Variation of Preharvest Sprouting Resistance and Its Associated Genetic Loci in Rice

Abstract

Genetic resources, reservoir of diversity during evolution process, are essential sources of valuable alleles for crop improvement program and bio-industry. Among the plant genetic resources, rice is an important crop for a carbohydrate source over the world and is widely used for genomic studies as a model plant because of relatively small genome size about 370Mb. Useful traits have been selected during the domestication, and the resistance for pre-harvest sprouting (PHS) caused by early breakage of dormancy is one of the domesticated traits. In that PHS reduces grain yields quality in cereal crops, its regulation is crucial for cereal crop improvement under the global climate changes. To unveil molecular basis for PHS as a quantitative trait, we firstly surveyed naturally occurred variations of seed germinability in diverse rice germplasm. Among 205 rice accessions, 90 accessions were resistant to PHS, and the accessions of tropic and subtropic regions revealed more resistance to PHS. The increase of germination ratio after detaching from panicle at harvesting time was detected, and this might be caused by a weakened mechanical barrier. Phytohormone abscisic acid (ABA) and gibberellins (GAs) are known for its role in embryo-imposed dormancy which is acquired during the developmental process. In this study, ABA levels was maximized at 10 DAF and decreased thereafter similar to previous reports, but PHS susceptible accessions maintain high or similar level of ABA compared to PHS resistant accessions. In this regard, ABAs perception and signal transduction might be key factors for the seed dormancy acquisition and its breakage is in natural population. To survey significant polymorphism between PHS resistant and susceptible group, we sequenced representative 21 accessions and acquired approximately a total of 4.27 million genome-wide SNPs. Among these SNPs, about 20 thousand SNP loci showed distinguished allele distribution between PHS resistant and susceptible group by enrichment analysis, and some regions presenting high number of differentiated SNP loci were overlapped with previously reported QTLs. Among the distinguished SNPs by PHS group, we selected 39 candidate SNP loci for further analysis and 144 rice germplasm were genotyped in these loci. Finally, six and seven SNP loci revealed significant allele distribution in japonica rice for PHS and GI, respectively, and three loci were significant for PHS and GI in indica rice. Based on the finally acquired SNP loci which significantly associated with PHS, we constructed regression equation for PHS (adjusted R2 of 0.401 in japonica germplasm), and validation of this regression equation showed R-square value of 0.296 (p-values < 0.001) between predicted values and actual measurements. The relative low value of adjusted R2 and correlation between predicted and actual measurements is reasonable in that PHS as a quantitative trait is affected by complex genetic factors, and further study might increase the resolution power for the phenomenon of seed germination and dormancy.