Identification of soybean mutants with low cesium accumulation, characterization of the causative gene, and field evaluation

Abstract

The radiocesium contamination of food poses a serious threat to food safety because radiocesium has a long half-life and emits harmful radiation during decay. Therefore, it is important to reduce radiocesium uptake by crops. In this study, we identified low-cesium-accumulating soybean mutants from an ethyl methanesulfonate-induced mutant population. Through the phenotypic screening of the population based on the seed cesium concentration, 10 candidate low-cesium-accumulating mutant lines were selected. Five of these exhibited significantly reduced seed radiocesium concentrations compared with the original variety, and one had accumulated an excessive amount of sodium. Since plant intracellular sodium ion homeostasis is regulated by the salt overly sensitive (SOS) pathway, sequence analysis of GmSOS1, which encodes an Na⁺/H⁺ antiporter in this pathway, revealed that the high-sodium-accumulating line contains a mutation in this gene. Additionally, two more gmsos1 mutant lines were isolated from the mutant population. In the field trial, the three independent gmsos1 mutant lines possessed lower seed cesium concentrations than the wild type. These results indicate that GmSOS1 is responsible for seed cesium accumulation, and gmsos1 mutants are potential breeding materials for reducing hazardous radiocesium accumulation in soybeans.

Introduction

An enormous amount of radiocesium was released into the agricultural land in eastern Japan following the nuclear accident at the Tokyo Electric Power Company’s Fukushima Daiichi Nuclear Power Plant (TEPCO–FDNPP) in March 2011. Since radiocesium has a relatively long half-life (2.06 years for ¹³⁴Cs and 30.2 years for ¹³⁷Cs) and emits harmful β and γ radiation during its decay, strategies aimed at reducing radiocesium accumulation in crops should be developed and implemented. As the chemical properties of cesium (Cs) are similar to those of potassium (K), the uptake and distribution of Cs are considered to occur through the same transport mechanism used by K. In fact, increasing K fertilizer application has resulted in decreased radiocesium uptake by various crops such as rice and soybeans (Hirayama et al. 2018, Ishikawa et al. 2018, Kato et al. 2015, Matsunami et al. 2021). As continuous K fertilizer application is labor intensive and expensive, “safer” crops that accumulate a small amount of radiocesium should be cultivated to reduce radiocesium contamination of agricultural products.

Previously, the genetic variation for the seed Cs concentration was surveyed in soybeans (Glycine max L.) using a set of mini-core collections representing a large genetic diversity (Kaga et al. 2012) to enable the selection of candidate accessions with traits for low Cs accumulation (Takagi et al. 2015). Subsequently, Uda et al. (2021) selected ‘Onihadaka’ as a low-Cs-accumulating candidate from the same mini-core collection. Using quantitative trait loci-sequence analysis, these authors highlighted the candidate genes responsible for the difference in Cs uptake between ‘Enrei’ and ‘Onihadaka’. However, identifying the responsible gene is challenging because the two accessions are genetically diverse, and genes other than the target gene can also influence evaluation of the seed Cs concentration.

Induced mutations are useful for examining gene function and improving crops; therefore, several mutant populations have been developed in a range of crops, including rice, maize, wheat, barley, and soybean (Sikora et al. 2011). Since a mutant population is composed of lines with the same genetic background, it is easier to identify the effects of mutant genes and exploit them as breeding materials, rather than relying on genetically diversified genetic resources. Genetic variability, as a result of mutations induced by various mutagens, has contributed to modern plant breeding and has played a major role in the development of superior plant varieties with improved productivity and quality (Oladosu et al. 2016). For example, the ‘Kosuzu’ soybean cultivar, is a γ-ray induced direct-use mutant cultivar, which was selected due to its earlier maturity and improved lodging resistance against cv. ‘Natto Shouryu’, which possesses small seeds and is well suited for use in the “natto” fermentation process (Hashimoto et al. 1988). Induced mutants can therefore be used to improve superior varieties within a short period of time, rendering them useful in situations where a rapid response is required, such as in the development of “safer” crops. To date, various mutant soybean populations have been developed using chemical mutagens such as ethyl methanesulfonate (EMS) and N-nitroso-N-methylurea (NMU) (Anai 2012, Cooper et al. 2008, Li et al. 2017, Tsuda et al. 2015). For example, Tsuda et al. (2015) constructed a high-density mutant library consisting of the DNA and seeds of 1,536 plants by repeating the EMS treatment for a Japanese soybean cultivar ‘Enrei’. This library has been successfully applied to select mutants with a phenotype of interest based on phenotypic screening. In the case of saponin, a major secondary metabolite in soybean seeds, novel compositional variants have been obtained, such as the deletion of 2,3-dihydro-2,5-dihydroxy-6-methyl-4H-pyran-4-one (DDMP) saponin through phenotypic screening (Krishnamurthy et al. 2019).

Despite these advances, to the best of our knowledge, no mutant soybean lines with low Cs accumulation have been reported. However, several mutant lines with reduced Cs accumulation have been identified by the phenotypic screening of rice. For example, Ishikawa et al. (2017) demonstrated that a mutant of OsSOS2, which encodes a serine/threonine-protein kinase required for the salt overly sensitive (SOS) pathway, exhibited reduced radiocesium concentrations in its grains. Notably, the SOS signaling pathway, which comprises SOS3, SOS2, and SOS1, mediates cellular signaling under salt stress to maintain ion homeostasis. More specifically, SOS3 encodes a myristoylated calcium-binding protein, wherein SOS3 interacts with and activates the serine/protein kinase SOS2, leading to activation of the downstream target SOS1, an Na⁺/H⁺ antiporter (Ji et al. 2013). The reduced Cs⁺ uptake in the ossos2 mutant may be closely related to the lower expression levels of several K⁺ and Na⁺ transporter genes due to the K⁺/Na⁺ imbalance in the roots caused by the OsSOS2 mutation (Ishikawa et al. 2017). The ossos2 mutant exhibited no difference in yield compared to the wild type, indicating that phenotypic selection is a viable approach for obtaining mutants suitable for use as breeding materials.

