Skip to main content
NCBI home page
Springer logoLink to Springer
. 2025 Feb 3;138(2):43. doi: 10.1007/s00122-024-04808-5

QTL mapping and whole-genome sequencing analysis for novel genetic resources associated with sucrose content in soybean [Glycine max (L.) Merr.]

Dongho Lee 1,2,3, Tri D Vuong 2, James G Shannon 1, Qijian Song 3, Feng Lin 1,✉, Henry T Nguyen 2
PMCID: PMC11788231  PMID: 39894897

Abstract

Key message

A major QTL for sucrose content was mapped on chromosome 8 in PI 506593. The novel genetic variants and candidate genes were further identified within the major QTL.

Abstract

Sucrose in soybean [Glycine max (L.) Merr.] contribute to animal feed efficiency and natural sweetness of soy products. Thus, identifying novel genetic resources, such as quantitative trait loci (QTL), associated with sucrose content in soybean is essential for enhancing seed values. In this study, two recombinant inbred line populations derived from the same high sucrose donor parent, PI 506593, were used to identify significant QTLs. A total of 11 sucrose-related regions on chromosomes (Chrs.) 4, 5, 6, 8, 10, and 13 were identified using QTL analysis. Among them, four QTLs (qSUC_08.1, qSUC_08.2, qSUC_08.3, and qSUC_08.4) were clustered in the interval of 40,597,410–42,861,364 bp on Chr. 8, which was considered major QTL region. A desirable marker at 41,834,095 bp was tested in two populations, showing that two phenotypically extreme groups were efficiently differentiated. We further identified 44 and 54 candidate genes with non-synonymous mutations in the major QTL region based on the annotations of Wm82.a2.v1 and Wm82.a5.v1 assemblies, respectively. Among 54 candidate genes from Wm82.a5.v1, Protein Variation Effect Analyzer (PROVEAN) revealed that 18 genes contained 34 variants that had deleterious impacts on biological functions. RNA-seq analysis highlighted five candidate genes that were highly expressed in pod and seed tissues during reproductive stages and other plant parts. A gene, Gm_Wm82_23219 (Glyma.08G293800, Wm82.a2.v1) encoding proline-rich protein 4-like, was highlighted in both PROVEAN and RNA-seq analyses. Novel findings in this study will be valuable genetic resources in soybean breeding programs that aim to improve efficiency in animal feed and human food.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00122-024-04808-5.

Introduction

Soybean [Glycine max (L.) Merr.] is the leading protein meal source, accounting for 69.7% (247 million metric tons) of global protein meal consumption, followed by rapeseed (Brassica napus L.) (13.1%, 47 million metric tons) in 2023 (http://soystats.com). Soybean meal is primarily used to feed non-ruminant animals, such as poultry and swine, as a main source of high-quality protein in feed formulations (Pope et al. 2023). Soybean meal meets three fundamental requirements in animal feeding programs, which are (1) Having one or more essential nutrients, (2) Being available to meet the demand of regular usage on a large scale, and (3) Being cost-effective to use (Dozier and Hess 2011). Besides protein, soluble carbohydrates are also essential components in animal feed for improving metabolizable energy (ME) efficiency. Soybean seeds generally contain up to 15% soluble carbohydrates on a dry-weight basis (Feng et al. 2005). The three main components consisting of soluble carbohydrates are sucrose, raffinose, and stachyose, of which sucrose typically accounts for the largest portion of soluble carbohydrates in cultivated soybean seeds (Sui et al. 2020; Wang et al. 2023a).

Among soluble carbohydrates in soybean, sucrose is the only nutritionally beneficial component, which is readily digestible by non-ruminant animals and turns into the source of ME (Jo et al. 2018; John 2008; Parsons et al. 2000; Wang et al. 2023a). Animal feeding programs recommend feed formulations with high sucrose content because sucrose has a significantly higher ME value (3900 kcal/kg) than starch (2918–3396 kcal/kg) (John 2008; Ostezan et al. 2023). Additionally, as the global vegetarian population has increased in recent decades, soy products have garnered popularity as a substitute for meat protein in vegetarian diets (Simeone et al. 2022; Wang et al. 2023b). Sucrose is a pivotal contributor to natural sweetness in soy food products, such as edamame, miso, natto, tofu, and soymilk (Rosset et al. 2012; Sui et al. 2020; Wang et al. 2023a). Therefore, the development of high sucrose soybean varieties has become a goal of breeding programs to improve animal feed efficiency and natural sweetness to meet consumer preference for soy products.

Quantitative trait loci (QTL) mapping strategies have been mainly used to identify genomic regions governing complex quantitative traits (Rani et al. 2023). For sucrose content in soybean seed, Maughan et al. (2000) first reported 17 sucrose-related QTLs on chromosomes (Chrs.) 5, 7, 8, 13, 15, 19, and 20 using an interspecific cross between a large-seeded breeding line and a wild soybean (G. soja) line. Kim et al. (2005, 2006) identified 17 sucrose-related QTLs on Chrs. 2, 11, 12, 14, 15, 16, and 19 using two F2:10 recombinant inbred line (RIL) populations sharing the same high sucrose donor line, Keunolkong. A major QTL for sucrose content was identified on Chr. 11 in V99-5089, derived from a cross between V71-370 and PI 87013 (Saghai Maroof and Buss 2008). Skoneczka et al. (2009) used two F2:3 populations derived from PI 200508 to identify another major QTL on Chr. 6 for high sucrose content. Later, three sucrose-related QTLs on Chrs. 7, 11, and 20 were identified by Wang et al. (2014), of which the QTL on Chr. 11 was the same QTL that Saghai Maroof and Buss (2008) previously identified. Zeng et al. (2014) identified three sucrose-related QTLs on Chrs. 5, 9, and 16. There have been more QTL mapping studies conducted for sucrose content in soybean using different donor parents and larger marker data sets (Akond et al. 2015; Knizia et al. 2023; Lee et al. 2016; Liu et al. 2023; Patil et al. 2018; Salari et al. 2021).

However, only a few variants and genes have been identified to be associated with high sucrose content. A three-base pair deletion leading to an amino acid deletion (rs2W331-) and a missense mutation (rs2T107I) in raffinose synthase 2 (RS2, Glyma.06G179200, Wm82.a2.v1) coding regions were responsible for increased sucrose and reduced raffinose and stachyose content (Dierking and Bilyeu 2008, 2009; Kerr and Sebastian 2000; Skoneczka et al. 2009). Although mutations in RS3 (Glyma.05G003900, Wm82.a2.v1) (rs3snp5/rs3snp6) alone did not increase the sucrose content, the combination of the RS2 and RS3 mutations significantly increases sucrose content (Hagely et al. 2020; Jo et al. 2018; Lee et al. 2023a; Thapa et al. 2019). A natural mutation in the D-myo-inositol 3-phosphate synthase 1 (MIPS1) gene (Glyma.11G238800, Wm82.a2.v1) derived from V99-5089 was associated with elevated sucrose content and significantly reduced raffinose and stachyose content in soybean (Rosso et al. 2011; Saghai Maroof and Buss 2008).

Recently, a new high-quality reference genome of Williams 82 (Wm82.a5.v1) was assembled using the combination of PacBio HiFi sequencing and Bionano optical mapping (Garg et al. 2023). The Wm82.a5.v1 conveys much smaller gaps and scaffold breaks than previous versions, such as Wm82.a1.v1.1 (Schmutz et al. 2010), Wm82.a2.v1 (Song et al. 2016), and Wm82.a4.v1 (Valliyodan et al. 2019). The previous reference genome assemblies contained 56,691 gaps (42.29 Mbp), whereas Wm82.a5.v1 is near gapless (only 14.25 Kbp gaps across the genome), of which ten chromosomes were completely reconstructed without any gaps (Garg et al. 2023). The high-quality reference genome recently developed will provide a robust and accurate understanding of the genetic architecture underlying complex quantitative traits of interest.

In this study, a high sucrose soybean germplasm, PI 506593, was crossed with two low sucrose breeding lines, S16-14161 and S16-11651, to develop two RIL mapping populations. The populations were genotyped with BARCSoySNP6K Illumina Infinium BeadChips (Song et al. 2020) and phenotypically evaluated in three field environments to identify sucrose-related QTLs. Furthermore, we compared the whole-genome sequence (WGS) of the three parental lines after alignment with the most widely used soybean reference genome assembly (Wm82.a2.v1) and the new soybean reference genome assembly (Wm82.a5.v1) to identify candidate genes with novel SNP and Indel variants within the major QTL in PI 506593. Protein Variation Effect Analyzer (PROVEAN) and RNA-seq analysis were further used to investigate the biological function of candidate genes, the impact of novel variants, and the gene expression levels of candidate genes across tissues and developmental stages. This study provides valuable information on QTL and candidate genes controlling sucrose content in soybean.

Materials and methods

Plant materials

PI 506593 was selected from the genome-wide association study panel used by Lee et al. (2023b) based on the sucrose content and stability across multiple environments. PI 506593 is a high sucrose large-seeded (32 g 100 seeds−1) maturity group VI soybean germplasm originating from Japan (https://www.ars-grin.gov/). Two low sucrose breeding lines, S16-14161 and S16-11651 (Chen et al. 2022), developed by the University of Missouri—Fisher Delta Research, Extension, and Education Center (FDREEC), were crossed with PI 506593.

