Genomic-assisted haplotype analysis and the development of high-throughput SNP markers for salinity tolerance in soybean

Abstract

Soil salinity is a limiting factor of crop yield. The soybean is sensitive to soil salinity, and a dominant gene, Glyma03g32900 is primarily responsible for salt-tolerance. The identification of high throughput and robust markers as well as the deployment of salt-tolerant cultivars are effective approaches to minimize yield loss under saline conditions. We utilized high quality (15x) whole-genome resequencing (WGRS) on 106 diverse soybean lines and identified three major structural variants and allelic variation in the promoter and genic regions of the GmCHX1 gene. The discovery of single nucleotide polymorphisms (SNPs) associated with structural variants facilitated the design of six KASPar assays. Additionally, haplotype analysis and pedigree tracking of 93 U.S. ancestral lines were performed using publically available WGRS datasets. Identified SNP markers were validated, and a strong correlation was observed between the genotype and salt treatment phenotype (leaf scorch, chlorophyll content and Na(+) accumulation) using a panel of 104 soybean lines and, an interspecific bi-parental population (F8) from PI483463 x Hutcheson. These markers precisely identified salt-tolerant/sensitive genotypes (>91%), and different structural-variants (>98%). These SNP assays, supported by accurate phenotyping, haplotype analyses and pedigree tracking information, will accelerate marker-assisted selection programs to enhance the development of salt-tolerant soybean cultivars.

Figure 1. The phenotypic evaluation for salinity tolerance.

(A) Effect of salt treatment (120 mM NaCl) on four diverse genotypes after 1 week of treatment under greenhouse conditions. Fiskeby III (HN105, salt-tolerant control), Hutcheson (HN001, salt-sensitive control) along with PI 561271 (HN074) and PI 548657 (HN071) showed a high level of tolerance and sensitivity out of 104 germplasm lines tested, respectively. (B) Correlation coefficients of leaf scorch score (LSS) and SPAD ratio were calculated from 104 soybean genotypes evaluated for salt tolerance. The highlighted circle shows the most tolerant and sensitive genotypes. (C) Dendrogram showing phenotypic (LSS and SPAD ratio) variability relationship between 104 soybean accessions based on phenotypic data. The Euclidean distance (horizontal axis) between objects is used as the distance measure; the clustering was performed using NTSYS software.

Figure 2

Manhattan plots of GWAS for (A) LSS and, (B) SPAD ratio, in 106 soybean lines using WGRS dataset. Negative log₁₀-transformed P values of SNPs from genome-wide scan for salinity traits using EMMAX model including kinship and population structure are plotted against positions on each of the 20 chromosomes; (C) genes underlying significant trait-associated SNPs on Chr. 3. The horizontal line denotes the calculated threshold value for declaring significant association.

Figure 3. Hierarchical clustering of 129 diverse lines based on SNP information of WGRS data.

Clusters were observed in 23 lines [denoted with ‘C’ and ‘W’] and 106 lines [denoted with ‘HN’]. Lines representing structural variation are highlighted with different colors (see details in Fig. 4A). Lines with known salinity reaction are shown with pointed arrow in each group.

Figure 4

(A) Structural variation in the GmCHX1 gene. Blue box represents exon, green bar represents intron, brown box represents insertion of Ty1/Copia transposon in exon 3, and gray bar represents 3′ and 5′ UTR. Dotted lines indicate the exon position. Red lines indicate that this region is deleted in SV-3 (Valliyodan et al. unpublished). (B) Schematic graph shows the position of SNP/Indels at for the GmCHX1 (40621077-40634451) gene. For clear visualization, 10 genotypes from each SV group were selected (for all other lines see Supplementary Table S3). SNP in back background are different from the reference genome (W82). The asterisk (*) above gene structure represents the approximate position of the SNP. SNPs used for the KASPar assay is denoted by black asterisk. The number above the SNP matrix shows position from start codon. PI line with underline represents wild soybean genotypes (G. soja).

Figure 5. Sample genotyping plots of a diverse set of 106 WGRS lines and F₈ RILs from population (PI 483463 x Hutcheson).

Plots generated from Roche 480 II software during KASPar assay genotyping of M1-M6 SNP markers. A-E: KASPar SNP graphs of 106 diverse soybean lines. F: KASPar SNP graph of RIL population. Genotype signal: Green- Mutant, Red- Heterozygote, Blue- WT (W82), Grey- Non template control. Heterozygote signal highlighted by circle represents artificial heterozygote.

Figure 6. Partial pedigree tracing of salt-tolerant and salt-sensitive lines.

Green represents salt-tolerant, light-red represents salt-sensitive, grey unknown. Asterisk (*) denotes that genotype of these lines studied in the present investigation using WGRS information.