Identification Of The Major QTL For Seeds Per Silique And Seed Weight With Construction Of Near-isogenic Lines In Oilseed Rape

Rapeseed is not only one of the main edible vegetable oils, but also one of the potential candidates for bio-diesel in the world.1,000 seeds weight and seeds per silique are two important yield components for oilseed rape. However, both traits showed typical quantitative phenotypic variation and controlled by multiple genes, therefore, the genetic mechanisms are complicated and difficult to be understood. So far, although a number of QTLs for 1,000 seeds weight and seeds per silique have been published, fine mapping and map based gene cloning underlining related QTL are rarely reported. In this research, we focus on validating and fine mapping the previously maped two major QTLs for 1,000 seeds weight on linkage groups C8 (qSWC8)and seeds per silique on A6 (qSSA6) in SG-DH population. The work includes to develop near-isogenic lines by advanced backcross approach, to increase the marker density by integrating locus specific markers and to narrow down the QTL intervals through association analysis between the marker genotypes and their corresponding phenotypes of BC2F2、BC3F2、BC3F1and BC4F1 generations. Main results are as following:1. Based on the published SG map, using B.rapa and B.oleracea genomic sequence resources, we developed 81 and 97 locus specific markers for qSWC8 and qSSA6 respectively and as the result,26 and 20 could be integrated to the qSWC8 and qSSA6 regions, thus the average distance between markers in QTL regions were decreased from 1.51 cM to 0.36 cM for qSWC8 and from 13.50 cM to 1.47 cM for qSSA6.2. Using the new integrated C8 and A6 maps, mapping data of 1,000 seeds weight in ten environments and seeds per silique over nine locations were re-analysed in SG-DH population. The results showed that the QTL region for qSWC8 was clearly reduced from 25 cM to 11.9 cM by flanking markers C8SW76 and C8SW51, and narrow down the QTL interval for qSSA6 from 40 cM to 21.9 cM between A 6SS40 and A 6SS28.3. Based on the previously developed BC2F1 and BC3F1 progenes concerning qSWC8 and qSSA6, the populations of BC2F2 and BC3F2 were established by selfing the BC2F1 and BC3F1 plants and used for the next field observation. Meanwhile, by marker assistant selection, we develop the higher backcross generations of BC3F1 and BC4F1 for further fine mapping study.4. Using the BC2F2 and BC3F2 populations of 591plants, the validation of qSWC8 was performed by association analysis between marker genotypes of 10 markers covering the whole QTL region (11.9cM) and their correspondent phenotypes. The qSWC8 was mapped in a 0.8 cM region between the markers C8SW76 to C8SW24. The comparative analysis of 1,000 seeds weight between homozygous sister sub-NILs carrying "Gaoyou" fragment (n=68,3.244g) and NILs containing "Sollux" segment (n=64,2.910g) in the 0.8 cM target region showed significant difference of 0.224g (p<0.05) in 1,000 seeds weight, which corresponding to the physical distance of 2.03Mb in B. oleracea C8.5. In total of 742 plants from seven BC2F2 and one BC3F2 populations were used to validate qSSA6. Association analysis between marker genotypes of 10 markers covering the whole qSSA6 region (21.9cM) and their correspondent phenotypes. The qSWA6 was mapped in a 19.1 cM region between the markers A6SS40 and A6SS26. The further mapping was carried out applying 431 BC4F2 plants and then narrow down the qSWA6 region to 14.5cM. The comparative analysis of seeds per silique between homozygous sister sub-NILs carrying "Gaoyou" fragment (n=37,25.65seeds) and NILs containing "Sollux" segment (n=50,23.03seeds) in the 14.5cM target region showed significant difference of 2.61seeds(p<0.01).