摘要
【目的】对超级杂交稻两优培九影响产量及其构成因素性状的杂种优势位点进行定位,在此基础上探讨亲本培矮64S和9311的遗传差异与水稻产量性状的杂种优势间的关系,以探明水稻产量杂种优势的分子预测途径。【方法】应用经单粒传法获得后续世代的219个培矮64S×9311 F8重组自交系(RILs)株系材料与亲本培矮64S回交,并选用151个分布于水稻基因组12条染色体上的SSR多态性标记,构建回交群体RILs BCF1;构建基因组总长为1 617.7 cM、标记间平均距离10.93 cM和含151个分子标记的遗传图谱;采用分子标记技术和自由度不等的单向分组方差两组法、三组法分析,用SAS软件ANOVA分析、混合线性模型复合区间作图等方法,对回交RILs BCF1群体的产量性状及其构成因素的F1表型值进行相关分析、优势预测与QTL定位。【结果】本回交杂种群体RILs BCF1具备多种基因型,遗传变异丰富,性状平均值均显著高于亲本群体重组自交系RILs F8,共筛选到影响RILs BCF1群体产量及其构成因素性状杂种优势的阳性、增效位点74个;其中,三组法所筛选的阳性、增效位点数高于两组法,用这些阳性、增效位点所预测的遗传距离与产量F1性状值的相关性也显著提高;三组法所筛选产量性状的增效位点与两组法所筛选的增效位点完全一致;连锁紧密的位点有成簇分布的现象,每穗空粒数、每穗实粒数、结实率有6个杂种优势位点相同,并与3个产量杂种优势位点重叠,且均处在第7染色体上;通过逐步回归建立了对4个产量性状进行预测的回归方程模型;筛选到28个杂合型的特异性标记,它们与产量性状的表型值显著相关,使用特异性标记可使遗传距离与产量F1性状值的相关系数由全部标记的0.335提高到0.617;定位到3个与产量杂种优势相关的QTL和3个影响每穗实粒数杂种优势的QTL。其中,在第7染色体上影响每穗实粒数和产量杂种优势的QTL QGpp7和QHy7与影响每穗实粒数和产量杂种优势的增效位点的结果相符。【结论】通过增加筛选产量杂种优势阳性位点或增效位点数量、筛选影响杂种优势特异性分子标记的方法,可显著提高分子标记遗传距离与产量F1性状值的相关性,有效提高用分子标记遗传距离对杂种优势预测效率。定位了3个影响产量杂种优势的QTL及3个影响每穗总粒数杂种优势的QTL,分别在第2、3、7、11和12染色体上,其中,影响产量杂种优势的数量性状位点QHy7,贡献率为7.48%,可用于杂种优势的预测和杂交组合的选配。定位于第3染色体RM293—RM468的表型贡献率为14.9%的抽穗期QTL可用于早熟高产水稻的选育。
【Objective】 The heterosis loci and QTLs of yield and yield components were detected by using a RILsBCF1 population derived from a cross between Pei＇ai 64 S and 9311. The relationship was explored between the genetic variance of these two parental lines and yield heterosis in the resulted hybrid for predicting hybrid heterosis.【Method】Based on a population of 219 recombinant inbred lines（RILs） of F8 generation produced by single seed descendant method from the Pei＇ai 64S×9311 cross, a RILsBCF1 population was generated by backcrossing of RILs to Pei＇ai 64 S. With a total of 151 polymorphic SSR markers, a linkage map was constructed spanning 1 617.7 cM across the whole genome with an average marker interval of 10.93 cM. The correlation between genetic distances and F1 trait performance of RILsBCF1 and their prediction in yield and yield component traits were conducted respectively by using molecular marker analysis, one-way ANOVA with different freedoms, and composite interval mapping using mix linear model in SAS, together with heterosis prediction and QTL mapping. 【Result】The RILsBCF1 used in the study showed significant diversity with high segregation in multiple traits, and their average performance was significantly higher than that of RILs F8. In this RILsBCF1 population, 74 heterosis positive loci and effect-increasing loci were identified by two-group method and three-group method in yield and yield component traits, respectively. Compared with two-group method, three-group method could get more positive loci or effect-increasing loci to a certain degree and raise efficiency of predicting correlationship between genetic distances of both positive loci and effect-increasing loci and F1 traits＇ performances. The result of the effect-increasing loci detecting was the same in both two-group and three-group methods. Six heterosis loci were detected at the same regions for three traits（Sterile lemma per panicle, Grains per panicle and Pencentage seed setting）, overlapped with three yield effect-increasing loci clustered on chromosome 7. Based on the relationship between marker-effect values of yield effect-increasing loci using the three-group method and F1 trait performance, four multiple regression prediction models were constructed using a stepwise procedure. A total of 28 markers with heterozygous genotypes were identified to significantly increase the correlation coefficient between the genetic distances and the F1 trait performance from 0.335 to 0.617. Three QTLs for yield heterosis and three QTLs for grain per panicle heterosis were mapped using this RILsBCF1 population. The mapped loci of QTL QGpp7 for grain per panicle heterosis and QHy7 for yield heterosis matched the effect-increasing loci identified by the methods of two-group and three group analyses. 【Conclusion】The approaches of screening more positive loci or effect-increasing loci and specific markers which influent heterosis can increase the correlation coefficient between the genetic distances and the F1 traits performances, and thus can be applied more efficiently in predicting the yield heterosis of rice hybrids with genetic distance of molecular markers. The yield QTL QHy7 located on chromosome 7 with a yield increase contribution of 7.48% can be used for yield heterosis prediction and in hybrid rice breeding. A heading stage QTL located between RM293-RM468 on chromosome 3 with the contribution of 14.9% can be used for rareripe high yield rice breeding.
出处
《中国农业科学》
CAS
CSCD
北大核心
2014年第14期2699-2714,共16页
Scientia Agricultura Sinica
基金
国家重点基础研究发展计划("973"计划)(2006CB101700)
教育部博士学科点专项基金项目(20060533064)
关键词
超级杂交稻
产量性状杂种优势
杂种优势预测
QTL定位
super-yielding hybrid rice
yield component trait heterosis
heterosis prediction
QTL mapping
