silique-shatter resistance trait was controlled by quantitative minor-effect additive genes and environmental factors affected it.4.2A linkage map with175polymorphic loci and19linkage groups was constructed, which covered1382.8cM and its mean marker interval was7.9cM. The genetic distance of linkage groups ranged from33.5cM to107.7cM and the number of markers in linkage groups varied from4to20. 4.3QTL analysis of silique-shatter resistance was conducted with composite interval mapping method of Windows QTLCart V2.5software and the mean SSRI in Zhengzhou and Wuhan. Nine QTLs were obtained in Zhengzhou and four QTLs were obtained in Wuhan(LOD>2.0). In Zhengzhou,9QTLs were located in N1, N7, N8, N15and N18linkage groups, respectively. These QTLs could explain49.0%phynotype. There were epistatic interactions among these QTLs and their contribution to phynotype was about45.9%. These QTLs and their epistatic interactions in Zhengzhou could totally explain94.9%variance. In Wuhan,4QTLs were located in N1, N4, N7and N18linkage groups, respectively.These QTLs could explain38.6%phynotype. There were also epistatic interactions among these QTLs and their contribution to phynotype was about12.8%. These QTLs and their epistatic interactions in Wuhan could totally explain51.4%variance.4.4There were three QTLs which were detected both in Wuhan and in Zhengzhou. qSSRI2detected in Zhengzhou and qSSRI10detected in Wuhan were all located in N1linkage group and their positions were adjacent. qSSRI4from Zhengzhou and qSSRI12from Wuhan were all located in N7linkage group. Their positions were also adjacent. qSSRI9from Zhengzhou and qSSRI13from Wuhan were located in N18linkage group. Their positions were also adjacent.4.5The results of QTL mapping in this study were consistent with the results of association analysis of Raman et al (2011).And the results of Raman et al (2011) aggreed with the findings of a comprehensive transcriptome analysis of silique development and dehiscence in Arabidopsis and Brassica(Jaradat et al.,2010).更多还原