The Study On Molecular Genetic Base Of Brassica Napus Intersubgenomic Heterosis

Introgression of Brassica rapa genome (2n=2X=20, ArAr) into Brassica napus (2n=4X=38, AnAnCnCn) creates a type of new type B. napus (4X=38, Ar/nAr/nCnCn). A cross between the new type B. napus and natural B. napus produces stronger heterosis, which is nominated as intersubgenomic heterosis. There were previous ample experiments to prove its truth but molecular genetic mechanism of intersubgenomic heterosis was still not unveiled. To investigate the influence of B. rapa genomic introgression on B. napus genomic structural variations and molecular mechanism of intersubgenomic heterosis, we developed a recombination inbred line (RIL) population (called TH-RIL) of new type B. napus of138Fg lines derived from an fertile BC1F2individual with38chromosomes which came from a corss of H3(a semi-winter B. napus cultivar)/TM (a semi-winter B. rapa cultivar)//TM and its corresponding backcrossed F1population (TH-BC) created by crosses of each RIL and H3. Field trial was carried out at one location in China in two years. One hundred and thirty eight recombination inbred lines,138BC combinations and their two parents (also as controls) were planted to evaluate for yield and eight yield-related traits, including seed yield, seed number per pod, kio-seed weight, dry weight, flowering time, plant height, branch number per plant, and oil content. In addition, pod number per plant and oil yield were calculated.The phenotype data showed that compared with the better parent H3,77.5%of the RIL lines showed very obvious advantages on seed weight. Averagely almost half of the lines had advantages over H3in the respect of PH, SN and DW. The remaining traits, including OY, SY, PN, FT, were30%lower than H3in RIL lines. For SY trait,20.3%RIL lines exceeded H3. Such data demonstrated that the introgression of B. rapa into B. napus made phenotype have great variations. Middle parent heterosis (MPH) level ranged from SW (0.0%) to OY (47.1%) while the OPH ranged from BN (-7.4%) to OY (26.9%) while over-parent heterosis ranged from-7.4%(branch number) to26.9%(oil yield). About97%hybrid combinations exceeded H3for seed yield with an average of28%. This showed that when alien B. rapa genomic fragments were introgressed into B. napus, obvious heterosis appeared.Analysis of genomic structural variation and genetic mapping were conducted with872different kinds of markers, including545SSR markers amplified by289SSR prier pairs,51centromere-specific retrotransposon (RTc) markers, and248noncentromere-specific retrotransposon (RTnc).TH-RIL genetic map was constructed by628markers and has25linkage groups (LGs) covering2272.8cM with an average interval of3.9cM, including10LGs in A genome,7LGs in C genome except C6and C7, and8un-corresponding LGs with other published maps. The donor of C genome in TH-RILs only originated from C genome of H3, so partial C-genomic linkage groups could be successfully constructed because homoeologous recombination occurred between homoeologous chromosomes.Through the analysis of genome-wide structural variation in the TH-RIL population, we found that new type B. napus contained B. rapa-specific alleles and numerous new alleles, such as the occurrence of N-type alleles (being present in some TH-RILs but absent in two parents) and D-type alleles (being present in two parents but absent in some TH-RILs). The analysis showed that retrotransposon markers were reactivated during the development of materials. Ninety two N-type varied bands were sequenced and analyzed by three transposon-specific databases. Sixty five sequences contained the transposon fragments, which showed that these genomic variations occurred by the induction of transposons. Finally, we detected12RBIP sites to confirm the occurrence of retrotransposon reactivation again. Go annotation toward sequenced65variants found that45variants were involved with functional genes, mainly including metabolism, cellular process and stress. The analysis of correlation coefficients showed that the novel sequence variation had significant impacts on the investigated traits.Three data sets, including RIL population, BC population and middle parent heterosis data (MPH) were used to analyze QTL by QTL Cartographer V2.5software. A total of123QTL for yield and nine yield-related traits were identified with40repeatable micro-QTL (MR-QTL) and83significant-level (SL-QTL) over the threshold. Fifty eight QTLs were detected in TH-RIL population. Thirty-three of the41QTL were located in chromosomal varied regions that could explain the64.2%phenotypic variation, and two fifths of the QTL contributed positively to the improvement in the agronomic traits. B. rapa genomic fragment (Ar) also played a great role in seed yield and yield-related traits, and6QTL and1/5loci detected by single-marker analysis originated from B. rapa parent.The heterosis analysis was performed using the TH-BC and TH-RIL data sets. By the analysis of dominant degree,63(54.6%) of66QTL belonged to heterosis-related QTL, including7partial dominance loci,6full dominance loci and50over dominance loci. In total,92.0%QTL (58) were distributed in A genome. Based on allele origin, five types of allele Ar (directly derived from B. rapa), An (directly derived from B. napus), An and Cn (indirectly produced by B. rapa introgression in A genome and C genome, respectively, and L (unmapped alleles). As the first factor,47.1%QTL (32) originated from novel varied alleles (An) in A subgenome with56.3%(18) loci of32QTL conducting favorable effect, then next to15QTL (22.1%) from Ar with86.7%(13) of15derived Ar loci arising positive effect. Furthermore,14.1%QTL were mapped in the rearrangement region and80%of these loci play a positive role while23.9%loci of QTL were involved with retrotransposon reactivity and52.9%among them performed a positive function. This indicated that the retrotransposon reactivity could induce new allelic emergence which indirectly played a great role in phenotype traits.Additionally,536interacting pairs of epistatic effect were also found by QTL mapper2.0software. For epistasis,372interaction pairs detected in TH-BC and TH-MPH data sets were identified and A-A subgenomic interaction ranked the first (71.4%), then followed by A-C interaction (26.5%) and the remained C-C interaction (2.1%). Since there was five allelic origins (Ar, An, An, Cn and L),15interacted patterns could be detected among interaction of five origins. The most important interacted modes were Ar-An (20.8%) and An-An (19.3%) which averagely58.6%loci produced a positive effect. It suggested that intersubgenomic heterosis were mainly rooted in An induced B. rapa genome introgression into B. napus and direct-effect Ar genome by the dissecting of whether single QTL or two-QTL interaction which slightly more than half of alleles worked as a positive function. The genetic mode of intersubgenomic heterosis was a comprehensive function of epistasis, dominant effect, and over dominant effect.B. napus was an allopolyploid species and the higher homology between A and C genome provided the beneficial foundation for fixed heterosis study. We detected203interacted pairs in RIL population to analyze heterosis fixation. Six interacted pairs between homoeologous genomic fragments were considered as fixed heterosis loci which may be resulted from chromosomal rearrangements. The investigation of fixed heterosis could directly improve the performance of new type B. napus and indirectly enhance intersubgenomic heterosis.An emergence of lots of novel alleles and chromosomal rearrangements induced by interspecific hybridization of B. napus and B. rapa makes new type B. napus significantly distinct from natural B. napus at alleles and genomic structure. Each pair of newly-detected alleles in an intersugenomic hybrid displayed dominance or over dominance alone. A wealth of alleles, allelic combination and allelic interaction which less occurred in an inter-variety hybrid, constitute the origin of intersubgenomic heterosis. It may be the essence of intersubgenomic heterosis merging epistasis and dominance effect from partial dominance to over dominance. B. napus, as a model crop of the study of polyploid, intersubgenomic heterosis and fixed heterosis, could play a great referenced role for guiding the research and breeding of other crops.