QTL Mapping For Kernel Nutritional Quality Characters And Their Genetic Relationship In Maize

High-oil corn is fastly developing in quality maize breeding with its comprehensive kernel nutritional quality traits and higher commercial value. Previous researches about QTL mapping for kernel nutritional quality traits have been conducted with IHO, IHP and BHO high-oil corn germplasms and using single genetic background and testcross populations. In this study, 284 and 265 F2 populations and their F2:3 family lines were respectively developed from two crosses between the same high-oil corn inbred GY220 and two normal corn inbreds, 8984 and 8622. The germplasm background of GY220 was different from IHO and BHO. SSR markers were used to construct high-density genetic maps. QTL associated with three kernel nutritional quality characters were detected according to the LOD thresholds after 1000 permutations. Using composite interval mapping (CIM) method, QTL mapping of kernel oil, protein and starch concentration was done for the two environments individually and together. The interactions of detected QTL were identified using multiple interval mapping (MIM) method according to the result of CIM method. Conditional QTL mapping and joint QTL analysis for two or three different characters were done using CIM method of multiple traits analysis among three kernel nutritional quality and two main yield component characters. Our objectives were to reveal the molecular genetic mechanism of these characters and the genetic correlations among different kernel nutritional quality characters, and between these characters and yield components, and to select effective molecular markers with stable expression across different populations and environments. These results will do great help in fine mapping QTL associated with kernel nutritional quality characters and their map-based cloning, and in marker-assisted selection in high-oil maize breeding. The main results in this study were as follows:1. Totally, 665 SSR primers were employed to screen polymorphism between parents of two crosses, 8984×GY220 (P1) and 8622×GY220 (P2). 212 and 205 polymorphism markers were selected, respectively. Great differences in polymorphism markers were found between the two sets of parents. Only 104 pairs of markers were in common, which accounted for 49.1% and 50.7%, respectively. 185 and 173 pairs of SSR markers were selected respectively to construct the maize genetic linkage maps with the genetic distance of 2111.7 cM and 2298.5 cM (centimorgan) and an average of 11.41 cM and 13.29 cM using Mapmaker 3.0b.2. 53 QTL were detected for three kernel nutritional quality characters using both F2 and F2:3 families of the two crosses under two environments. 10 and 22 QTL were detected in P1F2 population and P1F2:3 families, repectively. In P2F2 population and P2F2:3 families 6 and 19 QTL were detected. No common QTL were detected in both populations. 3. 22 QTL for kernel oil concentration were detected. Of these, 18 QTL were on identical or similar chromosomal bin locations with previous studies. qzOIL1-10-1 and qzOIL1-4-1 had the same chromosome regions as endosperm mutant genes accB and bt2, respectively. 5 and 7 QTL were detected in P1F2 population and P1F2:3 families, and 3 and 9 QTL for oil in P2F2 population and P2F2:3 families. These QTL were located on chromosome 1, 3, 4, 5, 6, 7, 8 and 10. Two identical QTL were detected in both F2 population and F2:3 families from the same cross. No identical QTL was detected in both F2:3 families. 5 QTL were simultaneously detected in F2:3 families under one environment and the average of two enviroments. The contribution of single QTL to phenotypic variation varied from 5.0% to 26.5%. Of 12 QTL with contributions higher than 10%, 11 QTL were detected by previous researchers. qlOIL2-8-1 and qzOIL2-6-1 had stability between different environments and consistency in different studies. They could be used as the main objevtive QTL in further studies and in MAS. Except one, all the favorable alleles of 21 QTL were from the high-oil corn parent GY220, which clearly showed high degree of kernel oil gene pyramiding through long-term selection.4. Of 12 QTL for kernel protein concentration, 9 QTL were on identical or similar chromosomal bin locations with previous studies. qpPRO1-3-1 had the same chromosome regions as endosperm mutant gene sh2. 3 and 6 protein QTL were detected in P1F2 population and P1F2:3 families. No protein QTL was detected in P2F2 population. 3 QTL were detected in P2F2:3 families. These QTL were located on chromosome 1, 3, 5, 6, 8, 9 and 10. No identical QTL was detected in both F2 population and F2:3 families, and in F2:3 families under two environments. 2 QTL were simultaneously detected in F2:3 families under one environment and the average of two enviroments. The contribution of single QTL to phenotypic variation varied from 3.9% to 17.5%. Of 6 QTL with contributions more than 10%, 4 QTL were detected by previous researchers. qxPRO1-8-1 and qzPRO1-1-1 could be used as the main objective QTL in further studies and in MAS with their stability between different environments and consistency between different studies. The favorable alleles of 5 QTL were contributed by GY220, and 5 and 2 QTL were from 8984 and 8622, repectively.5. 19 QTL for kernel starch concentration were detected. 9 QTL were on identical or similar chromosomal bin locations with those previous studies. qpSTA1-5-2 had the same chromosome regions to endosperm mutant gene ae1. 2 and 9 starch QTL were detected in P1F2 population and P1F2:3 families, and 3 and 7 QTL in P2F2 population and P2F2:3 families. These QTL were located on chromosome 2, 4, 5, 6, 8 and 10. Three identical QTL were detected in both F2 population and F2:3 families from the same cross. One identical QTL was detected in F2:3 families under two environments. Two QTL were simultaneously detected in F2:3 families under one environment and the average of two enviroments. The contribution of single QTL to phenotypic variation varied from 2.9% to14.0%. Of 6 QTL with contributions higher than 10%, 5 QTL were detected by previous researchers. qzSTA2-8-1 and qzSTA2-6-1 had stability between different environments and consistency between different studies, which could be used as the main objective QTL in further studies and in MAS. In all favorable QTL for starch concentration, only one was from GY220, and 10 and 8 QTL were from 8984 and 8622, repectively.6. The numbers of QTL expressing additive, partially dominance, dominance and over-dominance effects were 12, 25, 7, 9. Partially dominance and additive effects all played important role in the heredity of kernel nutritional quality characters. Very low interaction effects between detected QTL were estimated.7. The results of conditional QTL mapping showed that 100-kernel weight and ear-kernel weight had the greatest effect on kernel protein concentration. Their effects on kernel oil and starch concentration were also significant. Kernel oil concentration was also affected greatly by kernel protein and starch concentration. 3 and 2 QTL associated with oil and starch concentration were not affected by other characters, which could be pyramided in single QTL level. The QTL controlling different kernel nutritional quality characters mapped on chromosome 1, 2, 4, 5, 6, 7, 8 and 10, and QTL for them and yield components on chromosome 1, 2, 4, 5, 6, 7, 8,9 and 10, showed pleiotropy or tight linkage.