With these considerations in mind, the current study aimed to identify mutant lines with low levels of Cs accumulation. To achieve this, the high-density mutant library was searched to identify mutant lines with low Cs accumulation levels, and five lines with consistently low Cs contents were identified. One of the selected lines was a mutant of GmSOS1, a homolog of AtSOS1 and OsSOS1, which encode the Na⁺/H⁺ antiporter in the salt overly sensitive (SOS) pathway. A reduction in radiocesium levels in the field was subsequently confirmed using segregating populations derived from crosses between the original variety and gmsos1 mutants. The findings of this study are expected to contribute to the breeding of soybean varieties with low Cs accumulation, to reduce radiocesium contamination in soybeans and soybean-derived. Furthermore, the successful identification of multiple low cesium lines through the phenotypic screening of the mutant population in this study suggests that this strategy can be applied not only to harmful cesium accumulation, but also to the improvement of important agricultural traits, such as disease resistance and seed composition in soybean.

Materials and Methods

Plant materials and field conditions

The EMS-induced mutant library of the Japanese cultivar ‘Enrei’ (Tsuda et al. 2015) was referred to in this study. This mutant library consisted of associated DNA and seeds from 1,536 M2ʹ plants. In the phenotypic screening, 905 lines were used, which provided >50 M3ʹ seeds.

In addition to the 10 lines selected through phenotypic screening, three gmsos1 mutant lines (EnT-2953, EnT-3153, and EnT-3612) and their segregating populations (F₂, F₃, and F₄) developed from the crosses between the gmsos1 lines and the original variety (cv. Enrei) were used in field cultivation experiments. For this purpose, two fields in the Fukushima Prefecture with different soil characteristics were employed, namely Fields A and B (Table 1). Field A is a converted paddy field located 55 km northwest of TEPCO–FDNPP. The planting distance in Field A was 75 cm between rows and 15 cm between plants. Field B is an upland field located 65 km northwest of TEPCO–FDNPP. The planting distance in Field B was 75 cm between rows and 15 cm between plants.

Table 1.Soybean cultivation fields used in this study

Name	Field type	Year	Sowing date	Fertilizer (kg/ha)				Exchangeable ion concentrations1) (mg/kg)			137Cs concentration (kBq kg–1)
				N	P2O5	K2O		23Na	39K	133Cs
A	Converted paddy field	2018	May 29	30	100	–		12.2	65	0.218	2.46
		2019	May 28	30	100	–		12.3	53	0.240	2.11
		2020	May 29	30	100	–		11.3	60	0.235	1.88
B	Upland field	2020	June 29	30	100	100		5.4	366	0.187	n.d.2)

¹⁾ Exchangeable ion concentrations were measured using soil sampled after cultivation.

²⁾ ‘n.d.’ represents no data.

Analysis of ¹³³Cs and other elements in the plant samples

In the phenotypic screening of low-Cs-accumulating lines from the mutant population, the half-cotyledon of eight seeds from each M3ʹ line was freeze-dried. The freeze-dried cotyledons were individually ground into fine powder using a multibead shocker (Yasui Kikai, Osaka, Japan). Each ground sample (20 mg) was placed in a 15 mL metal-free centrifuge tube (INA·OPTIKA, Osaka, Japan) and digested in concentrated nitric acid (300 μL) at 95°C for 2 h. The volume of the digests was then made up to 10 mL by the addition of ultrapure water. The ¹³³Cs concentrations of the samples were determined using inductively coupled plasma–mass spectrometry (ICP–MS; Agilent 7700x, Agilent Technologies, Tokyo, Japan) with 5 μg/L ¹¹⁵In as an internal standard. To analyze the mineral content of the candidate low-Cs-accumulating mutant lines and cross populations between gmsos1 and ‘Enrei’, a sample of 10 seeds from each plant was dried at 105°C for 20–24 h; afterwards, the dried seeds were ground and digested as described above. The ²³Na, ³⁹K, and ¹³³Cs concentrations of the seeds were also determined by ICP–MS using 5 μg/L ¹¹⁵In as an internal standard. The ²³Na, ³⁹K, and ¹³³Cs concentrations of each sample were converted into the corresponding exchangeable element contents (mg/kg DW).

Search for candidate SOS genes in soybean

SOS genes were searched in the published Williams 82 genome sequence and annotated genes of soybean (Wm82.a2.v1 at Phytozome v13.0, https://phytozome-next.jgi.doe.gov/info/Gmax_Wm82_a2_v1) using the amino acid sequences of the SOS genes in Arabidopsis (i.e., AtSOS1 (AT2G01980), AtSOS2 (AT5G35410), and AtSOS3 (AT5G24270)) as queries. Genes were selected with a sequence homology of ≥60%; the genes that exhibited homology to the total length of amino acids were considered homolog candidates. The RNA-sequencing expression levels of the homologs in the soybean tissue, pod, root hair, leaves, root, nodules, seed, shoot apical meristem, stem, and flower were obtained from JBrowse of the Phytozome software (https://phytozome-next.jgi.doe.gov/).

Screening of gmsos1 mutants from the mutant library

The gmsos1 mutants were isolated from the soybean mutant library, as previously described by Tsuda et al. (2015). To screen the candidate mutants, 1,536 M2ʹ soybean plants were subjected to high-resolution melting (HRM) analysis using an ABI ViiA7 Real-Time PCR System (Applied Biosystems, Pleasanton, CA, USA). The mutation-bearing DNA amplicons detected through HRM analysis were sequenced using the ABI PRISM 3500xL Genetic Analyzer (Applied Biosystems). The primers used for HRM and sequence analyses are listed in Supplemental Table 1. The primers used to confirm the sequences of the coding regions of GmSOS1 (Glyma.08G092000) and of the AtSOS2 and AtSOS3 homologues (Glyma.04G235900, Glyma.06G128700, Glyma.13G166100, and Glyma.17G113700) in the gmsos1 mutants are shown in Supplemental Table 2. The sequences of GmSOS1 and of the AtSOS2 and AtSOS3 homologues in ‘Enrei’ were retrieved from DAIZU-net (https://daizu-net.dna.affrc.go.jp/ap/top).