The crosses of S16-14161 × PI 506593 and S16-11651 × PI 506593 were made in the summer of 2019 at the FDREEC in Portageville, MO, USA. The F1 and F2 generations were grown in the East Campus Plant Growth Facility at the University of Missouri, Columbia, MO, USA, and F3 seeds were harvested in October 2020. The F3 and F4 generations were grown in the soybean winter nursery in Costa Rica until mid-June 2021. The modified single-pod descent method was used throughout the generation advancement process (Fehr 1987). A total of 140 and 207 F4:5 RILs constituted two mapping populations, named P593_RIL1 (S16-14161 × PI 506593) and P593_RIL2 (S16-11651 × PI 506593), respectively, and were subsequently used for QTL analysis.

Field experiments

The two RIL populations and the parental lines were planted in three field environments. It included the FDREEC, Portageville, MO, USA (36.42°N 89.70°W) in 2021 and 2022 (named FDREEC_21 and FDREEC_22, respectively) and the Bradford Research and Extension Center (BREC), Columbia, MO, USA (38.89°N 92.19°W) in 2022 (named BREC_22). Ten seeds of each line were planted in 75 cm wide rows in hill plots spaced 30 cm apart at the FDREEC and 60 cm apart at the BREC in a randomized complete block design (RCBD) with two replications. Seeds were harvested at maturity for further sucrose quantification.

Phenotyping for sucrose content

The F5:6 and F6:7 RIL populations harvested in 2021 and 2022, respectively, were phenotyped for sucrose content. Ten intact soybean seeds (fully matured, spotless, and no damage) were sampled from each plot and used to quantify sucrose content using the established High-Performance Liquid Chromatography (HPLC) protocol described by Valliyodan et al. (2015) with minor modifications. Briefly, a powder sample was obtained by grinding seed samples with Thomas Wiley Mini-Mill (Arthur Thomas Co., Chadds Ford, PA, USA) fitted with a 20-mesh screen. The samples were lyophilized for two days using a Labconco freeze-dry system (Labconco, Kansas City, MO, USA). HPLC-grade water of 900 μL was added to 90.25 (± 0.15) mg of each lyophilized sample in a 2 mL centrifuge tube. After incubating tubes at 55 °C with 200 rpm agitation for an hour, the tubes were vortexed for 30 s. After 20 min under room temperature, 900 μL HPLC-grade acetonitrile was added to each tube. Then, the suspension was centrifuged for 30 min at a 14.0 × 1000 min−1 × g speed. The supernatant was diluted five times using 65% HPLC-grade acetonitrile to prepare the final samples. The Agilent HPLC-ELSD (Evaporative Light Scattering Detection) 120 series (Agilent, Santa Clara, CA, USA), equipped with the Prevail Carbohydrate ES columns (5 μm 250 × 4.6 mm) and guard columns (7.5 × 4.6 mm) (Grace Davison Discovery Sciences, Deerfield, IL, USA) were used to quantify sucrose content. The calibration curves were generated using the standard mixtures prepared in HPLC-grade water with 50, 100, 300, 500, and 1000 μg/mL concentrations.

Genotypic data

The genomic DNA of the F5:6 RIL populations and parental lines were isolated from the youngest trifoliate leaves using an established cetyltrimethylammonium bromide (CTAB) protocol with minor modifications as previously described (Vuong et al. 2010). Marker genotyping was performed in the Soybean Genomics and Improvement Laboratory, USDA-ARS, Beltsville, MD, using the BARCSoySNP6K Illumina Infinium BeadChips (Illumina Inc., San Diego, CA, USA) (Song et al. 2020). The SNP alleles were called using the Genotyping Module 2.0 in GenomeStudio software (https://www.illumina.com/). The physical positions of SNP markers were aligned to Wm82.a2.v1. The WGS data of parental lines were generated using the Illumina NovaSeq PE150 platform at a depth of 30 × by Novogene Corporation Inc. (Sacramento, CA, USA). Sequencing libraries with an insert size of ~ 350 bp were constructed incorporating the standard protocol.

Variant identification

The WGS data of the three parental lines, S16-14161, S16-11651, and PI 506593, were further used for variant identification. The raw Illumina reads were trimmed using Trimmomatic version 0.39 (Bolger et al. 2014). The clean data were aligned to two reference genome assemblies, Wm82.a2.v1 and Wm82.a5.v1, respectively, using the BWA-MEM (Li and Durbin 2009). The genome assemblies and annotation files of Wm82.a2.v1 and Wm82.a5.v1 were obtained from Soybase (https://www.soybase.org/) and Garg et al. (2023), respectively. The entire variant calling processing followed the industry-standard pipeline in GATK (available at https://gatk.broadinstitute.org/hc/en-us). The variants were annotated using the ANNOVAR tool (Wang et al. 2010). The physical positions of QTL regions and genes on Wm82.a5.v1 were retrieved from a liftover file with corresponding gene IDs and physical positions between Wm82.a2.v1 and Wm82.a5.v1. The genes containing unique variants in a donor parent were further evaluated.

Genetic map construction and QTL analysis

A genetic linkage map was constructed with a 6K SNP dataset for each population using JoinMap version 5.0 software (Van Ooijen 2018). A logarithm of the odds (LOD) score from 2 to 15 was used to cluster linkage groups. Regression mapping was set for the mapping algorithm, and map distances were calculated using Kosambi’s mapping function. The QTL analysis was carried out using MapQTL version 6.0 software based on a genetic linkage map, SNP alleles, and phenotypic data (Van Ooijen 2009). Successive procedures, including interval mapping, automatic cofactor selection, and multi-QTL method, were performed for QTL analysis. The genome-wide LOD threshold was determined using a 1000-permutation test at the 0.05 probability level of significance, resulting in the LOD threshold of 3.2 being selected for both populations in this study. The significant QTL region was defined by two flanking markers encompassing the marker with a LOD score higher than 3.2. The QTL identified in this study were named according to the following rule: qSUC_chromosome.QTL number (e.g., qSUC_08.1, qSUC_08.2, etc.).

Candidate gene identification

First, all genes within the QTL region on Wm82.a2.v1 and Wm82.a5.v1 were respectively retrieved. Afterward, candidate genes with non-synonymous SNP and Indel variants present only in PI 506593 were collected based on the variant identification results on Wm82.a2.v1 and Wm82.a5.v1. The functional annotations of those candidate genes were studied based on Wm82.a5.v1 using Soybase (https://www.soybase.org/). For unannotated candidate genes, coding sequences were used to perform BLAST in NCBI (https://blast.ncbi.nlm.nih.gov/Blast.cgi) based on soybean (Glycine max), wild soybean (Glycine soja), common bean (Phaseolus vulgaris), and Arabidopsis (Arabidopsis thaliana) genome databases. The biological impacts of amino acid changes derived from non-synonymous SNP and non-frameshift Indel variants were predicted by PROVEAN version 1.1.5 software (Choi and Chan 2015). The amino acid sequences used for PROVEAN were based on Wm82.a5.v1, and the significant threshold of ≤ − 2.5 was used, as suggested by Choi and Chan (2015). RNA-seq data analysis was conducted based on two gene expression data sets publicly available from Severin et al. (2010) and Almeida-Silva et al. (2023), respectively, as described in Lee et al. (2023b). The PROVEAN and RNA-seq analysis were conducted for the candidate genes identified in Wm82.a5.v1.

Statistical analysis

The broad-sense heritability (H2) was calculated using the H2cal function in the inti R package (https://rdrr.io/cran/inti/src/R/H2cal.R). Analysis of variance (ANOVA) was conducted using the PROC MIXED procedure of SAS software version 9.4 (SAS Institute, Cary, NC, USA). Genotype was used as a fixed effect, and Environment, Genotype × Environment interaction, and replication were used as random effects.

Results

Phenotypic variations of sucrose content in two RIL populations

The overall description of the sucrose content of two RIL populations (P593_RIL1 and P593_RIL2) across three environments is shown in Table 1. In P593_RIL1, there was transgressive segregation for sucrose content with a normal distribution (Shapiro–Wilk value from 0.988 to 0.992) in all environments (Table 1). PI 506593 consistently showed higher sucrose content (8.2%) than S16-14161 (5.5%) across environments. P593_RIL1 had the highest average sucrose content in BREC_22 (8.3%) and the lowest in FDREEC_21 (6.9%). The coefficient of variation (CV) ranged from 10.4% to 12.9%, and the broad-sense heritability was 0.83. Similarly, the sucrose content of P593_RIL2 showed normal distribution (Shapiro–Wilk value from 0.975 to 0.996) and transgressive segregation in all environments (Table 1). PI 506593 consistently showed higher sucrose content (8.2%) than S16-11651 (6.5%) in P593_RIL2 across environments. The sucrose content of P593_RIL2 also varied across environments, following a similar trend as P593_RIL1, with the highest average sucrose content observed in BREC_22 (7.8%) and the lowest in FDREEC_21 (6.6%). The CV ranged from 10.0 to 12.6%, and the broad-sense heritability was 0.84. The ANOVA showed that except for the genotype × environment interaction in P593_RIL1, all other factors had significant effects (Suppl. Table S1).

