Cloning And Functional Analysis Of A Major QTL,qSS.C9,For The Number Of Seeds Per Silique In Brassica Napus L.

Rapeseed is one of the most important sources of both vegetable oil and oil extraction meal in the world. Yield enhancement thus has been one of the most important goals of rapeseed production and genetic improvement. As a complex quantitative trait, the seed yield of a rapeseed plant is comprehensively determined by three components: the number of seeds per silique(NSS), number of siliques per plant and thousand seed weight. Recently, several quantitative trait loci(QTLs) for yield-related traits have been identified from different populations. However, only few studies reported QTL mapping for NSS in B. napus. Up to now, 35 QTLs related to seeds per silique have been reported, but none of these QTLs has been cloned thus far. To dissect the genetic basis of NSS, F1 plants of the cross of HZ396 × Y106 were used to develop a double haploid(DH) population. The DH population, consisted of 140 lines, was used for map construction and QTL analysis. In field experiments across three seasons and two locations in China 140 doubled haploid lines and their corresponding parents were evaluated for silique-traits. QTL meta-analysis revealed that 3 consensus QTL for NSS. For the unique QTL in the linkage group C9, the additive effects of qSS.C9, which explained 57.77% of the phenotypic variance of NSS and favorable alleles originated from Y106(Zhang et al 2012). In this study, near-isogenic line(NIL) for qSS.C9 was used to consecutively back cross with the recurrent parent HZ396. Then, we constructed two mapping populations to isolate the qSS.C9 locus. With the integration of classic map-based cloning strategy and comparative genomics, we successfully cloned the qSS.C9 gene(BnaC9.SMG7b). Furthermore, the felmale sterile molecular mechanism of HZ396 and the function of BnaC9.SMG7 b in megasporogenesis development were studied by cytological and molecular biological methods.There rsults were described as follows:1. QTL analysis of the q SS.C9 locus in the BC3F2 population. To investigate the effect of qSS.C9, we constructed one near-isogenicline(BC3F2) in the HZ396 background by consecutive backcrossing. QTL analysis based on the BC3F2 population showed that the qSS.C9 locus explained 61.9 % of phenotypic variance with additive and dominant effects of 7.4 and 6.4 NSS.2. The fine mapping of the qSS.C9 gene. To fine mapping the qSS.C9 locus, two populations were developed by backcrossing Y106 and W11(high-NSS parent) as the donors to HZ396(low-NSS parent) consisting of 16,466 and 1,088 individuals,respectively. To precisely map the qSS.C9, we assayed these BC4F2 populations with three SCAR markers, SCC9-136, SRC9-298 and SRC9-397, narrowing down to a 140-kb region in the reference Brassica rapa A10 genome, a 280-kb region in the reference Brassica oleracea C9 genome and a 261-kb region in the reference Brassica napus C9 genome.3. The cloning of the qSS.C9 gene. According to the Brassica database(http://brassicadb.org/brad/index.php), there are 45 predicted genes in the 280-kb candidate region of B. oleracea. We scanned the predicted proteins and concluded that 16 of them are unlikely to be associated with the trait of NSS according to the available function annotations. For each of the left 29 genes, we comparatively sequenced the genomic fragment covering the promoter region and the complete coding region from both homozygous NILs. The results showed that two predicted genes in NIL(Y106) individually homologous to Bol043637 and Bol043640 were deleted in NIL(HZ396), while no reproducible DNA polymorphism was observed for the other genes. The genetic complementation suggested BnaC9.SMG7 b, the ortholog of AtSMG7(At5g19400), corresponds to the qSS.C9 gene.4. Evolution analysisof BnaC9.SMG7 b. The phylogenetic, protein sequence and syntenic analysis indicated that BnaC9.SMG7 b has the closest relationship with BolC9.SMG7 b. According to the sequence variations of BnaC9.SMG7 b, four haplotypes were identified from these B. napus accessions. The first haplotype(HAP1) characterized with the deletion of BnaC9.SMG7 b. The HAP3 and HAP4 which displayed five SNP variations relative to HAP2, but none of the nucleotide changes had an effect on BnaC9.SMG7 b coding sequence. Further statistical analysis indicated that HAP2, HAP3 or HAP4 showed a significantly higher value