Quantitative Trait Analysis of Heterosis-relevant Loci using Molecular Markers in Oilseed Rape (Brassica napus L.)

摘要

Molecular markers are increasingly used to detect quantitative trait loci (QTL) that are responsible to the trait of interest. The central theme in hybrid breeding is heterosis and it is important to put the first emphasize on the QTL relevant for heterosis. The genetic action of the QTL found furthermore is necessary to be studied, as this will become the basis for decision on which breeding method to be used. A doubled-haploid (DH) population from a cross between "Express 617" and "V8" was developed using microspore culture. A counterpart back-cross population (BC) was developed by crossing each line in DH population with the ‘Express 617’. The populations would make possible estimation of the heterosis effect and the subsequent QTL analysis. Phenotypic performance was studied in a greenhouse trial during Summer 2007 and in field for two years (2005/06 and 2006/07) in four locations, two in Lower Saxony and two in Hesse. Microsatellites and AFLP® were used to genotype the DH and develop genetic maps. The genetic map comprised all the chromosomes of Brassica napus (Chromosome N01-N19) and covered 1792 cM. The QTL analysis was done using QTL-Network 2.0 based on a backbone map with 269 markers. Three sets of data, DH, BC, and mid-parent heterosis (MPH) were analysed for each trait.Greenhouse trial showed that BC tended to perform better than DH in almost all traits observed. Heterosis distributions were varied among traits, which range spanning from -31% (leaf area) to 84% (shoot dry mass). The correlation between BC and DH among the traits observed were medium, but in hypocotyl length it was high (0.81) and in specific leaf weight (SLW) was very low.Seed yield for 2006/07 was clearly lower than the earlier planting year due to mild winter and dry period during flowering time which caused the yield to go down 30% in average. Mid-parent heterosis in 2006/07 was higher than the previous year. Strong correlations between BC and MPH indicated that MPH was controlled more by variation in BC. Thousand seed mass appeared to be unconnected to the other traits, as well as hypocotyl length and leaf area. Biomass (fresh shoot mass) in general might show a weak correlation with yield.From greenhouse trial there was no heterosis-relevant main-effect QTL detected. Eleven main-effect QTL nevertheless was found from DH and BC populations; however, they were not co-located, making it impossible to indirectly estimate dominance effect. There were eleven heterosis-relevant main-effect QTL detected in field trials: four from 2005/06 planting year and seven from 2006/07. As already indicated from phenotypic relationship, QTL for thousand seed mass were mostly different from the other traits. Clusters of main-effect QTL were called ‘active regions’ and each was seen as a single region, ignoring the traits. Twenty six active regions were detected on 15 of the 19 chromosomes. Applying “active region” as basis for interaction unit raises the percentage of interaction involving at least one active region from 5% to 65%. This suggested that, at least in case of epistasis, it is better to see a cluster of QTL as single unit. Certain interactions involving the same trait could be found clustering in the same location. Such feature may indicate existence of transcription factors. QTL effects related to heterosis from early traits (greenhouse trials) were evidently dominated by epistatic origin, since no dominant QTL was detected directly from MP data set. Nevertheless, four main-effect QTL were found from BC data set, estimating difference between genetic additive and dominance effect; two of them, responsible for dry shoot weight and fresh shoot weight, located at the same region in N14. Partial dominance played in all the QTL for heterosis which additive and dominance effects could be estimated. Five of the seven such QTL showed partial dominance toward ‘Express’, the other two toward "V8". Most of the estimates were from QTL for TSM. Based on these results, it is indicative that the additive-dominance model is not enough to explain the phenotypic expression.The results showed that epistasis was evidently the source of heterosis in the early developmental stage. Partial dominance was found to play role in the traits observed at end stages of life, although epistasis was still abundant. Phenotypic correlation between shoot dry weight at early stage and yield could not be strongly supported by the QTL analysis. Further genetic analysis involving expression of gene, such as e-QTL can be used to explain the mechanism behind the correlation, which further can be applied in developing more effective marker-assisted methods in hybrid cultivar breeding.