The splicing pattern of GmSOS1 in EnT-2953 was determined by performing reverse transcription–PCR (RT–PCR). Total RNA was isolated from the roots of EnT-2953 plants using the RNeasy Plant Mini Kit (Qiagen, Valencia, CA, USA); subsequently, portions of the total RNA (1 μg) were subjected to RT–PCR using the QuantiTect Reverse Transcription Kit (Qiagen) and the provided RT-primer mix (total volume = 20 μL). The resulting complementary DNA (cDNA) products (1 μL) were subjected to PCR analysis using two primer sets designed to amplify the cDNAs derived from Glyma.08G092000 (5ʹ-ATGGAGGAAGAACAACAACAAC-3ʹ and 5ʹ-CTAGCGAAAAGATAGCGTGC-3ʹ). The amplified fragments were cloned into the pCR4 Blunt-TOPO vector (Invitrogen, Carlsbad, CA, USA), and the nucleotide sequences were confirmed using the 3500xL genetic analyzer (Applied Biosystems) and the BigDye-Terminator ver.3.1 cycle sequencing kit (Applied Biosystems). The primers are listed in Supplemental Table 3.

Genotyping of GmSOS1

To genotype the young leaves obtained from the segregating populations of the gmsos1 mutants (EnT-2953, EnT-3153, and EnT-3612), derived cleaved amplified polymorphic sequence (dCAPS) markers were developed to detect single-base substitutions that distinguish between the ‘Enrei’ and gmsos1 mutants. The primers used for dCAPS marker analyses are listed in Supplemental Table 4. The amplicons were digested with BbsI (EnT-2953), StuI (EnT-3153), or HphI (EnT-3612); subsequently, the restriction fragments were separated on a 3% agarose gel.

Analysis of exchangeable ions in soil

After cultivation, soil samples from Fields A and B were collected at a depth of 0–15 cm from five locations (four corners and the center) and then mixed. The soil samples were dried at 40°C for 3 d and then sieved to <2 mm. After extraction for 2 h using a 1 M NH₄OAc solution at a soil/solution ratio of 1:10, the exchangeable ²³Na, ³⁹K, and ¹³³Cs concentrations (mg/kg DW) in the soil samples were determined by ICP–MS (Agilent Technologies) using 5 μg/L ¹¹⁵In as an internal standard.

Analysis of ¹³⁷Cs concentrations in plant and soil samples

The ¹³⁷Cs concentrations of the seed and soil samples from Field A in 2018–2020 were determined using germanium detectors with multichannel analyzers (GC2520-7500SL and GCW2523–7905-30 U-ULB; Canberra Japan KK, Tokyo, Japan). The measurement results were decay-corrected to the harvest date (November 1) of the year of cultivation.

Results

Phenotypic screening of low-Cs-accumulating lines from a mutant soybean population

The M3ʹ mutant population used for screening the low-Cs-accumulating lines was harvested from a field in Tsukuba, Ibaraki, at a distance of >150 km away from the TEPCO–FDNPP. Given that the concentration of radiocesium (¹³⁴Cs and ¹³⁷Cs) in the seeds was too low to be measured, the nonradioactive Cs (¹³³Cs) content was determined, which was the only natural isotope of Cs present in a measurable amount in the soybean seeds. Given that there is no convincing evidence of discrimination by biological systems among ¹³³Cs, ¹³⁴Cs, and ¹³⁷Cs (Avery 1995), ¹³³Cs is expected to behave exactly like radiocesium in soybean plants. Although the distribution pattern of the seed ¹³³Cs concentration in the mutant population was continuous, four lines with a relative value of ≤0.5 were identified (Fig. 1). Considering the possibility of phenotypic errors due to the cultivation environment, the 10 mutant lines with the lowest seed ¹³³Cs concentrations were selected and cultivated in Field A. K fertilizer was not applied in Field A; hence, the exchangeable K content of the soil was 65 mg/kg, which is significantly lower than the recommended value (>208 mg/kg) for suppressing Cs absorption (Table 1) (MAFF, NARO and NIAES 2015). The exchangeable ion concentrations of ²³Na and ¹³³Cs in Field A were 12.2 and 0.218 mg/kg DW, respectively (Table 1). Of the 10 mutant lines, two lines were excluded from the analysis because they did not grow normally and few seeds were harvested, thereby rendering it difficult to determine their exact phenotypes for Cs accumulation. Five mutant lines, namely EnT-2377, EnT-3461, EnT-2953, EnT-2267, and EnT-2080, exhibited significantly reduced seed radiocesium (¹³⁷Cs) concentrations compared with the original variety ‘Enrei’ (Fig. 2A). Three mutant lines, namely EnT-3339, EnT-3308, and EnT-3882 showed no significant difference from ‘Enrei’. In addition, the ¹³³Cs concentrations were also measured in these eight mutant lines and ‘Enrei’. It was found that the seed ¹³⁷Cs concentrations of the mutant lines exhibited high positive correlations with their seed ¹³³Cs concentrations, indicating that the use of ¹³³Cs rather than ¹³⁷Cs was acceptable for phenotypic screening (Fig. 2C). Among the homologous elements, the seed K concentration did not differ obviously between the mutant lines and ‘Enrei’ (Supplemental Fig. 1). However, the concentration of the homologous element Na was significantly higher in the EnT-2953 line compared to ‘Enrei’ (Fig. 2B), suggesting that the gene related to the sodium transport system may have undergone mutation.

Fig. 1.

Frequency distribution of the seed ¹³³Cs concentration in the soybean mutant population (905 lines). Seed ¹³³Cs concentrations are shown relative to the average values of the population. The red blocks indicate the 10 lines selected as low-Cs candidates.