of NSS than HAP1, but no significant difference was observed among HAP2, HAP3 andHAP4. Therefore, the presence/absence of Bna C9.SMG7 b might be an important factor that affects the natural variation of NSS in B. napus, while the SNP in the promoter region of BnaC9.SMG7 b has no effect on NSS variation. In conclusion, it can be concluded that BnaC9.SMG7 b is a functional allele for improving NSS in the breeding of B. napus, and has been almost fixed in modern accessions.5. The effect of qSS.C9 on the grain yield of rapeseed. We compared the performances of six yield-related traits between NILs, and the results showed that NIL(Y106) showed extremely significant increases of NSS than NIL(HZ396) but substantial decreases of thousand-seed weight and number of siliques per plant; collectively, NIL(Y106) out-yielded NIL(HZ396) by 6.77 g seeds(58.7%) per plant. Thus, our results indicate that qSS.C9 is an elite allele for rapeseed yield potential through the increase of NSS.6. Expression patterns of BnaC9.SMG7 b and BnaC9.SMG7 c. RT-PCR, qPCR and ProBnaC9.SMG7b-GUS analysis indicated BnaC9.SMG7 b was expressed in cotyledons, rosette leaves, roots, young pedicels and pistils before and after pollination but not in stamens, petals, stems and mature siliques. Meanwhile, we also find that BnaC9.SMG7 c was expressed in various organs by RT-PCR and qPCR analysis. The subcelluar localization assay showed that the BnaC9.SMG7b-GFP and BnaC9.SMG7c-GFP fusion protein were both co-localized with the marker in the P bodies.7. The function study of BnaC9.SMG7 b. We stained callose in the cell plates of megasporocyte undergoing meiosis in ovules with aniline blue and observed the developmental process of female gametophyte. Comparable frequencies of aniline blue-stained callose bands were observed between NIL(HZ396) and NIL(Y106) throughout the whole meiosis stages. Differently, although the functional megaspores in most of the ovules(96.5%) underwent three times of mitosis and cellularization to generate mature female gametophytes in NIL(Y106), a high and consistent proportion(65.6%) of morphologically abnormal female gametophytes(or no female gametophyte) were observed from the early stage(Female Gametophyte1, FG1) to the late stage(FG5) in the developing ovules of NIL(HZ396). Therefore, these facts indicate that meiosis was initiated in most of the megasporecytes in NIL(HZ396) and progressed at least to anaphase II; however, the female meiosis ceased before the successful formation of a functional haploid(tetrad stage) in 65.6% ovules. We also scored the initial ovule numbers per silique of both NILs and compared them with the corresponding NSS. The ratio(65.6%) of abnormal female gametophytes was very close to the proportion(69.8%) of abortion ovules per silique in NIL(HZ396), indicating that the developmental arrest of megasporogenesis is directly responsible for the reduction of NSS in NIL(HZ396). RNA-seq analysis indicated that most of meiotic cell cycle-related genes were down-regulated expression in NIL(HZ396). In addition, we observed an identical and normal pollen development in both NILs. Hence, it can be speculated that BnaC9.SMG7 b has a conserved function for progression out of meiosis II like Arabidopsis SMG7 but the role is specifically restricted in ovules.RNA-seq data indicated that alternative splicing(AS) events were similar between NIL(HZ396) and NIL(Y106). Furthermore, we analyzed the expression levels of PTC-containing transcripts(+PTC) and normal transcripts(-PTC) from the alternative splicing mRNA products of a randomly selected single copy gene, BnaA03g28670 D. qPCR analysis showed that the accumulation of +PTC transcripts was not significantly elevated in NIL(HZ396) compared with in NIL(Y106), which is in accordance with the fact that no obvious difference in vegetative growth was observed between the two NILs. In addition, we transformed the complementation construct carrying Bna C9.SMG7 b into the Arabidopsis null SMG7 mutant(smg7-3). For all the resulting transgene-positive plants of the T2 generation, BnaC9.SMG7 b failed to complement the vegetative and developmental defects caused by the dysfunction of AtSMG7. Moreover, the altered expression level of +PTC transcripts(At2g45670) in smg7-3 was not reversed by the introduction of BnaC9.SMG7 b. Therefore, these data indicate that unlike AtSMG7, BnaC9.SMG7 b is not involved in the NMD process.