The Analysis Of Rice Grain Notched-kernel Traits And Fine Mapping Of A Novel Grain Shape Gene

Rice is one of the most important economic crops. Improving quality and yield potential of ricegrain is not only related to farmers’ income, but also to national food security. In the present study, QTLanalysis and map-based clone method was employed to analyze notched-belly kernel traits and grainshape traits1. We constructed a molecular marker genetic linkage map by F₂population derived from the crossof L1042（notched-kernel rice line） and Nanjing11（normal rice variety）. Two notched-kernel traits,percentage of notched grains （PNG） and notched severity （NS）, were found to be continuouslydistibuted and manifested a quantitative traits genetic model decided by multiple genes in the F₂population. Using Inclusive Composite Interval Mapping software, eleven quantitative trait loci （QTLs）including7PNG and4NS QTLs for notched-kernel traits were detected out, among which fourL1042-originated PNG, i.e. QTLs qPNG-2, qPNG-3, qPNG-5and qPNG-6could explain from9.3%to16.6%phenotypic variance of PNG, and two major L1042-originated NS QTLs, qNS-1and qNS-3,accounted for52.5%and42.9%of phenotypic variance, respectively. The QTL located between I3-11and S3-15-3interval of chromosome3contained qPNG-3and qNS-3simultaneously, suggesting thepresence of a real notched grain-related gene.2. QTL analysis for rice grain shape trait was performed in the same F₂population as above usingInclusive Composite Interval Mapping software. The results indicated that the grain shape traits werecontinuously distributed and exhibited quantitative-trait characteristics. Seventeen QTLs distributed on6chromosomes, including4grain length QTLs,6grain width QTLs, and7LWR QTLs. The majorgrain length QTL, qGL-1and qGL-3, could explain14.2%and17.0%phenotypic variance of grainlength, respectively. Three major grain width QTL, qGW-2a, qGW-2b and qGW-2c, which werecentralized on2chromosomes could account for14.0%,22.2%and23.5%phenotypic variance of grainwidth, respectively. The major length to width raitio QTL, qLWR-1and qLWR-2a could explain21.8%and13.1%phenotypic variance of grain LWR respectively. The major grain length QTL （qGL-1andqGL-3） and the major LWR QTL （qLWR-1and qLWR-2a） in this paper are reported for the first time.3. Grain shape is the important agronomic traits, which are related to grain yield. A mutant withlonger grain width, grain length and grain thickness more thousand-kernel weight was identified fromtissue-cultured progenies of Kittake. The mutant also showed reduced height, seed setting rate, grainnumber per spike. Genetic analysis indicated that the phenotype of grain width was controlled by asingle dominant gene. Using F₂-F5population derived from a cross between the mutant and an indicacultivar PA64, the mutant gene was narrowed down to a78kb between two InDel markers, LK64andLK57, on the short arm of chromosome2. Six open reading frames were predicted in this region, amongwhich the WD-domain ORF plays an important role in endosperm differentiation and development.However, there was no difference among six known genes by genome sequence comparison, thereforewe speculate the mutant phenomenon could be due to methylation and promoter mutation.