| Rice is an important crop in the world. With the reduction of arable land and increasing population, improving rice production has become the important way to solve the problem of food security. The development of hybrid rice has made an important contribution to improve rice yields. However, to date, hybrid indica rice accounts for approximately 80% of the total area of indica rice, hybrid japonica rice only accounts for approximately 5% of the total area of japonica rice in China. A huge gap between them brings an opportunity to the development of hybrid rice. Rice grain-filling is an important trait which affects grain weight and thus directly affects grain yield. Large panicle is one of the main superiorities of F1 hybrid rice, while the grain-filling and grain plumpness in hybrid japonica rice are poor on large panicles, which is the main manifestation of heterosis of F1 hybrid rice. Poor grain-filling and inferior grain plumpness result in decreases of not only the grain yield but also the commodity value. Therefore, the study of genetic basis and mechanism of rice grain-filling rate is important for determining rice breeding strategy of high quality and yield. The main results obtained from the two studies conducted are concluded separately as follows.1. Map-based cloning of rice grain-filling rate gene GFR1Previous studies in our laboratory has detected a main effect QTL qGFR10 controlling grain filling rate by using a backcross recombinant inbred lines (BC1F6) derived from crossing between two japonica rice varieties Ludao and C-bao (designated as GFRI in this study) .To study the genetic basis of GFR1, we isolated a NIL plant (NIL-gfrl) containing the allele of Ludao at the GFR1 locus with the background of C-bao from the chromosome segment substitution lines. Compared with C-bao, the grain-filling rate of NIL-gfr1 was significantly higher than that of C-bao at the early stages. There was no significant difference in the heading date, panicle number per plant, panicle length, number of spikelets per panicle and number of filled grains per panicle; while the seed setting rate,grain plumpness, grain weight and yield per plot of NIL-gfr1 were significant higher than C-bao, which indicated that the grain-filling rate increase of NIL-gfr1 can improve rice yields.A single segment segregating F2 population derived by crossing a single segment substitution line with C-bao was used to confirm and finely map the major QTL GFR1 controlling grain-filling rate. Genetic analysis indicated that GFR1 was a single Mendelian factor. Using 1614 plants with fast grain-filling rate in NIL-gfr1/C-bao F2 population, we narrowed GFR1 down to a 34-kb genomic region at the end of long arm on chromosomel.Four putative ORFs were predicted in this region. By comparing the sequences, we found that a deletion of three consecutive bases and a single nucleotide polymorphisms (SNP)were detected in the CDS of ORF1 (LOC_Os10g36400) between the Ludao allele(designated GFRl-1) and C-bao allele (designated GFRl-c). These differences result in a delection of alanine and a change from valine to alanine in amino acid sequence encoded by Ludao. Transgenic complementary and transgenic overexpression demonstrated that ORF1 is GFR1.Two male sterile lines, 9522A and Xu2A, were crossed, respectively, with C-bao and NIL-gfr1 to produce four combinations, 9522A/C-bao, Xu2A/C-bao, 9522A/NIL-gfr1 and Xu2A/NIL-gfr1 to evaluate the useful of GFR1 in hybrid rice breeding. The results indicated that the grain-filling rate, grain plumness and grain yield of F1 crossings beween NIL-gfr1 and sterile lines were higher than the F1 of crossings between C-bao and the sterile lines. Therefore, it is possible to improve grain filling rate of parents for yield increase of hybrid rice by using allele of GFR1 in Ludao.2. Detection of elite alleles of rice grain-filling rate in natural populationNinety-five varieties including 58 landraces and 37 elite varieties from the core germplasm collection were genotyped with 263 simple sequence repeat (SSR) markers.The GFR of the 95 varieties was evaluated at five stages, 7, 14, 21, 28 and 35 days after flowering (DAF) both in 2011 and 2012.There were significant differences in the GFR among the 95 varieties at five stages in both 2011 and 2012, with the CV ranging from 36.49% to 118.26% and broad sense heritability ranging from 87.5% to 96.8%. A population structure analysis identified seven subpopulations. A linkage disequilibrium (LD) analysis indicated that the levels of LD ranged from 60.3 cM to 84.8 cM and artificial selection had enhanced the LD of landraces.A time-course association analysis detected 24 SSR markers located on chromosomes 1, 2, 3, 4, 5, 6, 8, 9, 11 and 12 of rice at five stages. Amomg them, 12, 8, 2, 5 and 4 markers were detected at five grian-filling stages, respectively, in which seven markers ((RM480,RM5818, RM525, RM6361, RM6314, RM224 and RM72) were associated with two grian filling stages. The marker RM528 associated with grain-filling rate at 14DAF on chromosome 6 has the highest phenotypic variations, 25.87% and 27.19% in 2011 and 2012,respectively.Totally fifteen excellent parental combinations were predicted in this study, and the best parental combination is 'Nannongjing62401×Laolaihong',which could theoretically increase 4.086 mg grain-1 d-1 at the five stages.These carrier varieties with elite alleles can be used as parents to improve the GFR via pyramiding breeding. |
PDF Full Text Request