Table 1.

The overall description of sucrose content in parents (P1 and P2) and two mapping populations across three environments

Populationa Envb P1 P2 Mean ± SDc Range Shapiro wilk (w) Skewness Kurtosis Variance CV (%)d Heritability (H2)e
P593_RIL1 FDREEC_21 4.9 7.8 6.9 ± 0.9 4.5–9.4 0.991 − 0.01 3.02 0.89 12.9 0.83
FDREEC_22 5.4 8.0 7.2 ± 0.9 5.3–10.1 0.988 0.30 3.34 0.75 10.4
BREC_22 6.3 8.8 8.3 ± 1.0 5.8–10.8 0.992 0.15 3.00 0.93 11.2
Mean 5.5 8.2 7.4 ± 0.8 5.1–9.4 –
P593_RIL2 FDREEC_21 5.9 7.8 6.6 ± 0.9 4.4–9.1 0.990 0.07 2.61 0.83 12.6 0.84
FDREEC_22 6.5 8.0 6.9 ± 0.9 3.8–9.5 0.996 − 0.05 3.24 0.79 11.4
BREC_22 7.0 8.8 7.8 ± 0.9 6.1–10.8 0.975 0.54 3.10 0.78 10.0
Mean 6.5 8.2 7.1 ± 0.8 5.2–9.0 –
Open in a new tab

a P593_RIL1, a bi-parental mapping population from a cross between S16-14161 (P1) and PI 506593 (P2); P593_RIL2, a bi-parental mapping population from a cross between S16-11651 (P1) and PI 506593 (P2)

b FDREEC_21, Fisher Delta Research, Extension, and Education Center in 2021; FDREEC_22, Fisher Delta Research, Extension, and Education Center in 2022; BREC_22, Bradford Research Education Center in 2022

c SD, standard deviation

d CV, coefficient of variation

e H2, broad sense of heritability

Genetic linkage map for P593_RIL1 and P593_RIL2

Among 6,000 markers from the BARCSoySNP6K BeadChips, 1440 (24%) and 600 (10%) markers were polymorphic between the two parents of the P593_RIL1 (S16-14161 × PI 506593) and P593_RIL2 (S16-11651 × PI 506593), respectively (Table 2). The polymorphic markers were anchored on 20 chromosomes of the integrated genetic map with a total genetic distance of 2422.4 cM in P593_RIL1 and 2695.2 cM in P593_RIL2. In P593_RIL1, the total distance of each chromosome ranged from 74.9 cM (Chr. 7) to 187.5 cM (Chr. 13), and the genetic interval between two neighboring markers averaged from 1.3 to 2.6 cM, with an average of 1.7 cM. In P593_RIL2, the total distance of each chromosome ranged from 54.2 cM (Chr. 7) to 217.1 cM (Chr. 2), and the average genetic interval between two neighboring markers varied from 2.0 to 15.5 cM, with an average of 5.2 cM.

Table 2.

Distribution of SNP markers across 20 chromosomes

Chra P593_RIL1b P593_RIL2c
# Markers Total distance (cM) Average interval (cM) # Markers Total distance (cM) Average interval (cM)
1 48 114.8 2.4 21 176.1 8.4
2 99 163.2 1.6 14 217.1 15.5
3 74 113.1 1.5 35 81.7 2.3
4 84 131.3 1.6 22 159.4 7.2
5 56 123.3 2.2 18 120.2 6.7
6 75 159.7 2.1 23 115.4 5.0
7 57 74.9 1.3 27 54.2 2.0
8 104 141.4 1.4 24 118.5 4.9
9 79 117.6 1.5 31 187.2 6.0
10 80 143.2 1.8 28 127.4 4.6
11 57 146.2 2.6 23 93.0 4.0
12 67 97.8 1.5 18 124.1 6.9
13 82 187.5 2.3 61 176.4 2.9
14 62 82.0 1.3 51 133.1 2.6
15 62 107.3 1.7 16 81.0 5.1
16 69 101.2 1.5 39 189.2 4.9
17 83 121.3 1.5 44 102.6 2.3
18 59 75.8 1.3 37 130.0 3.5
19 74 115.6 1.6 29 125.5 4.3
20 69 105.2 1.5 39 183.1 4.7
Total 1440 2422.4 1.7 600 2695.2 5.2
Open in a new tab

a Chr, chromosome

b P593_RIL1, a bi-parental mapping population from a cross between S16-14161 and PI 506593

c P593_RIL2, a bi-parental mapping population from a cross between S16-11651 and PI 506593

Identification of QTLs associated with sucrose content

The QTL analysis of the two mapping populations identified 11 sucrose-related QTLs on Chrs. 4, 5, 6, 8, 10, and 13 (Table 3). In P593_RIL1, six QTLs were identified on Chrs. 6 (qSUC_06.1 and qSUC_06.2), 8 (qSUC_08.1 and qSUC_08.2), 10 (qSUC_10.1), and 13 (qSUC_13.1) and explained 9.1–14.9% of the phenotypic variation (Table 3). Among them, qSUC_08.1 and qSUC_06.1 had the highest LOD values of 5.9 and 5.3 in the FDREEC_21 and FDREEC_22, respectively. Only one QTL (qSUC_06.2) was detected with a LOD value of 4.1 in BREC_22. The additive effects of each QTL ranged from − 0.4 to -0.3, indicating that high sucrose-related alleles were inherited from the donor parent, PI 506593.

Table 3.

Summary of QTL associated with sucrose content in two mapping populations across three environments

Populationa Environment QTLb Chrc Marker interval (cM) Physical interval (bp) LODd PVE (%)e Addf
P593_RIL1 FDREEC_21 qSUC_08.1 8 112.9–114.8 42,146,252–42,532,915 5.9 14.0 − 0.4
FDREEC_21 qSUC_10.1 10 49.9–51.5 37,502,435–37,995,354 4.5 10.4 − 0.3
FDREEC_21 qSUC_13.1 13 71.5–76.7 25,268,844–25,927,261 4.0 9.3 − 0.3
FDREEC_22 qSUC_06.1 6 115.2–124.0 16,133,204–21,786,787 5.3 14.9 − 0.3
FDREEC_22 qSUC_08.2 8 114.8–115.9 42,532,915–42,861,364 3.3 9.1 − 0.3
BREC_22 qSUC_06.2 6 124.0–131.4 21,786,787–46,805,251 4.1 13.9 − 0.4
P593_RIL2 FDREEC_21 qSUC_05.1 5 77.5–86.1 34,272,637–35,101,291 5.2 7.8 − 0.3
FDREEC_21 qSUC_08.3 8 100.8–103.2 40,940,261–41,834,095 12.1 19.7 − 0.4
FDREEC_21 qSUC_13.2 13 116.9–117.1 32,525,645–32,570,888 3.7 5.5 − 0.2
FDREEC_22 qSUC_08.4 8 99.7–100.8 40,597,410–40,940,261 5.3 11.3 − 0.3
BREC_22 qSUC_04.1 4 106.0–125.4 46,793,849–48,002,103 3.4 6.3 − 0.2
BREC_22 qSUC_08.4 8 99.7–100.8 40,597,410–40,940,261 3.8 7.1 − 0.3
Open in a new tab

a P593_RIL1, a bi-parental mapping population from a cross between S16-14161 and PI 506593; P593_RIL2, a bi-parental mapping population from a cross between S16-11651 and PI 506593

b QTL, quantitative trait loci

c Chr, chromosome

d LOD, logarithm of the odds

e PVE, percentage variance explained

f Add, additive effect

In P593_RIL2, five QTLs were identified on Chrs. 4 (qSUC_04.1), 5 (qSUC_05.1), 8 (qSUC_08.3 and qSUC_08.4), and 13 (qSUC_13.2) and explained 5.5–19.7% of the phenotypic variation (Table 3). Among them, qSUC_08.3 and qSUC_08.4 showed the highest LOD values of 12.1 and 3.8 in FDREEC_21 and BREC_22, respectively, while qSUC_08.4 was the only QTL detected in FDREEC_22 with a LOD value of 5.3. Similarly, in P593_RIL1, the additive effects of each QTL varied from − 0.4 to − 0.2, indicating that PI 506593 positively contributed to an increase in sucrose content. The QTL analysis of two populations consistently pinpointed a QTL region on Chr. 8 harboring four QTLs (qSUC_08.1, qSUC_08.2, qSUC_08.3, and qSUC_08.4) at the physical interval of 40,597,410–42,861,364 bp across all environments (except BREC_22 in the P593_RIL1) (Table 3). This region should be considered the location of a major QTL conferring high sucrose content in PI 506593 for further investigation.

