| Rice (Oryza sativa.L) is one of the most important crops in the world, and yield related traits were always focused by researchers. Like other quantitative traits, rice yield traits are usually controlled by multiple loci. The panicle is a performance of rice yield, plays an important role in yield. In present study, the 1,840 recombinant inbred lines (RILs) were derived by single-seed descents from a cross between Oryza sativa ssp. indica cv.93-11 and ssp. indica cv. PA64s. A whole-genome sequencing based genotyping of the RIL was used for QTL mapping. Whole-genome sequencing acquired SNPs were used to construct recombination bins across the RIL population. A recombination bin represents a chromosomal block that is separated from the adjacent bin by at least 1 recombination within the RIL population. A linkage map was constructed, and the map contained 3,524 recombination blocks with the average length of 105.6 kb, which was in parallel to the expected minimum observation. The graphic genotypes were converted into a genetic map with total genetic distance of 1381.9 cM and an average of □0.392 cM be-tween adjacent recombinant blocks. Based on the result of QTL analysis, we can narrow down QTL fragments or extract genes using map-base clone strategy.In this study, one population of RILs derived from an indica two-lines-hybrid Liang-You-Pei-Jiu (LYPJ) was used. Meanwhile,132 RILs were selected to whole genome re-sequencing and constructed a high-density genetic map. Panicle related traits include panicle length, panicle number, primary panicle branch number, secondary panicle branch number, spikelet number, spikelet density, and yield per plant were surveyed, and used for QTL analysis. Other remaining 1,708 RILs were used to fine map 3 stable QTLs, named qPPB3, qSPBl, and qSN8. At the same time, one of stable QTL qPPB3 was dissected into single gene using BC3F2 populations derived from the cross between a target CSSL and genetic background parent,93-11 and InDel markers. The results should be useful for MAS breeding and map-based cloning of target QTLs. Main conclusions are as follows:1. Phenotypic performance of panicle related traits in parents and RILsPanicle length, spikelet per panicle, secondary branch number per panicle, and yield per plant showed significantly differenc between parents. In this study, SN was significantly positive related to PL, PPB, and SPB. In two different environments, PN or PL mainly affected YPP. The relationships of SN, PPB, and PL were negative to PN. Basic statistics for these traits in RILs showed that skewness and kurtosis of all traits were smaller than 1.0, except PN and YPP in 11HN. Their characteristics proved to fit for QTL analysis.2. QTL analysis of panicle related traits using RILs populationIn this study, a mapping population of 132 F14 RILs, derived from a cross between a Oryza sativa ssp. indica varity 93-11 and a photoperiod/temperature sensitive male sterile line PA64s by the single seed descent method, was used to detect quantitative trait loci (QTLs) for panicle related traits. Results indicated that 31 QTLs were extracted by the software MultiQTL 1.6, these QTLs distributed on each chromosome in rice. They accounted for 6%-16.5% of total phenotypic variation. Five of six QTLs (qPPB3, qPPB4, qPPB6, qPPB7a, and qPPB7b) for PPB were detected on chromosomes 3,4,6, and 7, and PA64s contributed favorable alleles (less primary panicle branch number). The other QTL qPPB8 for PPB was detected on chromosome 8, and the favorable allele was from 93-11. Six QTLs for PL were detected on chromosomes 1,4,5, 8,9, and 11, they explained 6%-13.5% phenotypic variation. PA64s contributed favorable alleles to qPL1, qPL9, and qPLll,93-11 contributed favorable alleles to qPL4, qPL5, and qPL8. Only one QTL was detected for PN on chromosome 5, and explained 13.4% phenotypic variation, the favorable allele of qPN5 was from PA64s. For SPB,4 QTLs were extracted. Explained 6.2%～10.3% phenotypic variation, favorable alleles of qSPBl, qSPB9a, and qSPB9b were from 93-11, and qSPB3 was from PA64s. Three QTLs (qSD3, qSD4, and qSDll) were detected for SD that explained 10%-16.5% phenotypic variation, and PA64s contributed favorable alleles to the three QTLs. Five QTLs were determined for SN to explain 7.9%-13.4% phenotypic variation, four of them on chromosomes 2,8,9, and 10 achieved alleles of 93-11, and PA64s contributed favorable allele to the other on chromosome 4. For the last trait, five QTLs were detected for YPP on chromosomes 4,6,8,11, and 12,93-11 contributed favorable alleles to three of them (qYPP4, qYPP6, and qYPP8) and PA64s to the others. Among these QTLs,4 pleiotropic loci were found. qSD4 and qSN4 on the region of chromosome 4 between SNP4-208 and SNP4-236 controlled spikelet density and spikelet number. Fragment of chromosome 8 between markers SNP8-40 and SNP8-71 controlled spikelet number, primary panicle branch number, and yield per plant. qSPB9b and qSN9 were detected on the same region of chromosome 9 between two markers SNP9-93 and SNP9-119, the fragment regulated secondary panicle branch number and spikelet number. The forth fragment was determined on chromosome 11 between two markers SNP11-123 and SNP11-171, it controlled panicle length and yield per plant.According to previous study, many of these detected QTLs were associated with reported genes or QTLs. These results provided basic materials and direction to MAS breeding and map-base clone of panicle related traits.3. Fine mapping and functional analysis of panicle related traits QTLsBased on the QTL analysis for panicle related traits in two different environments (2011 Hainan and 2011Hangzhou) using a mapping population of recombinant inbred lines (132 lines), derived from LYPJ. Using phenotyped remaining population (1,708 lines), linked-analysis combined with developed markers that were designed from InDel between parents was conducted. qPPB3, qSPBl, and qSN8 were further narrowed to 385kb,2.5kb, and 9.83kb, respectively. To further narrow down the region of qPPB3, a secondary F2 population was derived from a cross between targeted CSSL-GH18 and genetic background,93-11. After self-crossing, a BC3F2 population was used to fine mapping qPPB3 to a 34.6kb region.Analysis of candidate genes, D88 was suggested to be the candidate gene to qPPB3, there were 3 SNPs between the parents, the third SNP on exon 3 resulted in amino acid sequence changed. LAX1 was the only one candidate gene to qSPBl, one SNP was discovered between the parents. qSN8 was determined to the pleiotropic gene DTH8, and one SNP and one 8bp InDel in promoter region of DTH8 between the parents. A 3.57kb genomic DNA fragment of DTH8 from 93-11 was introduced into PA64s by Agrobacterium-mediated transformation, all T1 transgenic plant with DTH8 alleles of 93-11 exhibited phenotypes of late heading with increased spikelet number.Expression levels of 4 panicle related genes reported by others and 3 genes in this study was researched by real-time PCR, expression levels of D88, APO1, IPA1, DTH8, and LAX1 in PA64s was lower than that in 93-11, but expression levels of GN1a and DST was higher than that in 93-11. Lower expression of D88 resulted in less primary panicle branch number but higher tiller in PA64s, indicated that D88 regulated stigolactones level during the entire plant development. Stigolactones was a important hormone not only controlled shoot branching, but also regulated panicle branching. Besause of higher expression level of DTH8 in 93-11, larger spikelet number was obviously discovered in 93-11. Secondary panicle branch number of PA64s was reduced dramatically by expression level of LAX1 decline. Shoot branching, panicle branching, and yield related genes had unknown relationship, the rice yield related genes had pleiotropy to reach high yield. Pramiding of these important genes to one plant, and a ideal variety will be selected by breeders. |
PDF Full Text Request