| Oilseed rape (Brassica napus L.) is one of the most improtant oil crops worldwide, the oil from its seeds could support both edible and industrial requirement. As the reservation of non-renewable crude oil resources gets more and more limeted, it is strategically crutial for national safty and development of improving the yield and quality of renewable oil resources. During the past decades, the seed yield per unit of oilseed rape has been doubled, the fatty acid composition could be optimized for requirement of different usage, but the improvement of total oil content in seeds was slow and minor. Since the oil content in seeds is typically a quantitative trait, it is feasible to exploit the related genetic loci, unreval the variation mechanisms, and facilitate breeding practice via the quantitative trait loci (QTL) mapping. There has been a lot of studies referring QTL mapping related to oil content in seeds of B. napus, but some problems from those studies hamperred their application in breeding practice, for example, the number of identified QTL from single independent experiment is quite limited and always below 20, most of the single QTL contributes less than 10% of the phenotypic variation, and many QTL could be only detected from particular population or environmental condition. So, to effectively improve the complex trait seed oil content, it requires more manipulable genetic loci and their corresponding superior alleles, and furthermore, it will make the results of breeding practice better accordant with expectation when understand the proper conditions that particular alleles need to well function.In this study, seed oil content variation in a double haploid (DH) population, the TN DH population, and its derivate reconstruct-F2 (RC-F2) population were measured from totally 15 independent experiments including different years and different locations. The two parental lines of TN genetic population were both with moderate content of oil in seeds, that Tapidor had oil content 43.0±1.7% in seeds while Ningyou7 had 41.7±2.1%, on average. The oil content variation showed continuous distribution and transgressive segregation among the TN genetic population, the oil content varied from 33% to 51% extremely, and the average extent of oil content variation was 11%.Moreover, a kind of main inflorescence section harvest experiment has been carried out based on the TN DH popoulation, the main inflorescence of each line was labelled as 5 sections from bottom to tip along flowering, and siliques from each section were seperately harvested. It is notable that the direction of oil content variation along adjacent sections was different among lines, either increasing or decreasing. When under a certain polarity of plant, the different types of oil content variation along adjacent sections may probably due to divergent responses of different genotypes to the specific environmental conditions, under which the siliques of each section developed. Grouping lines according to the opposite types of oil content variation between each adjacent sections, theχ2 test for independence was applied to evaluate the correlation of genotype and phenotypic variation type, and 103 marker loci were found related to the divergence of oil content variation types along adjacent sections.Taking the seed oil content variation among TN DH population on each section as an independent experiment also, we did QTL mapping and epistatic interaction loci detection for each of the 20 independent experiments, respectively. A total of 195 original QTL, most intervals of which were overelapped with others, and 92 epistatic interaction involved loci, were identified from all the experiments. Among the original QTL,105 were defined as significant-large QTL (SL-QTL) which at the significant level p≤0.05, while the other 90 were defined as weak QTL which at the significant level p<0.5. Most of the epistatic loci were specifically identified from particular experiment; a large part of the epistatic loci were included in the original QTL intervals, while the other 36 epistatic loci were independent of QTL intervals. Half of the 103 marker loci, which were found related to the divergence of oil content variation types along adjacent sections, were also involved in the original QTL intervals.Applying the "meta-analysis"method, we integrated the overlapped original QTL to get refined intervals. Then the 195 original QTL were concluded as 59 independent "meta-QTL" and 28 weak-singleton QTLs (WS-QTL). The meta-QTL were sorted according to the significant level of the original QTL they included,50 meta-QTL had at least one significant original QTL included and were defined as meta-SL-QTL, and 9 meta-QTL had multiple weak original QTL included and were defined as meta-MR(micro-real)-QTL. As each meta-QTL included one or several original QTL which were corresponding to particular experiment(s), and the environmental conditions, temperature and sunlight illumination time, were recorded during the pivotal phase of seeds filling in each experiment, then it is feasible to analyse the environmental specificity that the meta-QTL required to well function. We found 28 meta-QTLs were specifically detected from particular conditions of temprature, or sunlight illumination, or both aspects, and notably several independent QTL which were detected from only single experiment indeed with the particular experiment had extreme condition(s). So, we suggest that the environment specific QTL should not be neglected.Through the common molecular markers on different genetic maps, we projected the reported QTL, which were related to seed oil content variation and identified from other genetic populations of B. napus, onto the TN genetic map that was used as reference. The multiple experiments showed significant improving on the detectable number of QTL:the QTL from only TN genetic population covered 80% of the QTL which were identified previously from other three populations. On the other hand, the QTL identified from other populations validated 30% of the QTL in TN population, including 25% of the weak QTL and 20% of the epistatic loci which were independent of QTL intervals.In addition to the comparison among different genetic populations, we furthermore compared the QTL that were identified from genetic map based mapping with associated markers that were identified from naturally varied breeding lines. The corresponding loci on TN genetic map of the associated markers were identified first, and then we found 25% of the QTL in TN population were validated by the associated loci, including half of the weak QTL and also 20% of the epistatic loci which were independent of QTL intervals. The inconsistent effects of particular QTL between different genetic backgrounds or environmental conditions were apparent through the comparisons.The unprecedented large amount of QTL related to seed oil content variation in B. napus that was revealled in this study, will provide rich targets for genetic improvement of such trait; While the uncovered relativity and specificity of QTL effects, are important references for choosing and combining of the most effective alleles in particular genetic design under a certain breeding condition. |
PDF Full Text Request