【作者】 崔法；

【导师】 王洪刚；

【作者基本信息】 山东农业大学 ， 作物遗传育种， 2011， 博士

【摘要】 小麦(Triticum aestivum L.)是世界上重要的粮食作物之一。普通小麦是异源六倍体(2n=6x=42),含有A、B和D 3个染色体组,基因组巨大,而且其中含有80%以上的DNA重复序列,导致遗传标记在不同材料间的多态性较差,小麦的遗传研究落后于玉米、水稻等作物。构建高密度小麦分子标记遗传连锁图,对控制产量相关性状QTL进行精细定位,明确其在染色体上的位置和效应,对于小麦产量相关性状的遗传改良具有重要意义。本研究利用3个关联性作图群体进行高密度遗传连锁图谱的构建,并进行产量相关性状的多元条件和非条件QTL分析,进而在QTL水平揭示产量及产量相关性状的遗传特点。本研究获得以下主要结果:⑴利用485个潍济(WJ)RIL家系和229个潍烟(WY)RIL家系进行基于PCR原理的分子标记遗传连锁图谱的构建,最终获得分别包含344和358个位点的分子标记遗传连锁图谱,它们分别覆盖基因组2855.51 cM和3010.72 cM,标记间的平均距离分别为8.30 cM与8.41 cM。由于相邻标记间距大于50.0 cM,潍济与潍烟连锁图谱分别各形成6个连锁断点。两图谱共包含69个公共位点。⑵从潍济和潍烟两个群体中分别随机选择175和172个家系进行基于多样性微阵列技术(Diversity Arrays Technology)的DArT标记的开发,结合DArT标记数据及PCR分子标记数据,进行遗传连锁图谱的加密,最终获得分别包含629和681个位点的潍济与潍烟加密连锁图谱,两图谱分别包含373和406个DArT标记位点,图谱全长分别为2777.54 cM和3004.62 cM,相邻标记间平均距离分别为4.42 cM和4.41 cM。由于相邻标记间距大于50.0 cM,潍济与潍烟连锁加密图谱分别形成8个和13个连锁断点。两图谱共包含270个公共位点。⑶将潍济与潍烟构建的两个加密图谱与另一基于179个家系潍麦8号与洛旱2号的F8:9 RIL家系构建的加密图谱进行整合,最终获得包含1113个位点的高密度整合图谱,其中,536个是基于PCR原理的分子标记位点,575个为DArT标记位点。整合图谱全长2946.98 cM,标记间平均距离为2.65 cM。在1113个位点中,群体间共同位点共494个,覆盖小麦基因组1607.21 cM,被用于基于巢式关联作图群体的QTL联合分析。⑷基于全部485/229个潍济/潍烟RIL家系的QTL分析共检测到533个控制21个产量及相关性状的加性QTL;其中,117个QTL在E1、E2、E3、E4中被至少重复检测到2次, 36个QTL贡献率超过10%,为环境间稳定表达的主效QTL,其中的19个QTL位于1B、1D、2A、2B、2D、5A、5D或6D染色体的同一区段,表现为一因多效,共涉及7个控制产量相关性状稳定表达的QTL富集区。基于175/172个随机选择潍济/潍烟RIL家系的QTL分析共检测到480个控制上述21个产量相关性状的加性QTL,其中75个QTL在E1、E2、E3、E4中被至少重复检测到2次,61个QTL贡献率超过10%,为环境间稳定表达的主效QTL,其中的32个QTL位于1A、1D、2A、2B、2D、3A、4A、5A、5B、6A、6B、7B或7D染色体的同一区段,表现为一因多效,共涉及13个控制产量相关性状稳定表达的QTL富集区。基于潍济与潍烟群体大群体检测到的QTL中,分别有22.22%和32.62%的QTL在基于小群体QTL分析中也被检测到。基于巢式关联作图群体的QTL联合分析共检测到185个控制上述21个产量相关性状的群体间等效QTL。⑸基于全部485/229个潍济/潍烟RIL家系的原因/结果条件QTL分析,首次揭示了籽粒大小与千粒重之间,千粒重、穗部性状与穗粒重之间以及穗长、各节间长与株高之间在单个QTL水平之间的关系。籽粒大小与千粒重之间条件QTL分析结果表明,籽粒宽度在单个QTL水平对千粒重贡献最高,其次是籽粒长度,粒径比在单个QTL水平对千粒重贡献最低;穗长、小穗数、穗粒数和千粒重与穗粒重的条件QTL分析结果表明,穗粒数在单个QTL水平对穗粒重贡献最高,其次是千粒重,小穗数与穗长在单个QTL水平对穗粒重贡献均较低;穗长及各节间长与株高的条件QTL分析结果表明,倒三节间长在单个QTL水平对株高贡献最高,其次是倒二与倒四节间长,而穗长在单个QTL水平对株高贡献最低,其次是穗下节间长。⑹基于大小群体QTL分析的检测结果表明,图谱密度与群体大小均能影响QTL检测的数目,且小群体易于夸大对QTL效应值的估计。更多还原

