Evolution And Association Analysis Of GHD7, GW2,GS5and QSW5in Rice

Rice is one of the most important crops in the world. As exploring population across the globe, it is urgent for rice breeders to improve its production and agronomic traits. Ghd7is an important regulator of heading date, plant height and yield in rice. GW2, qSW5, GS5and GS3are four important genes that dominate rice grain size. Candidate gene association analysis, taking advantages of abundance natural variations of a population, is very effective in excellent allele mining and functional SNP discovery. Moreover, a molecular evolution model of a gene implies important information of its function, which is very helpful in improving rice breeding.Firstly, Ghd7was deep sequenced in a natural population of104varieties,8haplotypes was discovered. Ghd7has at least two ancestors through a comparison to Ghd7alleles of3wild rice. A functional SNP, which is associated with plant height but not to heading date or yield in all three environments, is placed in gene promoter. Further analysis demonstrates that an expression level change caused by this SNP is efficient for the plant height variation. In addition, H4haplotype of Ghd7, which shows a big panicle along with modest plant height and heading date, is a favorable allele in rice breeding.Secondly, GW2, qSW5and GS5were deep sequenced in a natural population of127varieties. The evolutionary difference was detected between the3genes by genotype comparison to15wild rice. Of them, GW2and qSW5suffer natural selection during their evolution history, while GS5shows higher nucleotide diversity in its promoter. Moreover, genotype of another grain length gene, GS3, was also detected. The function of the4grain size gene was compared and the effect of genotype combinations were also estimated in this population. Although loss of function variation of GW2is not detected in this population, significant difference of grain length is found between two haplotypes, referred to GW2-2and GW2-3. Further experiment shows that this variation does not dependent on gene expression level, so it might be caused by the interaction with GS3. However, GS5contains a relatively small effect, and its function is masked by qSW5in natural population. And this may explain the reason why GS5escape from selection during evolution history. GS3and qSW5are two most effective genes that related to grain length and width respectively, but the combination of gs3and qsw5genotypes does not result in a big grain size. These above results are helpful for understanding the molecular mechanism of these genes and provide a potential for flexibility application of them in rice breeding programs.