Genome-wide Association Analysis Of Silique Density Of Major Inflorescence And Seed Density Within Per Silique In Brassica Napus L.

Rapeseed(Brassica napus),as one of the four oil crops,is not only widely cultivated in the world,but also one of the most important oil crops in China.At present,rapeseed breeding experts regard high yield as one of the main goals.As fructification organ,silique plays the dual role of "sink" and "source",and it is directly related to the accumulation of photosynthetic products,which may affect the yield of rapeseed.Silique per plant and seed per silique is the constituent elements of rapeseed yield.Silique density and seed density within per silique will affect silique per plant and seed per silique respectively,which have a direct or indirect effect on rapeseed yield.Therefore,increasing the density of rape silique and seed density within per silique become a possible way to improve the yield of rapeseed,and research on silique density of major inflorescence,seed density within per silique and their related traits have important significance for rapeseed breeding.In this study,213 natural populations of Brassica napus accessions(or lines)with different geographical origins were studied.A total of 23,767 SNP markers covering the whole genome were selected to assess population structure,relative kinship and linkage disequilibrium.Then the optimal model of each trait was selected,and genome-wide association studies were conducted with six traits.The main results are as follows:1.Phenotypic analysis of Silique density of major inflorescence,Seed density per within silique and their related traits: The results showed that the individuals in the natural population had different characteristics,and extensive phenotypic variations were observed for all traits.All of them were normally distributed or approximate normally distributed,which are characterized by typical quantitative traits.The phenotypic correlation coefficient of all traits indicated that there was a significant correlation between multiple traits.2.Analysis of population structure,relative kinship and LD: In this study,the natural populations of 213 Brassica napus could be divided into two subsets of P1 and P2 by a model-based population structure analysis using 8923 SNP markers.P1 subgroup includes 50 materials(23.5%),derived from Europen,Hubei province,the majority of the spring rape and a small amount of semi-winter material;P2 subgroup includes 163 materials(76.5%),most of them were from China and belonged to semi-winter oilseed rape.In the association panel,about 89.74 % kinship between any two lines is less than 0.02,59.91% kinship coefficient between any two lines was 0.These results indicate that the relationship between the natural population is relatively far,which has little effect on the association analysis results.By calculating the r2 value of pairwise markers in the same chromosome,we find that there was a big difference in decay distance among different chromosome.The average LD decay distance of A genome is bigger than C genome,indicating that Brassica rapa is widely used in breeding.3.Selection of optimal model for six traits: The optimum model whose observation P-values closest to the expected P-value,was chosen among the six models,i.e.GLM,Q,PCA,K,K+PCA,K+Q for GWAS.The Q model was the optimal model of seed density within per silique,valid silique length,valid length of major inflorescence in 2016 and valid length of major inflorescence in 2015.The optimal model for the seed per silique and valid length of major inflorescenc in 2015 was the PCA model.The K + Q model was chosen as the optimal model of silique density of major inflorescence and valid silique of major inflorescence in 2016.4.Association mapping of Silique density of major inflorescence,seed density within per silique and their related traits: A genomewide association analysis was performed using 23,767 SNP markers for the six traits.A total of 27 SNP markers,which were distributed on different chromosomes,related to the six traits were detected in two years(P<4.21×10-5).Some of these SNP sites are clustered on the same chromosome.The study found 17 SNP sites associated with silique density of major inflorescence and its related traits.7 SNP sites associated with silique density of major inflorescence were detected,and the individual locus accounting for 11.34%~15.96% of the total phenotypic variation.9 SNP sites associated with valid length of major inflorescence were detected,and single locus could explain the phenotypic variation of 9.67%~13.10%.1 SNP site associated with valid silique of major inflorescence,explaining the phenotypic variation of 11.56%.There were 10 SNP sites associated with seed density within per silique and its related traits.2 SNP sites associated with seed density within per silique were detected,explaining the phenotypic variation of 9.94% and 11.17%,respectively.6 SNP sites associated with valid silique length were detected,and the individual locus accounting for 9.81%~12.17% of the phenotypic variation.2 SNP sites associated with seed per silique were detected,explaining the phenotypic variation of 10.44% and 10.87%,respectively.5.Candidate genes prediction: By analyzing the genes in the LD region of associated SNP locus,we screened 38 candidate genes associated with all traits,including some unknown functional candidate genes located on or close to significant SNP sites.22 candidate genes related to valid silique of major inflorescence and its related traits were found.BnaA01g16940 D,GASA6,IBR5,GPS1,NPH4,RBX1,HOT5 and AHP3 were though to control silique density of major inflorescence and its related traits through the way of synthesis and signal transduction of endogenous hormones such as GA and IAA.ULT1,AtSK12,NSN1,OBE1 and RPK2 genes are regulators of the meristems or floral meristems to change the phenotype.Other genes,such as Mob1 and NAC036,are involved in cell division,cell growth and other processes of altering the phenotype.There were 16 candidate genes related to seed density within per silique and its related traits.KMD4 and UGT76C2 genes are involved in the regulation of cytokinin,and the genes i.e.AGL104 and ADC2 was found to be related to seed formation process.Other genes,such as MCCB、NGA2 and MATE,are involved in the growth and development of lateral organs,when abnormal expression lead to lateral organ variants.Two candidate genes ADC and UGT76C2,which overlapped in seed density within per silique and seed per silique,tended to have pleiotropism.