Integration of QTL Mapping and Whole Genome Sequencing Identifies Candidate Genes for Alkalinity Tolerance in Rice (Oryza sativa)

Abstract

Soil alkalinity is an important stressor that impairs crop growth and development, resulting in reduced crop productivity. Unlike salinity stress, research efforts to understand the mechanism of plant adaptation to alkaline stress is limited in rice, a major staple food for the world population. We evaluated a population of 193 recombinant inbred lines (RIL) developed from a cross between Cocodrie and N22 under alkaline stress at the seedling stage. Using a linkage map consisting of 4849 SNP markers, 42 additive QTLs were identified. There were seven genomic regions where two or more QTLs for multiple traits colocalized. Three important QTL clusters were targeted, and several candidate genes were identified based on high impact variants using whole genome sequences (WGS) of both parents and differential expression in response to alkalinity stress. These genes included two expressed protein genes, the glucan endo-1,3-beta-glucosidase precursor, F-box domain-containing proteins, double-stranded RNA-binding motif-containing protein, aquaporin protein, receptor kinase-like protein, semialdehyde hydrogenase, and NAD-binding domain-containing protein genes. Tolerance to alkaline stress in Cocodrie was most likely due to the low Na⁺/K⁺ ratio resulting from reduced accumulation of Na⁺ ions and higher accumulation of K⁺ in roots and shoots. Our study demonstrated the utility of integrating QTL mapping with WGS to identify the candidate genes in the QTL regions. The QTLs and candidate genes originating from the tolerant parent Cocodrie should be targeted for introgression to improve alkalinity tolerance in rice and to elucidate the molecular basis of alkali tolerance.

Keywords: Na+/K+ ratio; Oryza sativa; abiotic stress; genotyping-by-sequencing; quantitative trait loci; seedling stage.

Figure 1

Comparison of performance of Cocodrie and N22 under (upper panel) control and (lower panel) alkaline stress conditions, respectively.

Figure 2

Frequency distribution of Cocodrie/N22 RIL population for seedling stage alkalinity tolerance for various morphological and physiological traits. The arrowhead indicates the trait mean for Cocodrie (C), N22 (N), and RIL population (R).

Figure 3

Comparison of Cocodrie and N22 for various physiological and morphological traits under alkaline stress and non-stress environments. Single asterisk indicates significant difference between the means of Cocodrie and N22 for stressed and control environment at 0.05 level of probability.

Figure 4

Map positions of the QTLs for eleven morphological and physiological traits in the Cocodrie × N22 RIL population. N22 and Cocodrie alleles responsible for the increased mean are indicated in black and red font, respectively. Dark regions on the genetic map are the marker-saturated regions; light regions represent gaps between the markers.

Figure 5

Expression profiles of eight selected genes present in the alkalinity tolerance QTL regions under alkalinity stress (6 h after imposition of stress) in Cocodrie and N22. The selection was based on differences in high-impact variants between the parents. Selected genes included: 1-LOC_Os10g35040 (receptor kinase-like protein); 2-LOC_Os10g34000 (aquaporin protein); 3-LOC_Os10g35170 (semialdehyde dehydrogenase, NAD-binding domain-containing protein); 4-LOC_Os10g33970 (double-stranded RNA-binding motif-containing protein); 5-LOC_Os09g32550 (glucan endo-1,3-beta-glucosidase precursor); 6-LOC_Os09g32860 (OsFBX335-F-box domain-containing protein); 7-LOC_Os08g01560 (expressed protein); and 8-LOC_Os08g01720 (expressed protein). EF1α was used as the reference gene and gene expressions were expressed as log2 fold changes under alkaline stress compared with control in both parents.