Association Mapping Reveals Novel Genetic Loci Contributing to Flooding Tolerance during Germination in Indica Rice

Abstract

Rice (Oryza sativa L.) is the only cereal crop that possesses the ability to germinate under flooded or other oxygen-deficient conditions. Rapid elongation of the coleoptile is a perfect response to flooding during germination, with coleoptile length differing among various rice varieties. Despite multiple studies have uncovered valuable information concerning this trait by focusing on the physiological metabolism of oxygen stress, the underlying genetic mechanism still remains unknown. In the present study, we screened coleoptile lengths of 432 indica varieties germinated in two environments (normal and flooded) and found more variation existing in flooded coleoptile length (FCL) rather than in normal coleoptile length (NCL). With the phenotypic data of NCL, FCL and FTI (flooding tolerance index), a genome-wide association study was performed by using 5291 single nucleotide polymorphism (SNP) markers. We detected 2, 11, and 9 significant SNPs under a mixed linear mode for NCL, FCL, and FTI, respectively. Of these SNPs, five were shared by FCL and FTI. Haplotype and phenotype effect analysis on the highest ranking locus indicated one of the two haplotypes contributed to coleoptile elongation remarkably. To better understand the controlling gene of this locus, reported expression profile data was applied. We focused on LOC_Os06g03520, a candidate gene which was highly induced by anoxia (∼507 fold). Sequence analysis in 51 varieties demonstrated Hap.2 associated perfectly with flooding tolerance. Further studies on this gene may help explore the molecular mechanism of rice flooding tolerance during germination. We believe our discoveries may conduce to isolating major genes and aid the improvement of flooding tolerance in modern breeding programs.

Keywords: coleoptile; flooding; genome-wide association study (GWAS); germination; rice (Oryza sativa L.).

FIGURE 1

Principal component analysis and relative kinship analysis of the indica population. (A) The percentage of genetic variation explained by each of the first 10 principal components (PCs). (B) Distribution of relative kinship among 432 indica accessions, K represents relative kinship coefficients.

FIGURE 2

Manhattan plots and quantile–quantile plots of genome-wide association analysis. Manhattan plots, the red straight line show the threshold of P = 0.001; quantile–quantile plot, the red straight line represent the expected null distribution of P-values, the blue dots represent the observed distribution of P-values. (A) NCL, normal coleoptile length; (B) FCL, flooded coleoptile length; (C) FTI, flooding tolerance index.

FIGURE 3

Statistical analysis and candidate region estimation of seq-rs2699 and seq-rs2701. (A) Haplotypes consist of the two significant SNPs, Numbers indicates the amounts of corresponding accessions; (B) Phenotypic effect of each haplotype; (C) Local manhattan plots and LD heatmap around the peak on chromosome 6, the candidate region estimated using r² > 0.6.

FIGURE 4

Analysis of candidate gene LOC_Os06g03520. (A) Gene structure of LOC_Os06g03520, the blue regions show the exons; (B) eleven haplotypes detected in 51 varieties with extreme phenotypes; (C,D) phenotypic distribution of the 51 varieties, x-axis indicates numbers sort by accessions ID; y-axis represents phenotypes. (C) FCL, flooded coleoptile length, (D) FTI, flooding tolerance index.