Genome-Wide Association Studies Reveal that Diverse Heading Date Genes Respond to Short and Long Day Lengths between Indica and Japonica Rice

Abstract

Rice is a short-day plant. Short-day length promotes heading, and long-day length suppresses heading. Many studies have evaluated rice heading in field conditions in which some individuals in the population were exposed to various day lengths, including short and long days, prior to a growth phase transition. In this study, we investigated heading date under natural short-day conditions (SD) and long-day conditions (LD) for 100s of accessions and separately conducted genome-wide association studies within indica and japonica subpopulations. Under LD, three and four quantitative trait loci (QTLs) were identified in indica and japonica subpopulations, respectively, two of which were less than 80 kb from the known genes Hd17 and Ghd7. But no common QTLs were detected in both subpopulations. Under SD, six QTLs were detected in indica, three of which were less than 80 kb from the known heading date genes Ghd7, Ehd1, and RCN1. But no QTLs were detected in japonica subpopulation. qHd3 under SD and qHd4 under LD were two novel major QTLs, which deserve isolation in the future. Eleven known heading date genes were used to test the power of association mapping at the haplotype level. Hd17, Ghd7, Ehd1, and RCN1 were again detected at more significant level and three additional genes, Hd3a, OsMADS56, and Ghd7.1, were detected. However, of the detected seven genes, only one gene, Hd17, was commonly detected in both subpopulations and two genes, Ghd7 and Ghd7.1, were commonly detected in indica subpopulation under both conditions. Moreover, haplotype analysis identified favorable haplotypes of Ghd7 and OsMADS56 for breeding design. In conclusion, diverse heading date genes/QTLs between indica and japonica subpopulations responded to SD and LD, and haplotype-level association mapping was more powerful than SNP-level association in rice.

Minor Effects of 11 Dof Family Genes Contribute to the Missing Heritability of Heading Date in Rice (Oryza sativa L.)

“…Like our Dof family genes, whose effects are too small to be significantly associated with heading date in GWAS (Supplementary Figure S2). Heading date shows high heritability (0.75), but quantitative trait loci (QTLs) revealed by GWAS only explain 40.6% of the phenotype variation (Han et al, 2016). In this study, 22 Dof genes were associated with heading date via candidate genebased association analysis at the haplotype level (Supplementary Figure S4).…”

“…Heading date shows high heritability in linkage mapping populations and germplasm collections. Moreover, these core heading date genes explain only a fraction of the total heritability in both GWAS and biparental mapping populations (Xue et al, 2008;Huang et al, 2010;Yan et al, 2011;Huang et al, 2012b;Yan et al, 2013;Han et al, 2016;Yano et al, 2016). There must be many genes with minor effects spread widely across the genome.…”

“…A total of 529 rice accessions from around the world, composed of a core Chinese collection of 202 cultivars and a core world collection of 327 cultivars, were grown in Wuhan (114°21′E, 30°28′N; the average day length is more than 13.5 h from the middle of May to the beginning of August; 23-31°C) in the summer of 2013 and Hainan (110°01′E, 18°30′ N; the average day length was less than 12.5 h from December to the middle of March; 20-28°C) in the winter of 2013 (https:// www.timeanddate.com, https://www.climatestotravel.com). The details of field management and measurement of heading date are described in a previous study (Han et al, 2016). The 529 accessions were genotyped via sequencing and classified into four subpopulations: indica, aus, japonica, and Admixture (Chen et al, 2014).…”

“…Heading date had a high heritability of 0.75 across both conditions. However, SNP-based GWAS only identified a few loci that were significantly associated with heading date, and they only explained a small fraction of heading date variance (Han et al, 2016). Dof family genes were not significantly associated with heading date in GWAS (Supplementary Figure S2).…”

“…Candidate gene-based association analysis is an option to identify potential functions of related genes. Moreover, association mapping at the haplotype level is a very effective method to establish the relationships between genes and traits (Han et al, 2016;Chen et al, 2018;Lu et al, 2018). Haplotype development is easy at the single gene level, which is helpful to identify the association between related genes and traits at the haplotype level.…”