Fig. 2.

Seed Cs and Na concentrations in the low-Cs candidate mutant lines grown in Field A. The top 10 mutant lines with reduced seed ¹³³Cs concentrations (Fig. 1) were grown in Field A (2018). The mutant lines and the original variety ‘Enrei’ were cultivated twice and ten times in five individuals, respectively. Two of the ten mutant lines were excluded from analysis because they did not grow normally and the number of harvested seed was small. Seed concentrations of (A) ¹³⁷Cs and (B) Na. (C) Relationship between the seed concentrations of ¹³³Cs and ¹³⁷Cs. For the ¹³⁷Cs concentration, the data shown in (A) are the means ± SD of the harvested individuals. For the Na and ¹³³Cs concentrations, the same number of seeds from the harvested individuals was mixed in each replicate and subjected to ICP–MS analysis, and the average value of the two replicates is shown. The numbers in parentheses below the line names indicate the ranking of the seed ¹³³Cs concentration, as shown in Fig. 1. Bars with the same letter of the same case do not differ significantly from one another (Tukey–Kramer multiple comparison test).

Identification of the gmsos1 mutant lines

In rice, a mutant of OsSOS2, which encodes a serine/threonine-protein kinase required for the SOS pathway, has been reported to contain reduced radiocesium concentration in the grains (Ishikawa et al. 2017). Given that the ossos2 mutant exhibited lower concentrations of Cs and higher concentrations of Na in the roots compared to the original variety, it was hypothesized that the EnT-2953 line may possess an aberrant SOS pathway involving SOS1, SOS2, and SOS3. A BLASTP search showed that soybean contains at least two homologs (Glyma.13G166100 and Glyma.17G113700) of AtSOS2 (AT5G35410) and two homologs (Glyma.04G235900 and Glyma.06G128700) of AtSOS3 (AT5G24270). However, only one homolog (i.e., Glyma.08G092000) was detected for AtSOS1 (AT2G01980). This gene has previously been named as GmSOS1 (Zhang et al. 2022), and it shares 64 and 61% amino acid sequence identities with AtSOS1 and OsSOS1 (Os12g0641100), respectively. RNA-sequencing expression data for nine soybean tissues, retrieved from the Phytozome database, revealed that GmSOS1 is primarily expressed in the roots, suggesting that it may be involved in Na extraction in the roots (Supplemental Fig. 2). Considering that a mutated gene with one copy is more likely to influence the phenotype than a mutated gene with multiple copies, it was hypothesized that GmSOS1 could be the causative gene for low Cs accumulation. The coding sequence of GmSOS1 was subsequently confirmed using the genomic DNA of EnT-2953. A nucleotide G to A substitution at the 3ʹ-splite site of intron 5 was detected in the EnT-2953 line (Fig. 3A). Since EnT-2953 constitutes a homozygous mutant in the M2ʹ plants, it was considered a homozygous mutant in the M3ʹ generation used for phenotypic screening. The splicing pattern of EnT-2953 was confirmed using mRNA from the roots, and it was found that the sixth exon, which measures 102 bp and encodes 34 amino acids that form part of the Na⁺/H⁺ exchanger domain, was deleted in the transcripts (Fig. 3B).

Fig. 3.

Gene structure of Glyma.08G092000 as GmSOS1. (A) Schematic structure of Glyma.08G092000. Open and filled boxes indicate untranslated regions and exons, respectively. The mutation sites in EnT-2953, EnT-3153, and EnT-3612 are shown. (B) Structures of the Glyma.08G092000 amino acids. The gray and dotted boxes indicate the Na⁺/H⁺ exchanger motif and the cyclic nucleotide binding motif, respectively. The red boxes and lines indicate deleted regions and amino acid substitutions in the mutant lines, respectively.

Since the mutant population used in this study is thought to possess dense mutations throughout the entire genome (Tsuda et al. 2015), it is possible that genes other than GmSOS1 are responsible for the reduced Cs concentration in EnT-2953; hence, additional GmSOS1 mutant lines were searched in the same mutant library used for the phenotypic screening. Of the 1,536 mutant lines, two mutant lines (i.e., EnT-3153 and EnT-3612), which are likely to be functionally defective in GmSOS1, were identified. As shown in Fig. 3A, EnT-3153 and EnT-3612 carry a premature stop codon within the coding sequence of Glyma.08G092000 (GmSOS1). In EnT-3153, a nucleotide substitution of T to A causes the replacement of leucine with a premature stop codon at amino acid position 940. In EnT-3612, a G to A nucleotide substitution causes the replacement of tryptophan with a premature stop codon at amino acid position 786. EnT-3612 also exhibited a G to A nucleotide substitution, resulting in the replacement of valine with isoleucine at amino acid position 399. Both mutant lines demonstrated a heterozygous mutant genotype in M2ʹ plants, and EnT-3153 was not included in the 905 lines subjected to phenotypic screening. It was confirmed that the three gmsos1 mutants, namely EnT-2953, EnT-3153, and EnT-3612, possessed no mutations in the coding sequences of the AtSOS2 (Glyma.13G166100 and Glyma.17G113700) and AtSOS3 (Glyma.04G235900 and Glyma.06G128700) homologues.

Seed ¹³⁷Cs accumulation in the gmsos1 mutants

The exchangeable K content of Field A from 2018 to 2020 was 53–65 mg/kg (Table 1). In 2018, F₂ plants derived from the cross between ‘Enrei’ and EnT-2953 were planted. After sowing, the genotype of each individual was confirmed using the dCAPS marker. The seed ¹³⁷Cs concentrations of plants with the homozygous gmsos1 mutant genotype were significantly lower than those of plants with the wild-type and heterozygous genotypes (Fig. 4A). In 2020, F₂ plants derived from the crosses between ‘Enrei’ and either EnT-3153 or EnT-3612 were planted. Similarly, the seed ¹³⁷Cs concentration in the F₂ plants with homozygous mutant genotypes was significantly reduced (Fig. 4B, 4C). To evaluate the phenotypic stability for Cs accumulation in subsequent generations, the F₃ and F₄ progenies with homozygous mutant and wild-type genotypes, which were derived from crosses between ‘Enrei’ and EnT-2953, were planted in 2019 and 2020, respectively. The mutant genotype exhibited a significantly lower seed ¹³⁷Cs concentration than the wild-type genotype in both generations (Fig. 5). These results suggest that the mutations that occurred in the EnT-2953, EnT-3153, and EnT-3612 lines are recessive; moreover, homozygous gmsos1 mutant individuals possess lower Cs concentrations under low-K conditions.

