Identification of superior haplotypes and candidate gene for seed size-related traits in soybean (Glycine max L.)

Abstract

Seed size is an economically important trait that directly determines the seed yield in soybean. In the current investigation, we used an integrated strategy of linkage mapping, association mapping, haplotype analysis and candidate gene analysis to determine the genetic makeup of four seed size-related traits viz., 100-seed weight (HSW), seed area (SA), seed length (SL), and seed width (SW) in soybean. Linkage mapping identified a total of 23 quantitative trait loci (QTL) associated with four seed size-related traits in the F₂ population; among them, 17 were detected as novel QTLs, whereas the remaining six viz., qHSW3-1, qHSW4-1, qHSW18-1, qHSW19-1, qSL4-1 and qSW6-1 have been previously identified. Six out of 23 QTLs were major possessing phenotypic variation explained (PVE) ≥ 10%. Besides, the four QTL Clusters/QTL Hotspots harboring multiple QTLs for different seed size-related traits were identified on Chr.04, Chr.16, Chr.19 and Chr.20. Genome-wide association study (GWAS) identified a total of 62 SNPs significantly associated with the four seed size-related traits. Interestingly, the QTL viz., qHSW18-1 was identified by both linkage mapping and GWAS, and was regarded as the most stable loci regulating HSW in soybean. In-silico, sequencing and qRT-PCR analysis identified the Glyma.18G242400 as the most potential candidate gene underlying the qHSW18-1 for regulating HSW. Moreover, three haplotype blocks viz., Hap2, Hap6A and Hap6B were identified for the SW trait, and one haplotype was identified within the Glyma.18G242400 for the HSW. These four haplotypes harbor three to seven haplotype alleles across the association mapping panel of 350 soybean accessions, regulating the seed size from lowest to highest through intermediate phenotypes. Hence, the outcome of the current investigation can be utilized as a potential genetic and genomic resource for breeding the improved seed size in soybean.

Supplementary information: The online version contains supplementary material available at 10.1007/s11032-024-01525-1.

Keywords: Genetic mapping; Haplotype analysis; QTL; Seed size; Soybean.

Fig. 1

Phenotypic differences in seed size-related traits between the two parents. A Seed image demonstrating the seed size among the two parental soybean accessions viz., Heihe50 (HH50) and Dongnong60 (DN60); (B) Mean comparison of the seed size-related traits using student-t test (*indicates significant difference at p < 0.05, **indicates significant difference at p < 0.01, *** indicates significant difference at p < 0.001 and **** indicates significant difference at p < 0.0001); (C) Frequency distribution of four seed size-related traits in the F₂ population

Fig. 2

QTLs detected for four seed size-related traits in the F₂ population. A QTLs of 100-seed weight (HSW); (B) QTLs of seed area (SA); (C) QTLs of seed length (SL); and (D) QTLs of seed width (SW). The blue and red dashed lines represent LOD values of 2.5 and 3.0, respectively

Fig. 3

Distribution of the 23 QTLs and QTL Clusters/QTL Hotspots represented by partial genetic map identified for four seed size-related traits viz., 100-seed weight (HSW-yellow color), seed area (SA-pink color), seed length (SL-blue color) and seed width (SW-green color) across the 11 chromosomes. The QTLs for different seed size-related traits are represented by bars with different colors

Fig. 4

The Marker distribution, population structure and linkage disequilibrium (LD) analysis of 350 soybean accessions. A The presence of the 3,343,372 SNPs across 20 soybean chromosomes. Length of chromosomes (Mb) is represented by the horizontal axis, chromosome number is denoted by the vertical axis, and SNP density is depicted by the different colors (number of SNPs per window); (B) Relationship of 350 soybean accessions depicted by a kinship plot; (C) Genome-wide distribution of 3,343,372 SNP markers that are used for GWAS; (D) Population structure analysis of 350 soybean accessions (JL-Jilin province; HLJ-Heilonjiang province; LN-Liaoning province; HHH-Huanghuaihai region; and XN-Xinan region). E LD decay plot of 350 soybean cultivars using 3,343,372 SNP markers. The LD decay fitted with a smoothing spline regression model is represented by the red curve line. The blue vertical line intersection with the horizontal green line represents the half-decay of LD (r.^.2 = 0.34), and the genetic distance at this point corresponds to LD decay distance (71.6 kb)

Fig. 5

Genome-wide association study (GWAS) of four seed size-related traits viz., 100-seed weight (HSW), seed area (SA), seed length (SL) and seed width (SW) GWAS populations of 350 accessions across two environments (CC22 & JT23) plus combined environment (CE). The Manhattan plot of the GWAS for seed size across in 350 accessions. A-D Manhattan plot for four seed size-related traits viz., HSW, SA, SL and SW by MLM model in the CC22 environment; (E–H) Manhattan plot for four seed size-related traits viz., HSW, SA, SL and SW by MLM model in the JT23 environment; and (I-L) Manhattan plot for four seed size-related traits viz., HSW, SA, SL and SW by MLM model in the CE. The green line represents the threshold level of significance (-log₁₀P > 6.67), and the red line represents the threshold of -log10 (p value) after correction for the passing false-discovery rate (FDR), and the soybean chromosomes are represented by the numbers on the X-axis

Fig. 6

qRT-PCR showed the differential expression pattern of candidate genes among the two contrasting parental lines Dongnong60 (smaller seed) and Heihe50 (larger seed) at two developmental stages of 15DAF and 30DAF (DAF means days after flowering). A Relative expression levels of Glyma.18G242400; (B) Relative expression levels of Glyma.18G243100; (C) Relative expression levels of Glyma.18G243300. GmActin11 (Glyma.18G290800) and GmCYP2 (Glyma.12G024700) was used as two different internal controls. *indicates significant difference at p < 0.05, **indicates significant difference at p < 0.01, *** indicates significant difference at p < 0.001, respectively; ns, not significant (Student’s t-test, two-tail)

Fig. 7

Haplotype allele analysis underlying three haplotype blocks. viz., Hap2, Hap6A and Hap6B identified for SW on Chr.02, Chr.06 and Chr.06, respectively. Grouping of genotypes and pairwise comparisons of genotypes was performed by using Tukey’s HSD test at P < 0.05. Common letters above the boxes represent the non-significant differences in SW, whereas different letters represent significant differences

Fig. 8

Haplotype and phenotypic analysis of the Glyma.18G242400 gene in the GWAS panel of accessions. A Location and information of SNPs in haplotype as well as the number of haplotype alleles of the Glyma.18G242400 identified in the GWAS panel; (B) Phenotypic effect of the different haplotype alleles of the Glyma.18G242400 on the different seed-size related traits

Fig. 9

The evolution and geographical distribution of Glyma.18G242400 different haplotypes. A The pie charts represent the percentage of accessions with different haplotypes in wild soybean, landrace, and improved cultivars; (B) The global geographical distribution of 4414 soybean accessions (including wild soybean, landrace and improved cultivars). The color represents the different haplotypes of germplasm