Contributions Of Grain-Size Major QTL To Grain Weight And Map-Based Cloning Of GS3.2, A Grain-Length Gene In Rice

Rice yield is determined by three yield components, such as number of grains per panicle, panicle number and grain weight. Grain weight is always affected by rice grain size. Grain size is not only a yield trait but also an appearance quality trait. On one hand, grain size is employed as a useful genetic tool in understanding the genetic basis of rice yield, on the other hand, grain size could also be employed as a powerful tool for breeding to increase the grain yield. In this study, grain-size QTL, mainly detected from two RILs between indica varieties, were used to develope QTL-NILs. Based on these QTL-NILs, these QTL were fine mapped and GS3.2 was cloned by a map-based cloning strategy. Meanwhile, according to the combinated QTL-NILs with qGL3, qGW2a and qGS5, the genetic effects and the interactors were discussed for the formation of grain weight. The main results are as follows:1. Three grain-size QTL were employed, which were qGS5 identified as a minnor QTL for grain width in RIL of MH63/ZS97B, qGW2a and qGL3 identified as two major QTL for grain width and grain length in RIL of SLG/ZS97B. NILs of the three QTL have been developped and employed to build an F1 (qGL3/qGW2a/qGS5), a combined QTL-NIL in ZS97B background. Finally, an F2 population of 957 individuals were derived from F1 (qGL3/qGW2alqGS5). By performing epistatic QTL and ANOVA analysis, no notable interaction between the three grain-size QTL were detected. The additive effects among these three grain-size QTL played the most important roles in the formation of grain weight. The genetic effects of the three grain-size QTL in grain weight formation were qGW2a> qGL3> qGS5.2. qGL3, a major QTL for grain length, was identified in a RIL population from a cross of two indica varieties, ZS97B and SLG, which was located on chromosome 3 in the region flanked by two SSR markers, RM156 and MRG2513. To analyse qGL3, a BC3F2 population of 396 individuals was developed. According to QTL analysis, two QTL, qGS3.1 and qGS3.2, were detected in the region of qGL3. qGS3.2 mainly affects grain length and explained grain length variation by 27.31% with additive effects of 0.65 mm provided by SLG allele. Based on the F2 population derived from the qGS3.2-NIL, qGS3.2 was finally located in region of about 16.2 kb between the two markers, SNP1 (SNP marker) and S161 (STS marker). In this region, there were 3 candidate genes, named as Os2、Os3 and Os4.3. By transformation experiments of three candidate genes, the QTL gene for grain length was confirmed. By the over-expression transformation, it was found that Os4 showed no phenotype difference in grain length. Based on the complementary transformation and mutant analysis of Os3, it was shown that Os3 was not GS3.2. By the results of over-expression, RNAi and complementary transformation of Os2, the grain length of transgenic plants were varied followed the expectation. It was proved that Os2 was GS3.2. According to the results, it was concluded that the GS3.2 is a minor grain-length QTL gene, located in the qGS3.2 region and the functional differences were mainly caused by the expression level of GS3.2.4. For biological function analysis of GS3.2, parafin section was carried out by employing the two qGS3.2-NIL types. It was found that qGS3.2 mainly affects the cell number in direction vertical. According to the Real-Time PCR results, the relative expression between GS3.2 and cell cycle genes were analyzed. It was suggested that GS3.2 affected cell number variations mainly by affecting 5 cell cycle genes.5. According to the results of GS3.2, we speculated function area might locate in the promoter region.189 micro-core collection were detected in the promoter region of GS3.2 based on the SNPs and In/Dels between the two alleles from ZS97B and SLG. According to the analysis, GS3.2 promoters were classified into 4 genotypes:Type 1, type 2, type 3 and type 4. The long-grain variety in our study, SLG, was belongs to Type 1, and the short-grain variety, ZS97B, was belongs to type 4. The 4 GS3.2 promoter genotypes were prevalent in both indica and japonica subgroups. By the analysis of natural variations in the micro-core collections in the GS3.2 promoter, it could be hypothesized that functional microarray may be from-1400 bp to -2000 bp, including 7 SNPs and 4 In/Dels.