Fig. 4.

Seed ¹³⁷Cs concentrations in three F₂ populations between the gmsos1 mutant and the ‘Enrei’ grown in Field A. (A) F₂ populations between the EnT-2953 and ‘Enrei’ samples grown in 2018. (B) F₂ populations between the EnT-3153 and ‘Enrei’ samples grown in 2020. (C) F₂ populations between the EnT-3612 and ‘Enrei’ samples grown in 2020. The F₂ plants were classified according to the GmSOS1 genotype as follows: Wild type (W), heterozygous (H), mutant type (M). Data are reported as means ± SD. Bars with the same letter of the same case do not differ significantly from one another (Tukey–Kramer multiple comparison test).

Fig. 5.

Seed ¹³⁷Cs concentrations in the progeny lines from the F₂ plants between the EnT-2953 and ‘Enrei’ samples in Field A. (A) F₃ lines grown in 2019. (B) F₄ lines grown in 2020. The F₃ and F₄ lines were bred through the self-fertilization of F₂ individuals in which the GmSOS1 genotype was confirmed. The GmSOS1 genotypes were defined as follows: Wild type (W) and mutant type (M). Data are reported as means ± SD. Bars with the same letter of the same case do not differ significantly from one another (Tukey–Kramer multiple comparison test).

Agronomic traits in the gmsos1 mutants grown in two field with different characteristics

To investigate the phenotypic stability of Cs accumulation under different cultivation conditions, F₄ lines of EnT-2953 were planted in Field B in addition to Field A. Since 100 kg/ha of K₂O fertilizer was applied in Field B as a basal fertilizer, the exchangeable K concentration in Field B was 366 mg/kg, which is higher than the recommended value (>208 mg/kg) for suppressing Cs absorption (Table 1). The exchangeable concentrations of Cs and Na in Field B were lower than those in Field A. Owing to the high exchangeable K concentration and the low ¹³⁷Cs concentration (<500 Bq/kg) in the soil from Field B, the seed ¹³⁷Cs concentration was unmeasurable. Thus, ¹³³Cs was measured instead of ¹³⁷Cs, and the seed ¹³³Cs concentrations were compared between Fields A and B. In Field B, the F₄ plants with the mutant genotype demonstrated significantly reduced seed ¹³³Cs concentrations than the ‘Enrei’ plants and the F₄ plants with the wild-type genotype (Fig. 6A). In Field A, there was no significant difference in ¹³³Cs concentrations between the mutant and wild-type lines at the 5% level. However, the mutant line showed lower ¹³³Cs concentrations than the wild-type line. In addition, the mutant and wild-type lines significantly differed in terms of their ¹³⁷Cs concentrations (Fig. 5B). Meanwhile, the mutant lines had a significantly higher seed Na concentration than the wild-type line in both fields (Fig. 6B), and the seed K concentration was not altered by the GmSOS1 genotype in either field (Fig. 6C). A comparative cultivation test was performed in Fields A and B using the F₂ population of EnT-3153, and similar results were obtained for the ¹³³Cs, ²³Na, and ³⁹K concentrations (Supplemental Fig. 3). These results suggest that gmsos1 reduces Cs accumulation and increases Na accumulation in soybeans regardless of the exchangeable K content and soil type.

Fig. 6.

Seed concentrations of ¹³³Cs, K, and Na in the F₄ lines between the EnT-2953 and ‘Enrei’ samples in Fields A and B in 2020. (A) Seed ¹³³Cs concentration. (B) Seed Na concentration. (C) Seed K concentration. The left and right sides show the data corresponding to Fields A and B, respectively. The F₄ lines were bred through the self-fertilization of F₂ individuals in which the GmSOS1 genotype was confirmed. The GmSOS1 genotypes were defined as follows: Wild type (W) and mutant type (M). Data are reported as means ± SD. Bars with the same letter of the same case do not differ significantly from one another (Tukey–Kramer multiple comparison test).

To confirm the effect of gmsos1 on agronomic traits, the 100-seed weight and seed weight per plant were also determined. In Field A, reliable seed weight data were not obtained in 2020 due to a lack of rainfall over an extended period, which resulted in a poor germination uniformity after sowing. Therefore, a preliminary comparison was performed in the EnT-2953 line using the data obtained from the F₂ and F₃ individuals cultivated in Field A, and the F₄ individuals cultivated in Field B. No differences in the 100-seed weight were observed between the mutant and wild types (Enrei and wild-type F₂, F₃, or F₄ individuals) (Fig. 7A). However, individuals with the gmsos1 genotype cultivated in Field A exhibited a significantly lower seed weight per plant than the wild type, whereas no significant differences were observed in the seed weights between the mutant and wild types cultivated in Field B (Fig. 7B).

Fig. 7.

100-Seed weights and seed weights of the F₂ plants between the EnT-2953 and ‘Enrei’ samples and its progeny lines grown in Fields A and B. (A) 100-Seed weight. (B) Seed weight per plant. The left, middle, and right regions show the data for the F₂ plants in Field A, the F₃ lines in Field A, and the F₄ lines in Field B, respectively. The F₃ and F₄ lines were bred through the self-fertilization of F₂ individuals in which the GmSOS1 genotype was confirmed. The GmSOS1 genotypes were defined as follows: Wild type (W) and mutant type (M). Data are reported as means ± SD. Bars with the same letter of the same case do not differ significantly from one another (Tukey–Kramer multiple comparison test).

