QTL Mapping For Agronomic And Kernel Morphological Traits In Wheat And Candidate Genes Predicting

In wheat(Triticum aestivum L.),most agronomic and kernel morphological traits were quantitative character.At present,two major methods(linkage analysis and genome-wide association analysis)were employed in studying quantitative trait.In this study,we used linkage analysis method to map quantitative trait locus(QTL)for the important agronomic traits and kernel morphological traits of wheat in multi environments,and employed genome-wide association analysis to study kernel morphological traits at different nitrogen levels.This study would provide adequate knowledge on the genetic mechanism of agronomic traits and kernel morphological traits for wheat,and could be a reference for marker-assisted selection(MAS)of these traits in breeding.The main results were as follows:1)Based on the phenotypical data from an introgression line(IL)population derived from Lumai 14 × Shaanhan 8675 planted in 10 environments,QTL analysis was conducted for agronomic traits using a SSR genetic map.In total,48 additive-effect QTLs and 833 pairs of epistatic-effect QTLs were detected,and 40 of the 48 additive-effect QTLs were involved in epistatic interactions.The 48 additive-effect QTLs were distributed on chromosome 1A,2A,3A,4A,6 A,7 A,1B,2B,3B,5B,6B,7B,2D,5D and 7D respectively,with LOD value ranging from 2.50 to 12.91 and explaining 1.33-22.70% of phenotypic variation.The QTL number for plant height,spike length,heading data,kernel number per spike,thousand-kernel weight,number of valid tiller,fertile spikelet number per main spike,grain weight per plant were 6,8,4,8,6,5,7 and 4 respectively.Especially,among these QTLs,q HD-1A,q PH-6A,q SL-2A and q HD-2D could be found across 8,7,5 and 5 environments,indicating that they constantly expressed in different environments.2)Moreover,wheat kernel morphological traits,such as kernel length,kernel width,kernel thickness,kernel length/width ratio,kernel length/thickness ratio,and kernel width/thickness ratio,are studied.The phenotypical data included a doubled-haploid(DH)population(derived from ‘Hanxuan 10' × ‘Lumai 14')identified under six environments,a BC3F6 population(derived from ‘Lumai 14' × ‘Jing 411',IL population 1)under three environments,and a BC3F6 population(derived from ‘Lumai 14' × ‘Shaanhan 8675',IL population 2)under three environments.Then 46 additive QTLs were detected in the DH population,and 20 additive QTLs were detected in each of the IL populations.Especially,a kernel-width QTL,q KW-6A,was located within the same interval in all three populations.q KT-7A-3,q LTR-4A,and q WTR-7A-1 mapped in the DH population were located in the same marker intervals as q KT-7A-1,q LTR-4A,and q WTR-7A-1,respectively,in IL population 2.q LWR-5A-2,q WTR-5A-2,and q WTR-5A-1 from the DH population were detected in five,four,and three environments,and explained 14.72,25.11,and 25.91%,respectively,of the phenotypic variation.q LTR-7A from IL population 1 and q LWR-5B from IL population 2,detected in all three environments,explained 6.10 and 10.66% of the phenotypic variation,respectively.On the other hand,negative alleles of q KW-6A for kernel width detected in all three populations were derived from ‘Lumai 14'.And under drought-stress conditions,the lines with this donor QTL segment obtained averaged increased effects of 5.22 and 4.52% in IL population 1 and IL population 2,respectively,in the two years,which indicated that q KW-6A played a key role during the development of kernel width.3)Using the above DH population,we further constructed a high-density genetic linkage map with 1630 SNP and 224 SSR markers,then analyzed the QTLs for kernel morphological traits under six environments.Finally,a total of 133 additive-effect QTLs and 29 pairs of epistatic-effect QTLs were identified.The 133 additive-effect QTLs distributed on nearly all the 21 chromosomes of wheat except 2A,and LOD value ranged from 2.53 to 25.85.They each explained phenotypic variation within the range of 0.89-42.65%.The number of QTL controlling kernel length,kernel length,kernel width,kernel thickness,kernel length/width ratio,kernel length/thickness ratio,and kernel width/thickness ratio were 26,14,21,31,18 and 23,respectively.Bioinformatics analysis showed that the gene Traes CS6A02G238000,locating on the segment of q KW-22-4(q KW-6A)and encoding beta-1,3-galactosyltransferase GALT1,was the most possible candidate gene for kernel width.4)Using 660 K gene chip technology,genome-wide association analysis were conducted for the kernel morphological traits of 374 wheat varieties grown under normal nitrogen level(CK)and low nitrogen level(LW),respectively.A total of 54 SNP markers were significantly associated with six traits of kernel morphological phenotype at P?0.0001.Among them,10 loci could be detected under two environments,indicating they might be included in constitutively expressed genes that were necessary for the development of wheat kernel morphology.In addition,there were 18 loci specially identified under CK,and we speculated that the 18 loci were associated with genes affecting kernel morphology under normal nitrogen conditions.Also 26 SNPs were specially detected under LW,indicating that the genes involved in these loci might be induced by low nitrogen level and played important roles in nitrogen utilization efficient.