Genetic Analysis Of Ovule Number And Fine Mapping Of QON-A10 In Brassica Napus L

Oilseed rape(Brassica napus L)is an allotetraploid species of Brassicaceae family.It is an important oil crop around the world and has various secondary uses.During the breeding history of oilseed rape,more focus has been given to improve yielding ability to meet the increasingly global demands for oil and other uses.Seed number per silique(SN),as one of the components of individual plant,directly determines the final seed yield,while the SN is largely controlled by the ovule number per ovary(ON).Therefore,to increase the yield potential of oilseed rape,much more attention is required to be paid to SN than ever before.Here,to understand the genetic basis of ON,two independent populations,including a double haploid(DH)population and a collection of natural inbreeding accessions,were utilized.The biparental DH population comprised of 180 lines was developed from the two inbreeding accessions,i.e.,7-5 and ZY50,differing largely in ON.The collection of natural inbreeding accessions consisted of 505 inbred accessions from diverse geographic regions.The DH population was planted in three consecutive seasons(2015-2018)for phenotyping,and the data was integrated for QTL analysis.The natural population was planted in two consecutive years(2016-2018)at four different locations(Wuhan,Sichuan and Ezhou).The ON phenotype data was subject to linkage and association analysis.The main results are as follows;1.Phenotypic analysis showed that both of the DH and natural populations display a wide range of variations in the ON across all the environments.The DH population showed a normal distribution and transgressive segregation for ON suggesting its suitability for QTL mapping analysis.Similarly,the natural population showed normal distribution for ON across different locations.The ON showed a board sense heritability of 0.86 and 0.93 for the DH and natural populations,respectively.2.A high density linkage map consisting of 2237 SNP and 65 SSR markers and covers a map length of 1951.5 c M with the average marker distance of 0.85 c M,was used for linkage mapping.Five QTLs were detected for ON,including qON-A3,qONA7-1,qON-A7-2,qON-A10,qON-C6.Three of them,qON-A7-1,qON-A7-2 and qON-A10,explained the phenotypic variations(PVE)of 15.5%,17.38% and 9.28%,respectively,which seemed to be the QTL with major effects.Except for qON-A3,other QTLs were repeatedly detected in two or more environments.3.Genome-wide Association Mapping(GWAS)identified several ON associated loci in the natural population.Single locus GWAS(SL-GWAS)was carried out individual environment and best linear unbiased prediction(BLUP)data.SL-GWAS identified 214 significant SNPs associated with ON in BLUP data.The PVE of these SNPs ranged from 5.03% to 7.33%.Multi locus GWAS(ML-GWAS)individually identified 40 and 48 ON-associated significant SNP via the Mr MLM and FASTMr MLM methods.Thess SNPs explain 3.73 to 21.20% of the total phenotypic variation.These SNPs are distributed on all chromosomes except for A05,C05 and C07.4.Integration of linkage mapping and association mapping identified four consensus genomic region associated with ON on chromosome A103,A07 and A10,respectively.The consensus genomic regions of chromosome A07 and A10 were further analyzed for candidate genes.Genes within 150 kb flanking region of significant SNPs were extracted.A total of 269 genes distributed in the flanking around 12 significant SNPs located in qON-A07-2,while 145 genes flanked around the 6 SNPs located in qONA-10.Based on the expression differences between the parental lines,54 and 37 genes were selected from the candidate regions of qON-A07-2 and qON-A10 as possible candidate genes.5.The qON-A10 was validated in the backcross populations from BC2F2 to BC5F2.The ON was significantly different between the ZY50-NILsnd 7-5-NILs in each population.Genotyping of the BC4F2 population in the ZY50 background identified 9 different recombinant types,for which the progeny testing of recombinant families was carried out.Based on the genotypes and phenotypes of these recombinant families,qON-A10 region was narrowed down to a 490-kb region on chromosome A10.Further fine mapping narrowed down the 490-kb region to a 90.6-kb interval.Within the target genomic fragment,18 coding genes were identified based on the Darmor-bzh genome.Among these,only four genes(Bna A10g15110 D,Bna A10g15120 D,Bna A10g15190 D and Bna A10g15220D)showed contrasting expression in buds between the parental lines.Regional association analysis of the delimited interval of qON-A10 identified three SNPs in Bna A10g15100 D.The sequence analysis of Bna A10g15100 D revealed that that these SNP are located in the second exon of Bna A10g15100 D.These SNPs grouped the natural population into two halophytes(Hap1 and Hap2).The Hap2 carries altered allele and have higher ON than Hap1.In summary,we examined the variation of ON in oilseed rape and identified the genomic loci controlling ON variations via linkage and association mapping.Validation and fine mapping of the consensus qON-A10 region identified possible candidate genes that might regulate the ovule developmental processes.This study provides primary insights to the genetic control of ON,not only constructing a solid foundation for the gene cloning of the candidate gene but also releasing some valuable molecular marker linked to ON for the yield enhancement in in oilseed rape.