Discussion

Soybean is a major crop grown worldwide. To ensure food safety, it is necessary to develop soybean varieties with low radiocesium accumulation and to elucidate the mechanism behind Cs accumulation in soybeans. In this study, candidate low-Cs-accumulating lines were searched in the mutant soybean population. Consequently, five low-Cs-accumulating lines were successfully identified based on phenotypic screening, and it was found that one of these lines was caused by a mutation in the GmSOS1 gene, which encodes an Na⁺/H⁺ antiporter in the SOS pathway. To the best of our knowledge, there are no reports that SOS1 is involved in plant Cs accumulation; however, it has been reported that the rice SOS2 mutant (ossos2) contains reduced Cs concentrations in its grains and shoots (Ishikawa et al. 2017). Such a reduction indicates that the SOS pathway may also be involved in Cs accumulation in soybeans. In ossos2 mutants of rice, Cs⁺ uptake by the roots was significantly decreased under low K⁺/Na⁺ conditions (Ishikawa et al. 2017). Because K⁺ transporters are suggested to mediate Cs⁺ uptake in plants (White and Broadley 2000, Zhu and Smolders 2000), the expression of four gene families (i.e., OsHAK, OsAKT, OsHKT, and OsCNGC) involved in K⁺ transport was considered. It was found that the transcript levels of several K⁺ and Na⁺ transporter genes, including OsHAK1, OsHAK5, OsAKT1, and OsHKT2, were significantly downregulated under low K⁺/Na⁺ conditions. This indicated that the reduced Cs⁺ uptake in ossos2 mutants may be caused by the suppressed expression of these genes through K⁺/Na⁺ imbalance in the root, which in turn can be attributed to the ossos2 mutation (Ishikawa et al. 2017). Based on homology searches of the soybean genome, soybean reportedly contains 70 putative K⁺ transporter genes (Rehman et al. 2017). However, very few studies have investigated the K⁺ transporters present in soybeans. Consequently, the mechanism responsible for reducing the seed Cs concentration in the gmsos1 mutant is unknown at this time; nonetheless, it may be attributed to changes in the expression of K⁺ transporters, as in the case of ossos2.

In all field cultivation experiments involving siblings from the cross performed between the gmsos1 mutants and ‘Enrei’, the individuals possessing the homozygous mutant genotype contained lower seed Cs concentrations than those with the homozygous wild-type genotype (Figs. 2A, 4, 5, 6A, Supplemental Fig. 3A). Although the degree of decline in the Cs concentration varied depending on the cultivation field and year, the Cs concentration was consistently lower than that of the wild type, ranging from 45.1 to 83.8%. Previously, Takagi et al. (2015) investigated a set of mini-core collections and identified several soybean germplasm accessions with low Cs accumulation. However, the position, number, and effect of genes controlling low Cs accumulation in these soybean germplasms are still unclear, which limits their potential use as breeding materials. In contrast, the low-Cs-accumulating mutants identified in our study originated from an elite soybean variety, which has favorable agronomic traits. Additionally, DNA markers associated with genes controlling low-Cs-accumulating mutants could be easily developed and applied in breeding practice. These factors make the mutants identified our study promising breeding materials for developing soybean varieties with low Cs accumulation.

In the current work, it was found that the gmsos1 mutants contained higher seed Na concentrations than those with the homozygous wild-type genotype (Figs. 2B, 6B, Supplemental Fig. 3B). In contrast, the seed concentrations of K, an element homologous to Cs and Na, were similar between the mutant and wild-type plants (Fig. 6C, Supplemental Figs. 1, 3C). Notably, K is an essential element for plants, and since the seed K concentration in the gmsos1 mutant was unaffected, it can be considered a useful breeding material. Furthermore, the gmsos1 mutants showed no significant differences in their 100-seed weights compared to the wild type (Fig. 7A), and no obvious abnormalities were observed in seed germination or growth (data not shown). On the other hand, the mutants demonstrated significantly lower seed weights per plant than the wild-type plants under the low-K⁺ conditions of Field A (Fig. 7B). Although further cultivation tests are necessary to address this concern, it was considered that the yield may decrease in the mutants under K-deficient conditions because the seeds attempt to accumulate a certain amount of K despite the reduced amount of K uptake. In addition, considering that the loss-of-function gmsos1 soybean created using the CRISPR-Cas9 system displayed an exceptionally increased salt sensitivity (Zhang et al. 2022), cultivation tests should be conducted in fields with high salt concentrations. Given that the soybean germplasm exhibits a diversity of salt tolerance phenotypes (Phang et al. 2008), the gmsos1 mutation may reduce the salt tolerance, although its effect may vary depending on the genetic background. Several genes that play significant roles in soybean salt tolerance have been identified based on the quantitative traits loci (QTL) and genome-wide association studies using the soybean germplasm (Feng et al. 2023), and it may be possible to enhance the salt tolerance of the gmsos1 mutant by utilizing these genes.