Sucrose content of selected RILs in each population based on the desirable SNP marker

Within the major QTL region on Chr. 8, the most desirable SNP marker was selected based on the correlation with the sucrose content of RILs in each population. The SNP marker at 41,834,095 bp on Chr. 8 was selected for both populations. Fifteen RILs from each phenotypically extreme group (Top vs. Bottom) were selected from both populations, and their sucrose content and alleles of the SNP marker were described in Table 4. In P593_RIL1, the top 15 RILs showed average sucrose content ranging from 8.8 to 10.2%, and 12 RILs carried an allele derived from the donor parent. On the other hand, the bottom 15 RILs showed average sucrose content from 5.2 to 6.3%, and 13 RILs carried an allele derived from the low sucrose parental line. In P593_RIL2, the top 15 RILs showed average sucrose content ranging from 8.2 to 9.0%, but only six RILs carried an allele derived from the donor parent. However, all the bottom 15 RILs carried an allele derived from the low sucrose parent, with average sucrose content ranging from 5.2 to 6.0%. Overall, the desirable SNP marker demonstrated high efficiency in differentiating between two extreme groups, while it was only efficient in discriminating the low sucrose group in P593_RIL2.

Table 4.

Sucrose content of the selected lines from extreme groups based on the desirable SNP marker within the P593_RIL1 and P593_RIL2, respectively

Populationa Top 15 RILs Bottom 15 RILs
RIL ID Alleleb Sucrose (%)c RIL ID Allele Sucrose (%)
P593_RIL1 P593_RIL1_96 B 10.2 P593_RIL1_114 B 6.3
P593_RIL1_177 B 9.8 P593_RIL1_181 B 6.3
P593_RIL1_192 A 9.5 P593_RIL1_102 B 6.2
P593_RIL1_35 A 9.4 P593_RIL1_152 B 6.1
P593_RIL1_101 A 9.4 P593_RIL1_13 B 6.1
P593_RIL1_60 B 9.4 P593_RIL1_138 B 6.1
P593_RIL1_54 A 9.3 P593_RIL1_189 A 6.0
P593_RIL1_196 A 9.2 P593_RIL1_180 A 5.9
P593_RIL1_173 A 9.1 P593_RIL1_24 B 5.8
P593_RIL1_48 A 9.1 P593_RIL1_153 B 5.7
P593_RIL1_28 A 9.0 P593_RIL1_52 B 5.7
P593_RIL1_105 A 9.0 P593_RIL1_17 B 5.5
P593_RIL1_120 A 9.0 P593_RIL1_15 B 5.4
P593_RIL1_38 A 8.8 P593_RIL1_89 B 5.2
P593_RIL1_61 A 8.8 P593_RIL1_155 B 5.2
P593_RIL2 P593_RIL2_88 B 9.0 P593_RIL2_114 B 6.0
P593_RIL2_100 B 9.0 P593_RIL2_116 B 6.0
P593_RIL2_47 A 8.8 P593_RIL2_261 B 6.0
P593_RIL2_27 A 8.8 P593_RIL2_300 B 6.0
P593_RIL2_212 B 8.7 P593_RIL2_269 B 5.9
P593_RIL2_76 B 8.7 P593_RIL2_117 B 5.9
P593_RIL2_254 A 8.6 P593_RIL2_63 B 5.8
P593_RIL2_146 A 8.5 P593_RIL2_164 B 5.8
P593_RIL2_3 A 8.4 P593_RIL2_215 B 5.8
P593_RIL2_220 B 8.4 P593_RIL2_178 B 5.7
P593_RIL2_256 B 8.3 P593_RIL2_128 B 5.7
P593_RIL2_240 A 8.3 P593_RIL2_305 B 5.5
P593_RIL2_202 B 8.3 P593_RIL2_106 B 5.4
P593_RIL2_187 B 8.2 P593_RIL2_267 B 5.3
P593_RIL2_210 B 8.2 P593_RIL2_159 B 5.2
Open in a new tab

a P593_RIL1, a bi-parental mapping population from a cross between S16-14161 and PI 506593; P593_RIL2, a bi-parental mapping population from a cross between S16-11651 and PI 506593

b Allele, SNP marker position at 41,834,095 bp based on Wm82.a2.v1. Allele A was derived from PI 506593, while allele B was derived from the low-sucrose parental line

c Mean sucrose content across three environments

Candidate gene identification

Based on the variant identification results on Wm82.a2.v1 and Wm82.a5.v1, the genetic variation and functional annotation of all possible candidate genes within the major QTL region on Chr. 8 were evaluated. Based on Wm82.a2.v1, 175 genes were located in the major QTL region on Chr. 8. Of these, 44 genes carried 85 non-synonymous SNPs and 11 Indels that were only present in PI 506593 (Suppl. Tables S2 and S3). Within the corresponding region of Wm82.a5.v1, 54 genes carried 132 non-synonymous SNPs and 31 Indels in PI 506593 (Suppl. Tables S4 and S5). Among them, 34 genes were common but showed slightly different variants between the two reference genomes (Fig. 1; Table 5). An additional 20 genes on Wm82.a5.v1 were also described in Table 6. The further analyses were conducted based on the 54 candidate genes identified in Wm82.a5.v1. Functional annotations of 54 candidate genes were obtained from Soybase (https://www.soybase.org/) (Tables 5 and 6). According to the BLAST results, the 11 unannotated candidate genes (genes without annotation information in Tables 5 and 6) were described as uncharacterized, hypothetical, unnamed, or missing proteins in soybean (Glycine max and Glycine soja), common bean (Phaseolus vulgaris), and Arabidopsis (Arabidopsis thaliana) (Suppl. Table S6).

Fig. 1.

Fig. 1

Open in a new tab

A comparison of the physical position of genes carrying non-synonymous and Indel variants within the major QTL region on Chr. 8 between Wm82.a2.v1 and Wm82.a5.v1 reference genome assemblies. The orange lines indicate the 34 overlapping genes between two reference genome assemblies

Table 5.

The 34 overlapping candidate genes that carried novel variants in the major QTL on Chr. 8 in PI 506593 on Wm82.a2.v1 and Wm82.a5.v1 reference genomes

Wm82.a2.v1 Wm82.a5.v1 Annotationb
Gene ID Position Variant typea Gene ID Position Variant type
Glyma.08G293000 40,766,950–40,775,274 SNP Gm_Wm82_23206 43,220,985–43,228,708 SNP Sin3 family co-repressor
Glyma.08G293100 40,795,741–40,800,913 SNP Gm_Wm82_23209 43,249,676–43,259,978 SNP Sin3 family co-repressor
Glyma.08G293800 40,883,195–40,885,068 Indel Gm_Wm82_23219 43,333,545–43,338,512 Indel Proline-rich protein 4-like
Glyma.08G293900 40,887,462–40,910,754 SNP Gm_Wm82_23220 43,341,530–43,364,303 SNP Phospholipid-transporting ATPase-like protein
Glyma.08G294100 40,942,975–40,947,622 SNP Gm_Wm82_23222 43,396,901–43,401,240 SNP -c
Glyma.08G294300 40,953,540–40,957,782 SNP Gm_Wm82_23224 43,407,555–43,415,567 SNP Galactosyltransferase family protein
Glyma.08G294800 40,993,962–41,001,629 SNP Gm_Wm82_23227 43,447,708–43,455,285 SNP Actin cross-linking protein
Glyma.08G295100 41,011537–41,012,029 SNP Gm_Wm82_23229 43,462,185–43,465,739 SNP Disease resistance protein (TIR-NBS-LRR class)
Glyma.08G295400 41,026,575–41,033,381 SNP/Indel Gm_Wm82_23231 43,479,691–43,485,976 SNP/Indel Disease resistance protein (TIR-NBS-LRR class)
Glyma.08G295700 41,062,905–41,071,504 SNP Gm_Wm82_23234 43,516,488–43,524,479 SNP AMP deaminase-like
Glyma.08G295800 41,100,889–41,103,542 SNP Gm_Wm82_23237 43,554,115–43,556,053 SNP Homeobox associated leucine zipper
Glyma.08G296300 41,215,611–41,223,730 SNP Gm_Wm82_23243 43,668,712–43,675,770 SNP Plant/MEB5-like protein
Glyma.08G296500 41,241,857–41,246,757 SNP Gm_Wm82_23245 43,694,878–43,699,270 SNP Mediator of RNA polymerase II transcription subunit 10
Glyma.08G297000 41,326,371–41,329,183 SNP Gm_Wm82_23251 43,779,198–43,781,761 SNP AP2-like ethylene-responsive transcription factor
Glyma.08G297400 41,358,747–41,363,041 SNP Gm_Wm82_23255 43,811,714–43,815,086 SNP Protein kinase superfamily protein
Glyma.08G297500 41,378,414–41,383,832 SNP Gm_Wm82_23256 43,831,982–43,836,271 SNP Protein IQ-DOMAIN 31-like isoform X6
Glyma.08G297700 41,412,445–41,415,775 SNP Gm_Wm82_23258 43,864,913–43,868,558 SNP F-box/RNI-like superfamily protein
Glyma.08G299500 41,767,501–41,769,958 SNP Gm_Wm82_23276 44,211,989–44,214,205 SNP Peroxisomal adenine nucleotide carrier 1
Glyma.08G299800 41,795,912–41,796,546 SNP/Indel Gm_Wm82_23279 44,240,488–44,241,228 SNP Chitinase A
Glyma.08G300400 41,830,891–41,843,406 SNP Gm_Wm82_23285 44,278,415–44,283,458 SNP –
Glyma.08G300500 41,865,484–41,867,085 SNP Gm_Wm82_23287 44,310,063–44,311,664 SNP Phytochrome kinase substrate 1
Glyma.08G301300 41,931,888–41,933,603 SNP Gm_Wm82_23293 44,376,527–44,377,790 SNP Ribosomal protein L34e
Glyma.08G301700 41,962,541–41,968,109 SNP/Indel Gm_Wm82_23298 44,407,335–44,411,704 Indel –
Glyma.08G302400 42,046,386–42,051,525 SNP Gm_Wm82_23302 44,491,765–44,496,137 SNP Pre-mRNA-splicing factor cwc22
Glyma.08G303000 42,114,246–42,121,684 SNP/Indel Gm_Wm82_23307 44,559,451–44,566,133 SNP/Indel DEAD-box ATP-dependent RNA helicase-like protein
Glyma.08G303300 42,134,782–42,139,026 Indel Gm_Wm82_23310 44,579,782–44,584,026 Indel –
Glyma.08G303500 42,150,074–42,151,723 SNP/Indel Gm_Wm82_23312 44,595,014–44,596,723 SNP/Indel Disease resistance protein (TIR-NBS-LRR class)
Glyma.08G303600 42,160,184–42,161,884 SNP Gm_Wm82_23313 44,605,184–44,610,595 SNP Disease resistance protein (TIR-NBS-LRR class)
Glyma.08G304300 42,265,594–42,273,748 Indel Gm_Wm82_23320 44,711,920–44,718,967 Indel WD repeat-containing protein 26-like
Glyma.08G305600 42,374,266–42,376,352 SNP Gm_Wm82_23334 44,821,303–44,823,057 SNP Disease resistance protein (TIR-NBS-LRR class)
Glyma.08G307300 42,551,411–42,553,771 SNP Gm_Wm82_23351 44,989,964–44,992,324 SNP Receptor serine/threonine kinase
Glyma.08G307400 42,556,212–42,559,048 SNP Gm_Wm82_23352 44,994,964–44,997,445 SNP/Indel Receptor-like protein kinase 1
Glyma.08G307700 42,603,547–42,605,822 SNP Gm_Wm82_23357 45,040,796–45,043,149 SNP 60S ribosomal protein L18A-1
Glyma.08G309200 42,818,069–42,821,065 SNP Gm_Wm82_23372 45,255,743–45,258,461 SNP Glycerol-3-phosphate acyltransferase 1
Open in a new tab

