DNA marker-assisted evaluation of cooked bean hardness of three soybean progeny lines

Abstract

Cooked bean hardness is an important trait for the processing of soybean products such as nimame, natto, miso, and soy sauce. Previously, we showed that cooked bean hardness is primarily affected by the pectin methylesterase gene Glyma03g03360, and that calcium content has a secondary effect on this trait. To establish a simple and timely method for the evaluation of cooked bean hardness, primers of amplification refractory mutation system-polymerase chain reaction (ARMS-PCR) were designed to detect a single-nucleotide polymorphism of Glyma03g03360 and subsequently used to evaluate three soybean progeny lines. The determined genotypes were compared to those identified using the cleaved amplified polymorphic sequence (CAPS) method. Seven out of 284 lines presented different genotypes, which were determined using the two methods: A genotypes were incorrectly assigned as heterozygous by CAPS, suggesting that ARMS-PCR is more reliable. Glyma03g03360 genotypes could be used to evaluate cooked bean hardness, except for intermediate values. Cooked bean hardness within the same genotype groups was significantly correlated with calcium contents. These findings indicate that ARMS-PCR is useful for a marker-assisted selection of soybean with soft-cooked beans and that calcium content may be used for additional selection.

Introduction

Soybean [Glycine max (L). Merr] is a major crop worldwide used in oil production, as a feed grain, and in processed foods. Soyfood has been consumed in Asian countries for over 1000 years and is becoming increasingly popular in other countries. In Japan, about 1 million tons of soybean, of which approximately 0.2 million tons are domestic, are used for food; about 0.5 million tons are used for tofu, and about 0.3 million tons are used for cooked beans (nimame), fermented steamed beans (natto), fermented steamed bean paste (miso), and as a soy source. Most domestic soybeans are used for these processed foods. Soybeans are first cooked and softened during the production of nimame, natto, miso, and soy sauce. Soft-cooked beans obtained with a shorter cooking duration are preferable for those four processed foods (Taira 1990, Yoshioka et al. 2009, Zhang et al. 2008). Harder beans require longer and stronger heating conditions, resulting in a darker color and unfavorable tastes (Mori and Taya 2008, Takemura 2001). Hard beans also result in an undesirable texture and an unfavorable ammoniac flavor in natto (Taira 1990). Therefore, the hardness of cooked beans is a main target of soybean breeding in Japan. Since measuring cooked bean hardness is both time and labor intensive, a simple evaluation method is required by breeders.

Previously, we demonstrated that the hardness of cooked soybeans is affected by genetic factors, and Glyma03g03360 polymorphism, which was recently renamed as Glyma.03g028900, is associated with the cooked bean hardness (Hirata et al. 2014, Toda et al. 2015). Glyma03g03360 encodes a pectin methylesterase (PME) of 490 amino acids in length (Toda et al. 2015, SoyBase, http://soybase.org). Pectin, which is a major component of primary cell walls and the middle lamella, is mainly constituted of linear chains of 1,4-linked alpha-D-galacturonic acid units. The carboxyl groups of the galacturonic acid units are partially esterified with methanol. During heating, the middle lamella pectin of the cell walls is depolymerized by β-elimination of methyl-esterified polygalacturonic acids, promoting an increase in cell separation, which softens the tissues (Pirhayati et al. 2011). PME converts the methoxyl groups into carboxyl groups on the polygalacturonic acid chain and releases both methanol and protons. β-elimination of pectins is primarily responsible for heat degradation at pH 6.1, and the rate of the β-eliminative cleavage of prepared pectin is dependent on the degree of esterification (Sajjaanantakul et al. 1989). Glyma03g03360 is expressed in the cotyledon and seed coat during the later stage of seed development (Toda et al. 2015), which probably decreases the degree of esterification in those tissues, resulting in harder cooked beans.

We identified four Glyma03g03360 polymorphisms, named as A, Aʹ, B, and Bʹ. Polymorphism Aʹ contains a single base substitution in exon 3, resulting in an amino acid substitution from threonine to serine; this amino acid is not predicted to be important for enzymatic activity. Polymorphism B contains a single-base substitution in exon 3, and polymorphism Bʹ contains a single-base deletion in exon 1 (Supplemental Figs. 1, 2). These polymorphisms are summarized in Supplemental Table 1. A stop codon is inserted in the B and Bʹ genotypes, resulting in amino-acid sequences lacking amino acids that are catalytically important (Bosch et al. 2005, Jimenez-Lopez et al. 2012). The cooked bean hardness of soybean carrying the A or Aʹ genotypes was found to be significantly higher than that of soybean carrying the B or Bʹ genotypes (Toda et al. 2015), which supports the suggested mechanism of softening tissues by the β-elimination of pectins.

