| Doubled haploid (DH) lines produced via in vivo haploid induction have become indispensable in maize research and practical breeding, so it is important to understand traits characteristics and correlation in DH and its corresponding haploids. The stalk is the primary institutions of the whole plant, which involved with the water and nutrition transport and affecting the lodge-resistance and silage quality. During the whole developmental stage of plant, the ideal plant architecture allows for more efficient light capture and solar energy utilization during photosynthesis, which could eventually maximize the unit area yield. It is critical to further understand and analyze the genetic architecture of stalk quality-related and plant morphology-related traits.In this study, a DH population derived from Zheng58xChang7-2 (ZD958) and its corresponding haploid population, were developed, genotyped and evaluated to investigate phenotypic characteristics, genetic architecture and the correlations of 28 stalk quality-related and plant morphology-related traits crossing two years and three locations. This study could provide the scientific basis for genetic study and practical breeding, and the major results of this study are listed as follows:1. A DH population including 181 DH lines was genotyped using with MaizeSNP3K chip. A subset of 1137 polymorphic SNPs with minor allele frequency of >0.05, missing rate <20% were left and used to construct a linkage map. The total length of the linkage map was 1426.83 cM with an average interval of 1.26 cM.2. Both in the DH and haploid populations, the phenotypic values of the traits evaluated in our study showed a large range. The genotype had significant effect on all traits. The heritability of all traits ranged from 0.49 to 0.96. These results suggested that the traits variation in two different ploidy populations were mainly controlled by genotype and the QTL analysis could be carried out.3. Correlation analysis indicated that the correlation coefficient between two population for all traits were 0.43-0.76. RPR was significantly (P<0.01) negatively correlated with IVDMD, WC, CP and WSC, while was significantly (P<0.01) positively correlated with FW, ADF, NDF and Cel. There was no significant correlation between RPR and Lig in two different ploidy populations. IVDMD was significantly (P<0.01) negatively correlated with WC, ADF, NDF and Cel, while was significantly (P<0.01) positively correlated with Lig, CP and WSC.There was no significant correlation between IUE-ID and PEH/FIAG-IL, the pairwise correlation of other plant morphology-related traits were significantly (P<0.01) positively. Compare the correlation in DH population to that in haploid population, the coefficient changed. Via QTL analysis for 28 traits, we got 107 and 78 QTL in DH and haploid population, respectively. The explained phenotypic variation of each QTL ranged from 1.60% to 24.10%, in which there was 54 QTL accounting for 10.20%-20.49% of phenotypic variation and distributing on 1.04-1.09bin, 1.11bin,2.02-2.03bin, 3.05bin,3.08bin,5.01-5.03bin,5.05bin,6.01bin,6.03-6.04bin,7.00bin,7.03-7.05bin and 8.04-8.05bin. A total of 33 common QTL were detected for the same traits between DH and haploid population. Traits with significant correlation were found always with co-located or close linked QTL.4. For RPR and IVDMD which associated with lodge-resistance and silage quality, we screened the genes located in the QTLs detected in 2.02bin,5.05bin and 8.05bin, the potential candidate genes predicated from the QTL support intervals for RPR and IVDMD were indirectly or directly involved with cellulose and lignin biosynthesis, which participated in cell wall formation.Seven genes-PAL, COMT, ccoAOMT, CCR, CAD, CesAlland CesA12, were homoloyly cloned. For most genes in miik stage and ripen stage, the relative gene expression in two haploid parent were higher 0.62-14.06 times than that in two corresponding diploid parents, which suggest that dosage compensation mechianism may be one of the main causes resulting in the common QTL detected bwteen DH and haploid population.5. Using both DH and haploid populations, GS was more appropriate for the selection of PHT, PEH and IUE-IL, while QTL-based marker-assisted selection was more suitable to FIAG-IL, IUE-ID and LL. For FIAG-ID and LW, the selection efficiency of GS and QTL-based marker-assisted selection was differnet in two ploidy populaitons, so during the selective breeding process in different ploidy population, the method should be selected based on the comprehensive factors, such as triats, marker density, QTL effect value, population types and so on.6. For plant quanlity-realted and morphology-related traits evaluated in this study, the ratio of PG_TS/h2 was<0.60 (except for Lig content), indicating that conventional phenotypic selection is more efficient than the QTL-based marker-assisted selection for improving those traits. Nevertheless, the chromosome regions harboring "ploidy change-insensitive QTL" and "ploidy change-sensitive QTL" which accounting for more than 10% of phenotypic variation could be used as a starting point for fine mapping, gene cloning, the effect mechanism and model of ploidy on gene expression and QTL presentation. |
PDF Full Text Request