a SNP, Single nucleotide polymorphism; Indel, Insertion and deletion

b Annotation, the annotation was based on Wm82.a5.v1 reference genome using Soybase (https://www.soybase.org/)

c -, Data not available

Table 6.

Additional candidate genes that carried novel variants in the major QTL on Chr. 8 in PI 506593 on Wm82.a5.v1 reference genomes only

Gene ID Position Variant typea Gene ID_Wm82.a2.v1b Annotationc
Gm_Wm82_23199 43,084,276–43,087,170 SNP -d –
Gm_Wm82_23200 43,094,522–43,095,517 SNP – –
Gm_Wm82_23213 43,297,226–43,298,074 SNP – –
Gm_Wm82_23216 43,328,678–43,328,950 SNP – –
Gm_Wm82_23218 43,331,428–43,333,293 SNP – –
Gm_Wm82_23226 43,439,054–43,442,901 SNP Glyma.08G294700 Dentin sialophosphoprotein-like isoform X1
Gm_Wm82_23230 43,479,691–43,485,976 SNP – Disease resistance protein (TIR-NBS-LRR class)
Gm_Wm82_23235 43,529,584–43,530,663 SNP – Putative nuclease HARBI1-like
Gm_Wm82_23236 43,548,217–43,552,351 SNP – Putative nuclease HARBI1-like
Gm_Wm82_23257 43,861,288–43,861,777 SNP – Flocculation protein FLO11-like isoform X4
Gm_Wm82_23273 44,158,625–44,159,176 Indel Glyma.08G299000 –
Gm_Wm82_23277 44,220,684–44,221,889 SNP/Indel Glyma.08G299600 Probable membrane-associated kinase regulator 6-like
Gm_Wm82_23278 44,227,342–44,229,272 SNP Glyma.08G299700 Chitinase A
Gm_Wm82_23284 44,276,125–44,276,906 SNP Glyma.08G300300 Chitinase A
Gm_Wm82_23299 44,444,016–44,444,536 SNP – Protein kinase superfamily protein
Gm_Wm82_23305 44,530,279–44,536,400 SNP/Indel Glyma.08G302800 Chromodomain helicase DNA-binding protein
Gm_Wm82_23345 44,919,545–44,922,716 SNP Glyma.08G306700 Disease resistance protein (TIR-NBS-LRR class)
Gm_Wm82_23353 45,001,129–45,004,003 SNP – Phototropin 1
Gm_Wm82_23356 45,029,632–45,035,122 SNP – Receptor-like protein 12-like
Gm_Wm82_23371 45,253,027–45,253,530 Indel – –
Open in a new tab

a SNP, Single nucleotide polymorphism; Indel, Insertion and deletion

b Gene ID_Wm82.a2.v1, the corresponding gene ID based on Wm82.a2.v1

c Annotation, the annotation was based on Wm82.a5.v1 reference genome using Soybase (https://www.soybase.org/)

d -, Data not available

A total of 132 non-synonymous SNPs and five non-frameshift Indels in the major QTL region on Chr. 8 were evaluated for their biological impacts on the candidate gene functions using PROVEAN (Fig. 2). Based on the suggested threshold (PROVEAN score ≤ − 2.5), the donor parent, PI 506593, carried 34 deleterious variants, which can significantly affect the biological functions of 18 candidate genes (Fig. 2a). Among these, the deletion of six amino acids from Lysine (K) 203 to Valine (V) 208 (PROVEAN score = − 20.674) in Gm_Wm82_23219 (Glyma.08G293800, Wm82.a2.v1) showed the most deleterious impact on the biological function (Fig. 2b). One amino acid substitution from Arginine (R) to Cysteine (C) at position 85 (PROVEAN score = − 7.845) in Gm_Wm82_23299 (Gene name was not available on Wm82.a2.v1) was the most deleterious non-synonymous SNP variant, followed by the amino acid substitution from Phenylalanine (F) to Serine (S) at position 495 (PROVEAN score = − 7.822) in Gm_Wm82_23352 (Glyma.08G307400, Wm82.a2.v1) in PI 506593. The average PROVEAN score of those deleterious variants was -5.288 (Fig. 2b).

Fig. 2.

Fig. 2

Open in a new tab

Results from PROVEAN analysis testing 132 non-synonymous SNP and five non-frameshift Indel variants within the major QTL region on Chr. 8 showed a a total number of deleterious and neutral variants and b PROVEAN scores of deleterious variants. The blue line indicates the average PROVEAN score of deleterious variants, which was − 5.288. The red line indicates the significant threshold, which was − 2.5

RNA-seq analysis utilized only genes with corresponding gene IDs for either Wm82.a1.v1 or Wm82.a4.v1 required for RNA-seq data search in Severin et al. (2010) and Almeida-Silva et al. (2023), respectively. The corresponding gene IDs between Wm82.a1.v1, Wm82.a2.v1, Wm82.a4.v1, and Wm82.a5.v1 were described in Suppl. Table S7. A total of 34 candidate genes were evaluated for their expression levels in different plant tissues using publicly available databases described by Severin et al. (2010) and Almeida-Silva et al. (2023), respectively (Fig. 3). According to the RNA-seq dataset from Severin et al. (2010), Glyma.08G301300 (Gm_Wm82_23293, Wm82.a5.v1), Glyma.08G302400 (Gm_Wm82_23302, Wm82.a5.v1), and Glyma.08G304300 (Gm_Wm82_23320, Wm82.a5.v1) showed noticeably higher expression levels in seed tissues during seed development stages than other candidate genes (Fig. 3a). Two genes, Glyma.08G293800 (Gm_Wm82_23219, Wm82.a5.v1) and Glyma.08G299000 (Gm_Wm82_23273, Wm82.a5.v1), showed higher gene expression levels during early pod development stages (from 1_cm_pod to Pod_shell_14DAF). According to the RNA-seq dataset from Almeida-Silva et al. (2023), Glyma.08G293800 and Glyma.08G301300 showed outstanding gene expression levels across 19 plant parts (> 1000 TPM) (Fig. 3b). The high gene expression levels of Glyma.08G293800 were observed in the shoot, leaf, flower, and pod, while Glyma.08G301300 was highly expressed in all plant parts.

Fig. 3.