【Abstract】 Wheat (Triticum aestivum L.) is a major food crop worldwide. As an allohexaploid carrying the genomes AABBDD (2n = 6x = 42), common wheat owns large genome size, 80% of which is repetitive DNA . Hence, molecular markers show lower level polymorphism in wheat, which makes the genetic research in wheat fall behind other crops such as maize, rice, etc. Knowing the precise position and effect of QTL for yield-related traits will be of great value for genetic improvement in yiled in wheat breeding programs. Thus, it is of importance to construct a high-density wheat molecular map. In present study, high-density wheat molecular genetic maps were constructed based on three related mapping populations. Based on the novel molecular genetic maps above, multivariate conditional and unconditional QTL mapping analysis were conducted to specify the genetic characteristics of yield and yield-related traits at the QTL level. The main results were as follows:⑴Two molecular genetic maps comprising 344 and 358 PCR-based marker loci were constructed based on 485 WJ-derived and 229 WY-derived recombinant inbred line (RIL) populations. The two genetic maps spaned 2855.5 cM and 3010.72 cM, respectively, with an average density of one marker per 8.30 cM and 8.41 cM. Due to the linkage distance > 50 cM between the adjacent loci, there were six linkage gaps each in the two genetic maps. Sixty-nine pairwise molecular marker loci were common to the two genetic maps.⑵To saturate the two genetic maps above, 175 and 172 progenies were randomly selected from 485 WJ-derived and 229 WY-derived RIL populations, respectively, for developing diversity arrays technology (DArT) markers. Combining the novel DArT markers and PCR-based markers, two high-density genetic maps were established, including 629 and 681 loci on the wheat chromosomes, 373 and 406 of which, respectively, were novel DArT marker loci. The two high-density genetic maps covered total lengths of 2777.54 cM and 3004.62 cM, with average densities of one marker per of 4.42 cM and 4.41 cM, respectively. The linkage distances > 50 cM between the adjacent loci resluted in eight and 13 linakage gaps in WJ and WY-derived high-density genetic maps, respectively. The two high-density genetic maps shared 270 common loci.⑶A high-density integrative genetic map was developed by combining the two WJ and WY-derived high-density genetic maps and the third high-density genetic map, which was constructed based on a 179 F8:9 RIL population derived from the cross between Weimai 8 and Luohan 2. It comprised 1113 loci, 536 and 575 of which, respectively, were PCR-based marker loci and novel DArT marker loci, covering a total length of 2946.98 cM, with an average density of one marker per of 2.65 cM. Of the 1113 loci, 494 loci were common among the three individual genetic maps, covering a total length of 1607.21 cM. The common loci were used for joint QTL mapping analysis based on the nested association mapping (NAM) population.⑷In total, up to 533 putative additive QTL for the 21 yield and yield-related traits were detected in the 485/229 WJ/WY RIL populations. Of these, 117 QTL showed significance in at least two different environments of E1, E2, E3 and E4, 36 of which accounted for no less than 10% of the phenotypic variation, being major stable QTL. Of these, 19 QTL showed pleiotropic effects and were co-located on chromosomal regions of 1B, 1D, 2A, 2B, 2D, 5A, 5D or 6D, referring to seven major stable QTL clusters. QTL analysis based on 175/172 WJ/WY RIL populations identified 480 QTL for the 21 yield and yield-related traits. Of these, 75 QTL were verified in at least two different environments of E1, E2, E3 and E4, 61 of which were major stable QTL explaining more than 10% of the phenotypic variation. Of these, 32 QTL showed pleiotropic effects and were co-located on chromosomal regions of 1A, 1D, 2A, 2B, 2D, 3A, 4A, 5A, 5B, 6A, 6B, 7B or 7D, referring to 13 major stable QTL clusters. In the WJ and WY population, 22.22% and 32.62% QTL detected based on large RIL populations have been verified in the QTL analysis based on 175/172 WJ/WY RIL populations, respectively. Joint QTL mapping analysis based on the nested association mapping (NAM) population detected 185 congruent QTL among the three RIL populations.⑸Possible genetic relationships between kernel dimensions (KD) and thousand-kernel weight (TKW), TKW, spike-related traits (SRT) and kernel weight per spike (KWPS), and plant height components (PHC) and plant height (PH) were detected using both multivariate conditional and unconditional QTL analysis based on 485/229 WJ/WY RIL populations. The results were as follows: at the QTL level,①kernel width (KW) has the strongest influence on TKW, next to kernel length (KL), but kernel diameter ratio (KDR) has the least level contribution to TKW;②kernel number per spike (KNPS) contributes the most to KWPS, next to TKW, both spikelet numbe per spike (SPN) and SL have lower level contribution to KWPS;③spike length (SL) contributes the least to PH, followed by the first internode length from the top (FIITL); the third internode length from the top (TITL) has the strongest influence on PH, followed by the second internode length from the top (SITL) and the fourth internode length from the top (FOITL).⑹Based on comparison of the QTL detected based on large/small populations using moderate/high-density genetic maps, we concluded that:①both density of the genetic map and population size have great influence on the estimation of QTL number; and②the limited population sizes can lead to overestimation of QTL effects更多还原