| The chief objective in maize genetic and breeding research is to improve grain quality and yield. Grain quality and yield are all complex traits controlled by many genes. Previous researchs about the molecular genetic mechanism for grain yield and quality have been conducted using normal maize germplasms and high-oil maize. But the genetic relationship between kernel quality traits and yield components were seldom researched. In this study, an elite high-oil maize inbred line GY220 was crossed with two elite normal corn inbred lines 8984 and 8622 to produce two connected RIL populations with 282 (Pop.1) and 263 (Pop.2) F7:8 families, respectively. The field experiments were conducted under four different environments, and 4 kernel quality traits, 8 yield components and 10 plant traits were measured. Two high-density genetic maps were constructed by SSR markers. QTL analysis for every traits was conducted by composite interval mapping (CIM) method under each environment and in average analysis, the interaction between detected QTL were analysis by multiple interval mapping (MIM) method, and joint QTL analysis among main yeild components were done using CIM method of multiple traits analysis. By using meta-analysis method in Biomercator 2.1, intergrated map was construted for the two connected RIL population genetic maps and consensus QTLs were identified. Our first chief objective was to reveal the molecular genetic mechanism of kernel quality traits, yield components and plant traits, and theirs genetic relationship. The second objective was to obtain major QTL with consistency across genetic backgrounds, generations and environments for kernel quality traits, yield components and plant traits. These results will provide help for breeding in improving grain yield and quality, for marker assisted selection, and fine mapping and gene cloning of key QTL.The main results were as follows:1. In total, 666 SSR primers were chose to screen polymorphism between two pairs of parents, GY220 and 8984, GY220 and 8622. 228 and 217 markers were in co-dominant segregation, respectively. 216 and 208 SSR markers were selected respectively to construct the maize genetic linkage maps with the genetic distance of 2 285.3 cM and 2 217.2 cM (centimorgan) and an average of 10.58 cM and 10.66 cM. Using bioMercator 2.1, the linkage length of the integrated map for two RIL populations was 2 349.90 cM, which included 313 SSR markers, and was 7.51 cM average distances between markers.2. Variances of genotypes, environments and genotype×environment interactions were significant or high significant for most traits for 3 kernel quality traits, 8 yield components and 10 plant traits in both RIL populations. All traits showed normal distribution in the two RIL populations with a wide range of variation and transgressive segregation exceeding both parent values. The broad sense heritability (hB2) estimates for all traits were different. In total, those for Lysine, EWP, GWP, KR and LOV were low. The tendency of phenotypic and genotypic correlations among traits was almost the same under every environmrnt across the two populations. Positive correlations were consistently observed for oil with protein and lysine, GWP and EWP with 100GW, EL, GPR and ED, 100GW with ED and EWP, EL with GPR and EWP, ED with RPE and EWP, PH with EH, TH and TL, EH with LNE, TH with TL, and THPH with TH, while consistent negative correlations were observed between starch with protein, 100GW and both GPR and RPE.3. Eighty-one QTL for 4 kernel quality traits, 173 for 8 yield components and 247 for 10 plant traits were detected using two RIL populations under 4 environments individually and in average, 9, 21 and 48 QTL among theirs explained up to 10% phenotypic variation, 0, 5 and 15 QTL explained up to 15%, respectively. Three, 5 and 4 QTL were stable among genetic backgrounds and environments. Interactions among detected QTL were few and the effects were low. Using multiple traits analysis, some new QTL were detected by joint analysis, which showed pleiotropy or tight linkage.4. For all the 81 QTL detected in both RIL populations for 4 kernel quality traits, 13 mQTL were detected using meta-analysis. They included 65 QTL and 47 QTL explained above 5% effect of phenotypic variations, accounting for 80.23% and 58.02%, respectively. mqQTL1-1, mqQTL3-4, mqQTL6-1 and mqQTL8-3 were the clusters of QTL for grain quality traits. mqQTL4-1 and mqQTL6-1 might be clustered with candidate genes which control the same and (or) different metabolic process, and further fine mapping those major QTL for grain oil content in these regions might be worth to be conducted. Those candidate genes, such as akh1, bt2, gpc1, glt1, su1, su3, dgat1, dzs23, o14, su2, PG9, zSTSII-1 and ploc1 could be cloned in further research. The contribution of favorable alleles for QTL included in mqQTL3-4, mqQTL5-1, mqQTL8-2, mqQTL8-3 and mqQTL9-2 were different between starch and oil, protein and lysine. It showed the gene for these traits might be pleiotropy and (or) tightly linked.5. For all the 173 QTL detected in both RIL populations for 8 yield components, 22 mQTL were detected using meta-analysis.They included 152 QTL and 102 QTL explained above 5% effect of phenotypic variations, accounting for 87.86% and 58.96%, respectively. myQTL1-1, myQTL3-2, myQTL7-1, myQTL8-1, myQTL9-1 and myQTL10-2 were the clusters of QTL for yield components. myQTL1-1, myQTL6-1, myQTL6-2, myQTL8-2, myQTL8-3 and myQTL9-2 might be clustered with candidate genes which control the same and (or) different metabolic process. Further research could be done to fine map major QTL for 100GW, GWP, KR and EL. Those candidate genes, such as Bx9, ACA1, ENO2, GAPC2, czog2, PG9, cko3, HPR and tgp could be cloned in the same regions. The contribution of favorable alleles for QTL included in myQTL1-2 and myQTL5-2 were different between EL with GPR and ED with RPE. It showed the gene for EL with GPR and ED with RPE might be pleiotropy and (or) tight linkage.6. For all the 247 QTL detected in both RIL populations for 10 plant traits, 27 mQTL were detected using meta-analysis. They included 216 QTL and 174 QTL explained above 5% effect of phenotypic variations, accounting for 87.44% and 70.45%, respectively. mpQTL1-3, mpQTL2-2, mpQTL3-1, mpQTL3-2, mpQTL3-4 and mpQTL8-2 were the clusters of QTL for plant traits. mpQTL1-1, mpQTL1-2, mpQTL1-4 and mpQTL8-1 might be clustered with candidate genes which control the same and (or) different metabolic process. Further research could be done to fine map major QTL for LNE, LOV and EH in these regions. Those candidate genes, such as LTP-2, COP11, SIP2-1, LPE1, ipk, rd1, tlr1, ct1 and clm1 would be cloned in the same origins. The contribution of favorable alleles for QTL included in mpQTL5-1 and mpQTL8-2 were different between PH and THPH. It showed the gene for these traits might be pleiotropy and (or) tight linkage.7. The improvement of kernel quality and grain yield in maize might be obtained through detection the molecular genetic mechanism for each kernel quality traits and yield components. Significant negative correlations were observed between oil with KR, protein with EWP, GWP, 100GW, EL, GPR and KR, and significant positive correlations were observed between starch with EWP, GWP, 100GW, GPR and KR. The contribution of favorable alleles for QTL located at chromosome 1, 3 and 8 were different between grain quality and yield components. It showed the gene for these traits might be pleiotropy and (or) tight linkage and the molecular genetic mechanism was pleiotropy and (or) tight linkage QTL controlled these traits. Accumulation and introgression of these chromosome interval with the QTL for 100GW at bins 1.05 (umc1603), 3.04-3.05; 100GW, EL and GPR at bin 7.02-7.03; and EWP and starch at bin 9.02-9.05 might increase grain weight while keeping kernel quality, and accumulation and introgression of these chromosome interval with the QTL for oil at bin 6.03-6.04 might increase kernel oil content while keeping high grain yield. |
PDF Full Text Request