Fig. 3

Open in a new tab

The candidate gene expression levels within the major QTL region on Chr. 8 using publicly available gene expression datasets from a Severin et al. (2010) and b Almeida-Silva et al. (2023), respectively. The read counts for a and b were normalized using Reads Per Kilobase of transcript per Million mapped reads (RPKM) and Transcripts Per Million (TPM) method, respectively

Discussion

Similar to other seed composition traits in soybean, sucrose content fluctuates in response to various environmental factors, such as temperature, precipitation, and air and soil humidity, that affect plant growth, flowering time, and seed development (Jo et al. 2018; Lee et al. 2023a, b; Tarumingkeng and Coto 2003; Wang et al. 2023a). In this study, the average sucrose content was the highest in BREC_22 and the lowest in FDREEC_21 (Table 1). Although many studies reported a negative correlation between temperature and sucrose content (lower temperature induces higher sucrose) during pod-filling stages, our study did not strictly follow this trend. However, according to meteorological data from nearby weather stations (https://www.wunderground.com/), from late September to October, BREC_22 had the lowest average temperature of 15.2 °C, while FDREEC_21 had the highest average temperature of 17.8 °C. This observation suggests that the sucrose content of soybean may be more sensitive to lower temperatures during the late reproductive stages (R6–R7) than during the early pod-filling stages (R3–R5). Soil moisture is another important environmental factor affecting the sucrose content. However, the relationship between soil moisture and sucrose content at present is unclear (Du et al. 2020; Wang et al. 2021; Wijewardana et al. 2019). We used the dew point temperature to determine the correlation between soil moisture and sucrose content. The dew point temperature provides an actual amount of water vapor in the air and positively correlates with soil moisture content (Leelamanie 2010). In this study, the sucrose content was negatively correlated with the dew point temperature during the pod-filling stages, indicating that the lower the soil moisture content, the higher the sucrose content. This finding was consistent with previous studies showing the negative correlation between soil moisture and sucrose content in soybean (Du et al. 2020; Wang et al. 2023a).

A total of 11 QTLs on Chrs. 4, 5, 6, 8, 10, and 13 were identified in two populations sharing a high sucrose donor parent across three environments (Table 3). Among them, QTLs on Chr. 8 (qSUC_08.1, qSUC_08.2, qSUC_08.3, and qSUC_08.4) were consistently detected across all environments in two populations except BREC_22 in P593_RIL1 (Table 3). Since two populations shared the same donor parent, a genomic region harboring these QTLs on Chr. 8 was identified as a major QTL region (the physical interval of 40,597,410–42,861,364 bp, Wm82.a2.v1) that underlies the high sucrose trait in PI 506593. In a previous study, this major QTL region was located near the QTL (43,602,528–43,670,249 bp) identified by Lee et al. (2016) from a population derived from a cross of Keunolkong × Iksan 10, although they did not overlap with our findings. Other minor QTLs identified on Chrs. 4, 5, 6, 10, and 13 in our study seemed unstable and inconsistent across environments (Table 3). Two QTLs, qSUC_06.1 and qSUC_06.2, were only identified on Chr. 6 in the P593_RIL1 at the physical intervals of 16,133,204–21,786,787 bp and 21,786,787–46,805,251 bp, respectively (Table 3). However, due to the lack of polymorphic markers on Chr. 6, the two QTLs constituted more than 50% of the entire chromosome, which spans approximately 30 Mbp. In other studies, most of the sucrose-related QTLs identified on Chr. 6 were located within this region (Liu et al. 2023; Patil et al. 2018; Skoneczka et al. 2009). A minor QTL on Chr. 13 (qSUC_13.1) identified in P593_RIL1 was located at the physical interval of 25,268,844–25,927,261 bp, which was close to the QTL that Maughan et al. (2000) previously identified at the physical interval of 26,196,486–28,912,864 bp (‘Sucrose 1–5’ in Soybase). The other minor QTLs on Chrs. 4, 5, 10, and 13 have not been reported near these regions from previous studies.

Desirable marker testing in both populations showed that the SNP marker positioning at 41,834,095 bp on Chr. 8 had a high efficiency in differentiating between phenotypically extreme groups in P593_RIL1. On the other hand, the SNP marker had a high discrimination power for only low sucrose lines in P593_RIL2. This discrepancy could occur due to the lack of SNP markers anchored on Chr. 8 in P593_RIL2. P593_RIL1 contained 104 SNP markers with an average interval of 1.4 cM, while P593_RIL2 contained only 24 SNP markers with an average interval of 4.9 cM on Chr. 8 (Table 2). Although the significant SNP marker identified in this study could be included in the marker-assisted selection pipeline for high sucrose soybean breeding along with other molecular markers, such as rs2W331-, rs2T107I, rs3snp5/rs3snp6, MIPS1, further investigations, such as fine mapping followed by molecular marker development, within the major QTL region, will increase the marker efficiency and precision.

In this study, the WGS data of the three parental lines, PI 506593, S16-14161, and S16-11651, constituting the two mapping populations, were analyzed to identify SNP and Indel variants within the significant QTL regions. This genomic information enabled us to narrow down the candidate genes within sucrose-related QTL regions identified by a conventional bi-parental mapping analysis. Furthermore, this study was the first to employ the high-quality reference genome assembly of Wm82.a5.v1 in QTL analysis compared to the most widely used reference genome version of Wm82.a2.v1. To date, most novel genomic findings in soybean research were identified and reported based on Wm82.a2.v1. However, some variants identified in Wm82.a2.v1 were not found in Wm82.a5.v1 and vice versa (Fig. 1; Suppl. Tables S4 and S5). This is mainly because most of the gaps and scaffold breaks in Wm82.a2.v1 were completed in Wm82.a5.v1, resulting in different numbers of total genes between Wm82.a2.v1 (56,044 genes) and Wm82.a5.v1 (58,287 genes). Specifically, the total length of Chr. 8 in Wm82.a5.v1 is 50,302,612 bp, whereas that of Chr. 8 in Wm82.a2.v1 is 47,837,940 bp.

We studied potential candidate genes associated with sucrose content in PI 506593 using PROVEAN and RNA-seq analysis. These analyses were informative in identifying promising candidate genes among the 54 genes within the major QTL region. However, further investigations, such as fine mapping, molecular marker development, and gene editing, will be required to pinpoint causal variants and genes. Three candidate genes, Gm_Wm82_23219 (Glyma.08G293800, Wm82.a2.v1), Gm_Wm82_23299 (Gene name was not available on Wm82.a2.v1), and Gm_Wm82_23352 (Glyma.08G307400, Wm82.a2.v1) were the most significant in PROVEAN (Fig. 2). Interestingly, the functional annotations of these candidate genes are proline-rich protein 4-like, protein kinase superfamily protein, and receptor-like protein kinase 1, respectively, which are responsible for the stress reaction against various abiotic stressors, such as drought, salinity, etc. (Cui et al. 2022; Kavi Kishor et al. 2015; Rajasheker et al. 2022) (Tables 5 and 6). Especially during the reproductive process, those genes play a key role in the onset of leaf senescence, which is crucial in disassembling macromolecules and reallocating nutrients into developing seeds (Yang et al. 2022). Sucrose accumulation begins at the early pod-filling stages and ceases when the seed color turns yellow, which is the onset of leaf senescence (Obendorf et al. 1997). Thus, the candidate genes could be responsible for the stress-induced leaf senescence and regulation of the soluble carbohydrate metabolic process (Yang et al. 2022). Also, a recent study identified a stable major QTL associated with sucrose content in peanut (Arachis hypogaea L.) and highlighted a candidate gene encoding receptor-like protein kinase (Li et al. 2023). We also identified six additional candidate genes, including Gm_Wm82_23229 (Glyma.08G295100, Wm82.a2.v1), Gm_Wm82_23230 (Gene name was not available on Wm82.a2.v1), Gm_Wm82_23312 (Glyma.08G303500, Wm82.a2.v1), Gm_Wm82_23313 (Glyma.08G303600, Wm82.a2.v1), Gm_Wm82_23334 (Glyma.08G305600, Wm82.a2.v1), and Gm_Wm82_23345 (Glyma.08G306700, Wm82.a2.v1), that were annotated as disease resistance protein (Tables 5 and 6). As a main signaling molecule, the translocation and utilization of sucrose are significantly affected by not only external environmental stressors but also biotic stressors (Aluko et al. 2021; Tauzin and Giardina 2014). Therefore, the variants significantly affecting the biological function of stress response-related genes could regulate the reallocation of sucrose, resulting in the modification of seed sucrose content in soybean.

Furthermore, a subsequent RNA-seq analysis highlighted five other candidate genes. They were Glyma.08G293800 (Gm_Wm82_23219, Wm82.a5.v1), Glyma.08G299000 (Gm_Wm82_23273, Wm82.a5.v1), Glyma.08G301300 (Gm_Wm82_23293, Wm82.a5.v1), Glyma.08G302400 (Gm_Wm82_23302, Wm82.a5.v1), and Glyma.08G304300 (Gm_Wm82_23320, Wm82.a5.v1). Of these, Glyma.08G293800, highlighted from PROVEAN results, also showed high gene expression levels in pod tissues (Fig. 3a). Since seed composition traits are closely related to reproductive stages, the identification of candidate genes with high gene expression levels in the pod and seed tissues is important (Lee et al. 2023b). Two genes, Glyma.08G301300 and Glyma.08G302400, were related to the post-transcriptional process (Ribosomal protein L34e and Pre-mRNA-splicing factor cwc22, respectively), but further investigation will be required to confirm the association with sucrose-related genes. The annotation of Glyma.08G299000 was unknown, uncharacterized, and unnamed in Glycine max, Glycine soja, Phaseolus vulgaris, and Arabidopsis thaliana, respectively, based on BLAST results, although it showed relatively high gene expression levels in pod tissues during early reproductive stages (Fig. 3a).