Based on the seed Cs concentration results obtained by phenotypic screening, four low-Cs-accumulating lines were identified in addition to the gmsos1 mutant lines. More specifically, in addition to the gmsos1 mutant EnT-2953, four lines stably showed significantly lower ¹³⁷Cs concentrations than ‘Enrei’ (Fig. 2A). Among them, EnT-2267 had a lower ¹³⁷Cs concentration than EnT-2953. In contrast to EnT-2953, this mutant line did not exhibit a high Na concentration (Fig. 2B), and so its causative gene was thought to differ from GmSOS1, potential rendering it an additional breeding material. In the future, it will be necessary to develop a segregating population by crossing with the original variety, evaluating the resulting phenotypic stability, and considering its usefulness as a breeding material. Moreover, it may be possible to identify additional mutants with low Cs accumulation levels using information obtained from other plant species. Previous analyses have shown that several genes, such as the high affinity K⁺ transporter (HAK) family and the cyclic nucleotide gated channels (CNGC), are involved in Cs accumulation in Arabidopsis and rice (Hampton et al. 2005, Qi et al. 2008, Rai et al. 2017). Mutants of the homologous of these genes may also lead to reduced Cs accumulation in soybeans. The mutant library used in this study has been applied not only to select mutants with phenotypes of interest through forward screening, but also to isolate specific gene mutants for functional analysis based on reverse genetics (Takagi et al. 2018, Yano et al. 2017, 2018). The soybean is of paleotetraploid origin, and as in the cases of SOS2 and SOS3, multiple homologue candidate genes may also exist in HAK and CNGC. By accumulating mutations of each candidate gene through reverse genetics, it may be possible to obtain low cesium accumulation lines. Further identification of low-Cs mutants, such as EnT-2267, and their causative genes will be expected to lead to the development of ‘safer’ soybean varieties.

Author Contribution Statement

K.T. and M.I. conceived and designed the research. K.T., Y.Y., Y.F., S.H., H.M., K.K., A.K., and T.A. conducted the experiments and prepared samples. K.T. and M.I. wrote the manuscript. All authors contributed to the development of this manuscript.

 Acknowledgments

We are grateful to the staff of the Tohoku Agricultural Research Center, National Agriculture and Food Research Organization, for their field and laboratory assistance. These staff include Ms. Aya Miura, Ms. Etsuko Shibayama, Ms. Yurie Yoshida, Mr. Masakatsu Ito, Mr. Takao Sakurai, Mr. Masanori Yoshida, Mr. Rikio Shishido, Mr. Takahiro Oyama, and Mr. Takao Ota. This work was supported by Cross-ministerial Strategic Innovation Promotion Program (SIP), “Building a Resilient and Nourishing Food Supply Chain Management for a Sustainable Future” (Grant Number JPJ012287; funding agency: Bio-oriented Technology Research Advancement Institution).

