Genome-wide association study reveals the QTLs and candidate genes associated with seed longevity in soybean (Glycine max (L.) Merrill)

Abstract

Background: Soybean is challenged with a problem of poor seed longevity, a complex trait and key target for breeding. Therefore, understanding the genetic basis of seed longevity is of great significance for mining favorable genes and prolonging the seed life. A genome-wide association study (GWAS) was conducted to understand the genetic background of seed longevity over two ageing methods (natural ageing and accelerated ageing) and seasons.

Results: This study evaluated seed longevity traits in a panel of sixty diverse soybean genotypes with different seed coat colors, seed sizes, and agro-morphological traits under natural and accelerated ageing across two seasons. Seed longevity related traits such as seed germination, seedling vigor index-I (SVI-I), seedling vigor index-II (RVI-II), reduction in vigor index-I (RVI-I), and reduction in vigor index-II (RVI-II) were recorded after 12 and 14 months after natural ageing and accelerated ageing during the year 2019 and 2021. Seed germination under natural aging averaged 45.2% (range: 4.8-100%) after 12 months and 38.1% (range: 0.0-90%) after 14 months. Accelerated ageing resulted in germination rates ranging from 0.0 to 80.6% with means of 43.38% and 40.7% in 2019 and 2021, respectively. Significant variability in SVI-I, SVI-II, RVI-I, and RVI-II was observed across genotypes and conditions. GWAS using 29,955 Genotyping-by-sequencing (GBS)-single nucleotide polymorphism (SNP) markers identified 71 significant SNPs linked to seed longevity traits. Chromosomes 1, 4, and 8 harboured common SNPs for seed germination, SVI-I, SVI-II, and RVI-I. QTL hotspots were detected on chromosomes 2 and 8, encompassing multiple SNPs within 189 bp and 19 bp, respectively. Twenty-eight candidate genes were identified, including Glyma20g27840 (encoding a LEA hydroxyproline-rich glycoprotein family) on chromosome 20 for germination and SVI-I, and Glyma08g24630 (encoding ATP-dependent RNA helicase A) on chromosome 8 for germination, SVI-II, and RVI-II. Several candidate genes involved in ROS scavenging, cell regulation, ATP production, metabolism, stress response, and seed development were also associated with seed longevity traits.

Conclusions: The SNPs associated with many longevity related traits and genes in the present study can be used for soybean breeding and functional studies of seed longevity in soybean after validation using linkage mapping or diverse population.

Keywords: Candidate gene; Genome wide association study; Natural ageing of seeds; QTLs; Seed germination; Seed longevity; Seedling Vigor.

Fig. 1

Box plot of seed longevity traits, (a) germination %, (b) seedling vigor index-I, (c) seedling vigor index-II, (d) reduction in vigor index-I, and (e) reduction in vigor index-II under two season accelerated ageing (AA-2019; AA-2021) and natural ageing after 12 months (NA-12) and 14 months (NA-14). The genotypes listed in upper and lower range of each boxplot indicate the genotypes with maximum and minimum values, respectively. The statistical analysis was performed using completely randomized design. The F-value provided in the figures represent the ratio of variance between genotypes to the variance within the genotypes. The p-value represents the statistical significance

Fig. 2

The principal component analysis of soybean genotypes using the seed longevity traits measured after accelerated and natural ageing. a) Scree plot of principal components with percent variance; b) Scatter plot of genotypes under PC1 and PC2. G%: seed germination, SVI-I: seedling vigor index-I, SVI-II: seedling vigor index-II, RVI-I: reduction in vigor index-I, RVI-II: reduction in vigor index-II. The 19, 21 and indicated the Accelerated ageing during 2019 and 2021. The 12 and 14 indicate the natural ageing after 12 and 14 months

Fig. 3

Bayesian-based population structure analysis. a) scree plot delta K and b) the bar plot of population structure with two sub-populations. The list of genotypes number with their names are given in Supplementary Table 1

Fig. 4

Manhattan and Q-Q plot of seed longevity traits (seedling vigor index-I&II, and reduction in vigor index-I&II) after accelerated ageing during 2019 (AA-19) and 2021 (AA-21). The red line indicate the p-value threshold above which the SNPs will significantly associated with the trait

Fig. 5

Manhattan and Q-Q plot of seed longevity traits (seed germination, seedling vigor index-I&II, and reduction in vigor index-I&II) after natural ageing for 12 months (NA-12) and 14 months (NA-14). The red line indicate the threshold p-value after p-value threshold above which the SNPs will significantly associated with the trait. The green line indicate the threshold p-value after standard p-value threshold. The dotted green line is threshold p-value after FDR correction

Fig. 6

Linkage disequilibrium decay plot using 29,955 SNP markers in soybean