Despite not being highlighted in PROVEAN and RNA-seq analysis, a novel variant in one candidate gene, Gm_Wm82_23224 (Glyma.08G294300, Wm82.a2.v1), which is closely related to sucrose accumulation, was identified in PI 506593 on both reference genomes. Gm_Wm82_23224 encodes a galactosyltransferase family protein, of which two well-known soluble carbohydrate-related genes, RS2 and RS3, also encode similar proteins, galactinol-sucrose galactosyltransferase (https://www.soybase.org/). Also, Gm_Wm82_23224 is homologous to the GATL4–6 genes, sharing 58–60% amino acid similarities in Arabidopsis thaliana. The GALT genes encode the galactosyltransferase family protein in Arabidopsis thaliana. Galactosyltransferase plays a key role in transferring galactose and regulating carbon partitioning between sucrose and raffinose in soybean seeds (Saravitz et al. 1987; Singer et al. 2023). Although this gene was not highlighted in PROVEAN and RNA-seq analysis, the variant is still worth highlighting due to the biological function of the gene directly related to sucrose accumulation.

Conclusion

In this study, QTL analyses using two mapping populations derived from the same donor parent, PI 506593, identified a major QTL region on Chr. 8 controlling sucrose content in soybean seeds. The WGS data of the parental lines aligned to Wm82.a2.v1 and Wm82.a5.v1 were used to identify SNP and Indel variants within the major QTL region and candidate genes. Based on PROVEAN and RNA-seq analysis, 18 and 5 candidate genes were highlighted, respectively. Among these, Gm_Wm82_23219 (Glyma.08G293800, Wm82.a2.v1) in PI 506593 carried the most deleterious non-synonymous variant and was highly expressed in pod tissues during the early reproductive stages in soybean. The findings in our study suggested that candidate genes and novel variants within the major QTL region on Chr. 8 can be useful genetic resources for improving carbohydrate profiles in new soybean cultivars aimed at animal feed efficiency and human consumption.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary file1 (XLSX 59 KB) (59KB, xlsx)

Acknowledgements

The authors would like to thank the soybean breeding team at the University of Missouri–Fisher Delta Research, Extension, and Education Center for their dedication and supporting the field trials. The authors also extend gratitude to the United Soybean Board (USB) for financial support, Dr. Aamir Khan for providing a Liftover file, and Haiying Shi for technical support on HPLC analysis. Last but not least, we would like to acknowledge Dr. Pengyin Chen, who led this project, although sadly, he is no longer with us.

Author Contribution statement

DL conducted population development, field experiments, phenotypic data collection, and data analysis and prepared the first draft of the manuscript. QS genotyped the population and edited the manuscript. TDV and FL supervised the overall research and edited the manuscript. JGS and HTN conceptualized, supervised, financially supported the overall research, and edited the manuscript. All authors revised and approved the manuscript for submission.

Funding

This work was supported by the United Soybean Board, project #1420–532-5639.

Data Availability

The datasets generated during and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Conflict of interest

The authors declare that they have no conflict of interest.