Literature Cited

•  Anai,  T. (2012) Potential of a mutant-based reverse genetic approach for functional genomics and molecular breeding in soybean. Breed Sci 61: 462–467.
•  Avery,  S.V. (1995) Caesium accumulation by microorganisms: uptake mechanisms, cation competition, compartmentalization and toxicity. J Ind Microbiol 14: 76–84.
•  Cooper,  J.L.,  B.J.  Till,  R.G.  Laport,  M.C.  Darlow,  J.M.  Kleffner,  A.  Jamai,  T.  El-Mellouki,  S.  Liu,  R.  Ritchie,  N.  Nielsen et al. (2008) TILLING to detect induced mutations in soybean. BMC Plant Biol 8: 9.
•  Feng,  C.,  H.  Gao,  Y.  Zhou,  Y.  Jing,  S.  Li,  Z.  Yan,  K.  Xu,  F.  Zhou,  W.  Zhang,  X.  Yang et al. (2023) Unfolding molecular switches for salt stress resilience in soybean: recent advances and prospects for salt-tolerant smart plant production. Front Plant Sci 14: 1162014.
•  Hampton,  C.R.,  M.R.  Broadley and  P.J.  White (2005) Short review: the mechanisms of radiocaesium uptake by Arabidopsis roots. Nukleonika 50: S3–S8.
•  Hashimoto,  K.,  T.  Nagasawa,  S.  Murakami,  I.  Watanabe,  K.  Kokubun,  S.  Sakai,  K.  Igita and  A.  Okabe (1988) A new soybean variety “Kosuzu”. Bulletin of the Tohoku National Agricultural Experiment Station 77: 45–61.
•  Hirayama,  T.,  M.  Takeuchi,  H.  Nakayama and  N.  Nihei (2018) Effects of decreasing radiocesium transfer from the soil to soybean plants and changing the seed nutrient composition by the increased application of potassium fertilizer. Bulletin of the Fukushima Agricultural Technology Centre 9: 1–10.
•  Ishikawa,  J.,  S.  Fujimura,  M.  Kondo,  M.  Murai-Hatano,  A.  Goto and  T.  Shinano (2018) Dynamic changes in the Cs distribution throughout rice plants during the ripening period, and effects of the soil-K level. Plant Soil 429: 503–518.
•  Ishikawa,  S.,  S.  Hayashi,  T.  Abe,  M.  Igura,  M.  Kuramata,  H.  Tanikawa,  M.  Iino,  T.  Saito,  Y.  Ono,  T.  Ishikawa et al. (2017) Low-cesium rice: mutation in OsSOS2 reduces radiocesium in rice grains. Sci Rep 7: 2432.
•  Ji,  H.,  J.M.  Pardo,  G.  Batelli,  M.J.  Van Oosten,  R.A.  Bressan and  X.  Li (2013) The Salt Overly Sensitive (SOS) pathway: established and emerging roles. Mol Plant 6: 275–286.
•  Kaga,  A.,  T.  Shimizu,  S.  Watanabe,  Y.  Tsubokura,  Y.  Katayose,  K.  Harada,  D.A.  Vaughan and  N.  Tomooka (2012) Evaluation of soybean germplasm conserved in NIAS genebank and development of mini core collections. Breed Sci 61: 566–592.
•  Kato,  N.,  N.  Kihou,  S.  Fujimura,  M.  Ikeba,  N.  Miyazaki,  Y.  Saito,  T.  Eguchi and  S.  Itoh (2015) Potassium fertilizer and other materials as countermeasures to reduce radiocesium levels in rice: Results of urgent experiments in 2011 responding to the Fukushima Daiichi Nuclear Power Plant accident. Soil Sci Plant Nutr 61: 179–190.
•  Krishnamurthy,  P.,  Y.  Fujisawa,  Y.  Takahashi,  H.  Abe,  K.  Yamane,  K.  Mukaiyama,  H.R.  Son,  S.  Hiraga,  A.  Kaga,  T.  Anai et al. (2019) High-throughput screening and characterization of a high-density soybean mutant library elucidate the biosynthesis pathway of triterpenoid saponins. Plant Cell Physiol 60: 1082–1097.
•  Li,  Z.,  L.  Jiang,  Y.  Ma,  Z.  Wei,  H.  Hong,  Z.  Liu,  J.  Lei,  Y.  Liu,  R.  Guan,  Y.  Guo et al. (2017) Development and utilization of a new chemically-induced soybean library with a high mutation density. J Integr Plant Biol 59: 60–74.
• MAFF, NARO and NIAES (2015) Factors and countermeasure of soybean with high radioactive cesium concentration. Investigation of factors and result of trial examination. https://www.maff.go.jp/j/kanbo/joho/saigai/pdf/youin_daizu_3.pdf (in Japanese).
•  Matsunami,  H.,  T.  Uchida,  H.  Kobayashi,  T.  Ota and  T.  Shinano (2021) Comparative dynamics of potassium and radiocesium in soybean with different potassium application levels. J Environ Radioact 233: 106609.
•  Oladosu,  Y.,  M.Y.  Rafii,  N.  Abdullah,  G.  Hussin,  A.  Ramli,  H.A.  Rahim,  G.  Miah and  M.  Usman (2016) Principle and application of plant mutagenesis in crop improvement: a review. Biotechnol Biotechnol Equip 30: 1–16.
•  Phang,  T.H.,  G.  Shao and  H.M.  Lam (2008) Salt tolerance in soybean. J Integr Plant Biol 50: 1196–1212.
•  Qi,  Z.,  C.R.  Hampton,  R.  Shin,  B.J.  Barkla,  P.J.  White and  D.P.  Schachtman (2008) The high affinity K⁺ transporter AtHAK5 plays a physiological role in planta at very low K⁺ concentrations and provides a caesium uptake pathway in Arabidopsis. J Exp Bot 59: 595–607.
•  Rai,  H.,  S.  Yokoyama,  N.  Satoh-Nagasawa,  J.  Furukawa,  T.  Nomi,  Y.  Ito,  S.  Fujimura,  H.  Takahashi,  R.  Suzuki,  E.  Yousra et al. (2017) Cesium uptake by rice roots largely depends upon a single gene, HAK1, which encodes a potassium transporter. Plant Cell Physiol 58: 1486–1493.
•  Rehman,  H.M.,  M.A.  Nawaz,  Z.H.  Shah,  I.  Daur,  S.  Khatoon,  S.H.  Yang and  G.  Chung (2017) In-depth genomic and transcriptomic analysis of five K⁺ transporter gene families in soybean confirm their differential expression for nodulation. Front Plant Sci 8: 804.
•  Sikora,  P.,  A.  Chawade,  M.  Larsson,  J.  Olsson and  O.  Olsson (2011) Mutagenesis as a tool in plant genetics, functional genomics, and breeding. Int J Plant Genomics 2011: 314829.
•  Takagi,  K.,  A.  Kaga,  M.  Ishimoto,  M.  Hajika and  T.  Matsunaga (2015) Diversity of seed cesium accumulation in soybean mini-core collections. Breed Sci 65: 372–380.
•  Takagi,  K.,  R.  Yano,  S.  Tochigi,  Y.  Fujisawa,  H.  Tsuchinaga,  Y.  Takahashi,  Y.  Takada,  A.  Kaga,  T.  Anai,  C.  Tsukamoto et al. (2018) Genetic and functional characterization of Sg-4 glycosyltransferase involved in the formation of sugar chain structure at the C-3 position of soybean saponins. Phytochemistry 156: 96–105.
•  Tsuda,  M.,  A.  Kaga,  T.  Anai,  T.  Shimizu,  T.  Sayama,  K.  Takagi,  K.  Machita,  S.  Watanabe,  M.  Nishimura,  N.  Yamada et al. (2015) Construction of a high-density mutant library in soybean and development of a mutant retrieval method using amplicon sequencing. BMC Genomics 16: 1014.
•  Uda,  S.,  I.  Fukuhara,  A.R.M.  Daniela,  K.  Kojima,  S.  Djedidi,  S.I.  Agake,  A.  Kaga,  H.  Matsunami,  H.  Moriyama,  T.  Yamada et al. (2021) Identification and analysis of genes causing varietal differences of cesium accumulation in soybean: Soybean genes for varietal Cs difference. Soil Sci Plant Nutr 67: 670–680.
•  White,  P.J. and  M.R.  Broadley (2000) Mechanisms of caesium uptake by plants. New Phytol 147: 241–256.
•  Yano,  R.,  K.  Takagi,  Y.  Takada,  K.  Mukaiyama,  C.  Tsukamoto,  T.  Sayama,  A.  Kaga,  T.  Anai,  S.  Sawai,  K.  Ohyama et al. (2017) Metabolic switching of astringent and beneficial triterpenoid saponins in soybean is achieved by a loss-of-function mutation in cytochrome P450 72A69. Plant J 89: 527–539.
•  Yano,  R.,  K.  Takagi,  S.  Tochigi,  Y.  Fujisawa,  Y.  Nomura,  H.  Tsuchinaga,  Y.  Takahashi,  Y.  Takada,  A.  Kaga,  T.  Anai et al. (2018) Isolation and characterization of the soybean Sg-3 gene that is involved in genetic variation in sugar chain composition at the C-3 position in soyasaponins. Plant Cell Physiol 59: 797–810.
•  Zhang,  M.,  J.  Cao,  T.  Zhang,  T.  Xu,  L.  Yang,  X.  Li,  F.  Ji,  Y.  Gao,  S.  Ali,  Q.  Zhang et al. (2022) A putative plasma membrane Na⁺/H⁺ antiporter GmSOS1 is critical for salt stress tolerance in Glycine max. Front Plant Sci 13: 870695.
•  Zhu,  Y.G. and  E.  Smolders (2000) Plant uptake of radiocaesium: A review of mechanisms, regulation and application. J Exp Bot 51: 1635–1645.