Haplotype analysis of key genes governing grain yield and quality traits across 3K RG panel reveals scope for the development of tailor-made rice with enhanced genetic gains

Abstract

Though several genes governing various major traits have been reported in rice, their superior haplotype combinations for developing ideal variety remains elusive. In this study, haplotype analysis of 120 previously functionally characterized genes, influencing grain yield (87 genes) and grain quality (33 genes) revealed significant variations in the 3K rice genome (RG) panel. For selected genes, meta-expression analysis using already available datasets along with co-expression network provided insights at systems level. Also, we conducted candidate gene based association study for the 120 genes and identified 21 strongly associated genes governing 10-grain yield and quality traits. We report superior haplotypes upon phenotyping the subset of 3K RG panel, SD1-H8 with haplotype frequency (HF) of 30.13% in 3K RG panel, MOC1-H9 (HF: 23.08%), IPA1-H14 (HF: 6.64%), DEP3-H2 (HF: 5.59%), DEP1-H2 (HF: 37.53%), SP1-H3 (HF: 5.05%), LAX1-H5 (HF: 1.56%), LP-H13 (3.64%), OSH1-H4 (5.52%), PHD1-H14 (HF: 15.21%), AGO7-H15 (HF: 3.33%), ROC5-H2 (31.42%), RSR1-H8 (HF: 4.20%) and OsNAS3-H2 (HF: 1.00%). For heading date, Ghd7-H8 (HF: 3.08%), TOB1-H10 (HF: 4.60%) flowered early, Ghd7-H14 (HF: 42.60%), TRX1-H9 (HF: 27.97%), OsVIL3-H14 (HF: 1.72%) for medium duration flowering, while Ghd7-H6 (HF: 1.65%), SNB-H9 (HF: 9.35%) were late flowering. GS5-H4 (HF: 65.84%) attributed slender, GS5-H5 (HF: 29.00%), GW2-H2 (HF: 4.13%) were medium slender and GS5-H9 (HF: 2.15%) for bold grains. Furthermore, haplotype analysis explained possible genetic basis for superiority of selected mega-varieties. Overall, this study suggests the possibility for developing next-generation tailor-made rice with superior haplotype combinations of target genes suiting future food and nutritional demands via haplotype-based breeding.

Keywords: haplotype-based breeding; 3K RG panel; cloned genes; haplotype mining.

Figure 1

Haplotype analysis of 120 cloned key genes associated with major grain yield and quality related traits in rice. The chosen 120 genes were previously functionally characterized and were reported to govern/regulate several major grain yield (87 genes) and quality related traits (33 genes) such as tiller number, flowering, panicle architecture, lodging resistance, single plant yield, grain amylose content, grain Fe and Zn concentration, grain size etc. For all these genes haplotype analysis was conducted and the numbers in blue within brackets indicates the number of haplotypes for that particular gene in the 3K RG panel. Interestingly, more than 75% of the genes had haplotypes ranging from 2 to 15 among the 3024 lines. This is a gold mine for the identification of superior haplotypes and utilizing the same in developing elite versions of rice via haplotype‐based breeding.

Figure 2

Establishing subset of 3K RG panel. A core panel comprising of 150 lines was developed based on diversity analysis using 559, 297 SNPs. (a) The chosen lines were from diverse geographical locations (32 countries). (b) Kinship among the 150 lines. (c) Two significant principle components were found in the panel. (d) SNP based UPGMA tree reveals the presence of two major clusters with five subgroups within the first major cluster and two subgroups in second major cluster.

Figure 3

Correlation analysis of various grain yield and quality related traits in the subset of 3K RG panel. Days to flowering, tiller number and panicle length were positively associated with single plant yield. Plant height is negatively related with days to flowering and on the other hand positively linked with panicle length. It was observed that the panicle length and tiller number were positively correlated and was negatively related to days to flowering and grain size. Grain Fe and Zn concentration were positively correlated. DTF, Days to Flowering; PBN‐Primary branch number per panicle and SPY‐Single plant yield; *P < 0.05, **P < 0.01, ***P < 0.005.

Figure 4

Haplotype analysis of Ghd7 and OsNAS3 across the 3K RG panel. (a) Ghd7, a key gene associated with heading date has about 14 haplotypes in the 3K RG panel with wide phenotypic variations. (b) Ghd7‐H8 was the most diverse one based on SNP and (c) interestingly was the earliest to flower, while Ghd7‐H6 took greater than 100 days to flower. (d) OsNAS3 that influences grain Fe and Zn concentration has three haplotypes in the 3K RG panel with significant phenotypic variations in the subset. OsNAS3‐H2 had the highest grain (e) Fe and (f) Zn profile. The geographical distribution of various haplotypes of (g) Ghd7 and (h) OsNAS3.

Figure 5

The tailored rice with superior haplotypes for grain yield and quality. The findings of this study could be employed to develop a designer rice genotype comprising superior haplotype combinations of the target genes such as MOC1‐H9 & IPA1‐H14 for higher tiller number, Ghd7‐H8 & TOB1‐H10 for early, Ghd7‐H14 & OsVIL3‐H14 for medium duration, Ghd7‐H6, SNB ‐H9 & TRX1‐H9 for late flowering, DEP3‐H2, DEP1‐H2 & SP1‐H3 for long panicles, SD1‐H8 for semi‐dwarf nature, LAX1‐H5, OSH1‐H4 & LP ‐H13 resulting in increased panicle branching, PHD1‐H14, AGO7‐H15 & ROC5‐H2 for high yield, along with GS5‐H4 for slender, GS5‐H5 & GW2‐H2 for medium slender, GS5‐H9 for bold grains, RSR1‐H8 for intermediate amylose content and OsNAS3‐H2 for increased Fe and Zn concentration in grains. B, Bold grain; MS, Medium slender type grain; S, Slender type grain.