Footnotes

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  1. Akond M, Liu S, Kantartzi SK et al (2015) Quantitative trait loci underlying seed sugars content in “MD96-5722” by “Spencer” recombinant inbred line population of Soybean. Food Nutr Sci 6:964–973. 10.4236/fns.2015.611100 [Google Scholar]
  2. Almeida-Silva F, Pedrosa-Silva F, Venancio TM (2023) The soybean expression atlas v2: a comprehensive database of over 5000 RNA-seq samples. Plant J 116(4):1041–1051. 10.1111/tpj.16459 [DOI] [PubMed] [Google Scholar]
  3. Aluko OO, Li C, Wang Q et al (2021) Sucrose utilization for improved crop yields: a review article. Int J Mol Sci 22(9):4704. 10.3390/ijms22094704 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bolger AM, Lohse M, Usadel B (2014) Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics 30(15):2114–2120. 10.1093/bioinformatics/btu170 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Chen P, Ali ML, Shannon G et al (2022) Registration of ‘S16-11651C’, a conventional soybean cultivar with high yield, resistance to multiple diseases, and broad adaptation. J Plant Regist 16(2):329–340. 10.1002/plr2.20224 [Google Scholar]
  6. Choi Y, Chan AP (2015) PROVEAN web server: a tool to predict the functional effect of amino acid substitutions and indels. Bioinformatics 31(16):2745–2747. 10.1093/bioinformatics/btv195 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Cui Y, Lu X, Gou X (2022) Receptor-like protein kinases in plant reproduction: current understanding and future perspectives. Plant Commun 3(1):100273. 10.1016/j.xplc.2021.100273 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Dierking EC, Bilyeu KD (2008) Association of a soybean raffinose synthase gene with low raffinose and stachyose seed phenotype. Plant Genome 1:135–145. 10.3835/plantgenome2008.06.0321 [Google Scholar]
  9. Dierking EC, Bilyeu KD (2009) New sources of soybean seed meal and oil composition traits identified through TILLING. BMC Plant Biol 9(1):89. 10.1186/1471-2229-9-89 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Dozier WA, Hess JB (2011) Soybean meal quality and analytical techniques. In: El-Shemy H (ed) Soybean and Nutrition. IntechOpen, London, UK, pp 111–124. 10.5772/1008 [Google Scholar]
  11. Du Y, Zhao Q, Chen L et al (2020) Effect of drought stress during Soybean R2–R6 growth stages on sucrose metabolism in leaf and seed. Mol Sci 21(2):618. 10.3390/ijms21020618 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Fehr WR (1987) Breeding methods for cultivar development. In Wilcox JR (Eds.) Soybeans: improvement, production, and uses. ASA, CSSA, and SSSA, Madison, pp 249–293
  13. Feng CB, Morsy M, Giannoccaro E et al (2005) Soybean seed sugar content and quantitative trait loci mapping. In: Li CJ et al (eds) Plant nutrition for food security, human health, and environmental protection. Tsinghua University Press, Beijing, pp 438–439 [Google Scholar]
  14. Garg V, Khan AW, Fengler K (2023) Near-gapless genome assemblies of Williams 82 and Lee cultivars for accelerating global soybean research. Plant Genome 16(4):e20382. 10.1002/tpg2.20382 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Hagely KB, Jo H, Kim JH et al (2020) Molecular-assisted breeding for improved carbohydrate profiles in soybean seed. Theor Appl Genet 133:1189–1200. 10.1007/s00122-020-03541-z [DOI] [PubMed] [Google Scholar]
  16. Jo H, Lee JD, Bilyeu KD (2018) Environmental stability of carbohydrate profiles in different soybean genotypes. Crop Sci 58:773–782. 10.2135/cropsci2017.08.0497 [Google Scholar]
  17. John PGK (2008) Sugar syrup: the new energy feed for poultry. World Poult Sci 24:12–13 [Google Scholar]
  18. KaviKishor PBK, HimaKumari P, Sunita MSL (2015) Role of proline in cell wall synthesis and plant development and its implications in plant ontogeny. Front Plant Sci 6:544. 10.3389/fpls.2015.00544 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Kerr PS, Sebastian SA (2000) Soybean products with improved carbohydrate composition and soybean plants. United States Patent 6,147,193
  20. Kim HK, Kang ST, Cho JH et al (2005) Quantitative trait loci associated with oligosaccharide and sucrose contents in soybean (Glycinemax L.). J Plant Biol 48:106–112. 10.1007/BF03030569 [Google Scholar]
  21. Kim HK, Kang ST, Oh KW (2006) Mapping of putative quantitative trait loci controlling the total oligosaccharide and sucrose content of Glycine max seeds. J Plant Res 119:533–538. 10.1007/s10265-006-0004-9 [DOI] [PubMed] [Google Scholar]
  22. Knizia D, Bellaloui N, Yuan J et al (2023) Quantitative trait loci and candidate genes that control seed sugars contents in the Soybean ‘Forrest’ by ‘Williams 82’ recombinant inbred line population. Plants 12(19):3498. 10.3390/plants12193498 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Lee JS, Kim SM, Kang S (2016) Fine mapping of quantitative trait loci for sucrose and oligosaccharide contents in soybean [Glycinemax (L.) Merr.] using 180 K Axiom SoyaSNP genotyping platform. Euphytica 208:195–203. 10.1007/s10681-015-1622-x [Google Scholar]
  24. Lee D, Vuong TD, Shanno G et al (2023a) Environmental stability and genetic effect of soybeans differing in mutant allele combinations between rs and mips1 genes for soluble carbohydrate profiles. Crop Sci 63(6):3326–3337. 10.1002/csc2.21094 [Google Scholar]
  25. Lee D, Lara L, Moseley D et al (2023b) Novel genetic resources associated with sucrose and stachyose content through genome-wide association study in soybean [Glycinemax (L.) Merr.]. Front Plant Sci 14:1294659. 10.3389/fpls.2023.1294659 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Leelamanie DAL (2010) Changes in soil water content with ambient relative humidity in relation to the organic matter and clay. Trop Agric Res Ext 13(1):6–10. 10.4038/tare.v13i1.3130 [Google Scholar]
  27. Li H, Durbin R (2009) Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics 25(14):1754–1760. 10.1093/bioinformatics/btp324 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Li W, Huang L, Liu N et al (2023) Identification of a stable major sucrose-related QTL and diagnostic marker for flavor improvement in peanut. Theor Appl Genet 136:78. 10.1007/s00122-023-04306-0 [DOI] [PubMed] [Google Scholar]
  29. Liu C, Chen H, Yu Q et al (2023) Identification of quantitative trait loci and candidate genes for seed sucrose and soluble sugar concentrations in soybean. Crop Sci 63:2976–2992. 10.1002/csc2.21080 [Google Scholar]
  30. Maughan PJ, SaghaiMaroof MA, Buss GR (2000) Identification of quantitative trait loci controlling sucrose content in soybean (Glycine max). Mol Breeding 6:105–111. 10.1023/A:1009628 [Google Scholar]
  31. Obendorf RL (1997) Oligosaccharides and galactosyl cyclitols in seed desiccation tolerance. Seed Sci Res 7:63–74. 10.1017/S096025850000341X [Google Scholar]
  32. Van Ooijen JW (2009) MapQTL® 6, Software for the mapping of quantitative trait loci in experimental populations of diploid species. Kyazma BV, Wageningen, Netherlands
  33. Van Ooijen JW (2018) JoinMap® 5, Software for the calculation of genetic linkage maps in experimental populations of diploid species. Kyazma BV, Wageningen, Netherlands
  34. Ostezan A, Prenger EM, Rosso L et al (2023) A chromosome 16 deletion conferring a high sucrose phenotype in soybean. Theor Appl Genet 136:109. 10.1007/s00122-023-04354-6 [DOI] [PubMed] [Google Scholar]
  35. Parsons CM, Zhang Y, Araba M (2000) Nutritional evaluation of soybean meals varying in oligosaccharide content. Poult Sci 79:1127–1131. 10.1093/ps/79.8.1127 [DOI] [PubMed] [Google Scholar]
  36. Patil G, Vuong TD, Kale S et al (2018) Dissecting genomic hotspots underlying seed protein, oil, and sucrose content in an interspecific mapping population of soybean using high density linkage mapping. Plant Biotechnol J 16:1939–1953. 10.1111/pbi.12929 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Pope M, Borg B, Boyd RD et al (2023) Qualifying the value of soybean meal in poultry and swine diets. J Appl Poultry Res 32(2):100337. 10.1016/j.japr.2023.100337 [Google Scholar]
  38. Rajasheker G, Nagaraju M, Varghese RP et al (2022) Identification and analysis of proline-rich proteins and hybrid proline-rich proteins super family genes from Sorghum bicolor and their expression patterns to abiotic stress and zinc stimuli. Front Plant Sci 13:952732. 10.3389/fpls.2022.952732 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Rani K, Kumar M, Razzaq A et al (2023) Recent advances in molecular marker technology for QTL mapping in plants. In: Wani SH, Wang D, Singh GP (eds) QTL Mapping in Crop Improvement: Present Progress and Future Perspectives. Academic Press, New York, pp 1–15 [Google Scholar]
  40. Rosset M, Prudencio SH, Beleia ADP (2012) Viscozyme L action on soy slurry affects carbohydrates and antioxidant properties of silken tofu. Food Sci Technol Int 18(6):531–538. 10.1177/1082013211433076 [DOI] [PubMed] [Google Scholar]
  41. Rosso ML, Burleson SA, Maupin LM et al (2011) Development of breeder-friendly markers for selection of MIPS1 mutations in soybean. Mol Breeding 28:127–132. 10.1007/s11032-011-9573-y [Google Scholar]
  42. Saghai Maroof MA, Buss GR (2008) Low phytic acid, low stachyose, high sucrose soybean lines. U.S. Patent 12/033,830
  43. Salari MW, Ongom PO, Thapa R et al (2021) Mapping QTL controlling soybean seed sucrose and oligosaccharides in a single family of soybean nested association mapping (SoyNAM) population. Plant Breeding 140(1):110–122. 10.1111/pbr.12883 [Google Scholar]
  44. Saravitz DM, Pharr DM, Carter TE (1987) Galactinol synthase activity and soluble sugars in developing seeds of four soybean genotypes. Plant Physiol 83(1):185–189. 10.1104/pp.83.1.185 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Schmutz J, Cannon S, Schlueter J et al (2010) Genome sequence of the palaeopolyploid soybean. Nature 463:178–183. 10.1038/nature08670 [DOI] [PubMed] [Google Scholar]
  46. Severin A, Woody JL, Bolon YT et al (2010) RNA-Seq Atlas of Glycine max: a guide to the soybean transcriptome. BMC Plant Biol 10:160. 10.1186/1471-2229-10-160 [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Simeone G, Bergamini M, Verga MC et al (2022) Do Vegetarian diets provide adequate nutrient intake during complementary feeding? A systematic review. Nutrients 14(17):3591. 10.3390/nu14173591 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Singer WM, Lee Y, Shea Z et al (2023) Soybean genetics, genomics, and breeding for improving nutritional value and reducing antinutritional traits in food and feed. Plant Genome 16(4):e20415. 10.1002/tpg2.20415 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Skoneczka JA, SaghaiMaroof MA, Shang, et al (2009) Identification of candidate gene mutation associated with low stachyose phenotype in Soybean Line PI200508. Crop Sci 49:247–255. 10.2135/cropsci2008.07.0403 [Google Scholar]
  50. Song Q, Jenkins J, Jia G et al (2016) Construction of high resolution genetic linkage maps to improve the soybean genome sequence assembly Glyma1.01. BMC Genomics 17:33. 10.1186/s12864-015-2344-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Song Q, Yan L, Quigley C et al (2020) Soybean BARCSoySNP6K: an assay for soybean genetics and breeding research. Plant J 104(3):800–811. 10.1111/tpj.14960 [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Sui M, Wang Y, Bao Y et al (2020) Genome-wide association analysis of sucrose concentration in soybean (Glycinemax L.) seed based on high-throughput sequencing. Plant Genome 13(3):e20059. 10.1002/tpg2.20059 [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Tarumingkeng RC, Coto Z (2003) Effects of drought stress on growth and yield of soybean. Kisman Sci Philos 702:798–807 [Google Scholar]
  54. Tauzin AS, Giardina T (2014) Sucrose and invertases, a part of the plant defense response to the biotic stresses. Front Plant Sci 5:293. 10.3389/fpls.2014.00293 [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Thapa R, Carrero-Colon M, Rainey KM et al (2019) TILLING by sequencing: a successful approach to identify rare alleles in soybean populations. Genes 10(12):1003. 10.3390/genes10121003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Valliyodan B, Shi H, Nguyen HT (2015) A simple analytical method for high-throughput screening of major sugars from soybean by normal-phase HPLC with evaporative light scattering detection. Chromatogr Res Int 2015:8. 10.1155/2015/757649 [Google Scholar]
  57. Valliyodan B, Cannon SB, Bayer PE et al (2019) Construction and comparison of three reference-quality genome assemblies for soybean. Plant J 100(5):1066–1082. 10.1111/tpj.14500 [DOI] [PubMed] [Google Scholar]
  58. Vuong TD, Sleper DA, Shannon JG et al (2010) Novel quantitative trait loci for broad-based resistance to soybean cyst nematode (Heterodera glycines Ichinohe) in soybean PI 567516C. Theor Appl Genet 121:1253–1266. 10.1007/s00122-010-1385-7 [DOI] [PubMed] [Google Scholar]
  59. Wang K, Li M, Hakonarson H (2010) ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data. Nucleic Acids Res 38(16):e164. 10.1093/nar/gkq603 [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Wang Y, Chen P, Zhang B (2014) Quantitative trait loci analysis of soluble sugar contents in soybean. Plant Breeding 133(4):493–498. 10.1111/pbr.12178 [Google Scholar]
  61. Wang X, Wu Z, Zhou Q et al (2021) Physiological response of soybean plants to water deficit. Front Plant Sci 12:809692. 10.3389/fpls.2021.809692 [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Wang T, Masedunskas A, Willett WC et al (2023a) Vegetarian and vegan diets: benefits and drawbacks. Eur Heart J 44(26):3423–3439. 10.1093/eurheartj/ehad436 [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Wang Z, Yu D, Morota G et al (2023b) Genome-wide association analysis of sucrose and alanine contents in edamame beans. Front Plant Sci 13:1086007. 10.3389/fpls.2022.1086007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Wijewardana C, Reddy KR, Bellaloui N (2019) Soybean seed physiology, quality, and chemical composition under soil moisture stress. Food Chem 278:92–100. 10.1016/j.foodchem.2018.11.035 [DOI] [PubMed] [Google Scholar]
  65. Yang F, Miao Y, Liu Y et al (2022) Function of protein kinases in leaf senescence of plants. Front Plant Sci 13:864215. 10.3389/fpls.2022.864215 [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Zeng A, Chen P, Shi A et al (2014) Identification of quantitative trait loci for sucrose content in soybean seed. Crop Sci 54(2):554–564. 10.2135/cropsci2013.01.0036 [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary file1 (XLSX 59 KB) (59KB, xlsx)

Data Availability Statement

The datasets generated during and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Articles from TAG. Theoretical and Applied Genetics. Theoretische Und Angewandte Genetik are provided here courtesy of Springer

RESOURCES