DNA marker-assisted selection (MAS) has been successfully incorporated into soybean breeding programs to select diverse traits such as soybean pod dehiscence and maturity time (Suzuki et al. 2010, Yamada et al. 2012). Cooked bean hardness can also be evaluated by DNA markers, because our previous studies indicated a primary effect of Glyma03g03360 on this trait (Hirata et al. 2014, Toda et al. 2015).

Dozens of single nucleotide polymorphism (SNP)-genotyping techniques have been described (Isobe et al. 2016, Jungerius et al. 2003). A real-time polymerase chain reaction (PCR) using fluorescent probes, high-resolution melting (HRM) post PCR analysis, and digital PCR (dPCR), which are simple and time-saving, have become more popular, but they require expensive equipment and/or expensive reagents. Sequencing-based methods have recently become popular too, but their costs are still high. Hybridization-based assays for SNP genotyping have been developed as low-cost and high-throughput methods, but they hold the risk for cross-hybridization and unspecific bindings (Hacia 1999, Li et al. 2012). PCR-electrophoresis-based methods would be preferable for many researches and breeders who require robust and simple SNP genotyping methods with low costs. In the present study, to establish MAS for cooked bean hardness, two sets of primers were designed for amplification refractory mutation system-polymerase chain reaction (ARMS-PCR) to detect the B and Bʹ genotypes of Glyma03g03360, although the cleaved amplified polymorphic sequence (CAPS) method has been established to detect the B genotype (Toda et al. 2015). CAPS is a two-step method in which DNA is first amplified by PCR, and then PCR products are digested with restriction enzymes. In contrast, ARMS-PCR is a one-step method; it allows genotyping solely by inspection of reaction mixtures after electrophoresis (Newton et al. 1989). In addition, ARMS-PCR is more cost-effective than CAPS.

In the present study, ARMS-PCR was performed to genotype Glyma03g03360 and the correctness of genotyping and the relationship between the genotype and cooked bean hardness were considered. Since calcium content was positively correlated with cooked bean hardness (Saio and Watanabe 1972, Toda et al. 2015), the effects of calcium content were also considered. Our data are useful for soybean breeders and manufacturers who have an interest in cooked bean hardness and plan to evaluate this trait using a simple method.

Materials and Methods

Plant materials

The soybean recombinant inbred line (RIL) population in the F₈ generation derived from a cross between ‘Satonohohoemi’ and ‘Fukuibuki’, and parental cultivars were planted in fields at Ishinazaka (39°32ʹN, 140°23ʹE) in the Kariwano Branch of the Daisen Research Station of Tohoku Agricultural Research Center, NARO, on 23th May, 2018.

Two soybean progeny lines in the F₅ generation derived from crosses between ‘Satonohohoemi’ and ‘Sachiyutaka’, and those between ‘Karikei-920’ and ‘Karikei-898’, and parental cultivars were planted in the same fields on the same day. These two progeny lines have been selected for good agricultural traits such as pest resistance and plant shape, but not for cooked bean hardness.

Each plot consisted of a single row with row spacings of 0.75 m, and 0.12 m between plants within each row. Starter fertilizers included 24 kg ha^–1 N, 120 kg ha^–1 P₂O₅, 80 kg ha^–1 K₂O, and no compost. Soils of the experimental fields were andosol. The parental cultivars and each line were harvested in bulk.

20 lines of A and B types each were randomly selected for analyses from the RIL population derived from a cross between ‘Satonohohoemi’ and ‘Fukuibuki’. Six lines of the Aʹ type and five lines of the B type were used from the progeny lines derived from a cross between ‘Satonohohoemi’ and ‘Sachiyutaka’, and three lines of the A type and five lines of the B type were used from the progeny lines derived from a cross between ‘Karikei-920’ and ‘Karikei-898’ for analyses, because only progeny lines with good agricultural traits were harvested.

DNA isolation and marker analysis

Total DNA was extracted from leaves using a BioSprint 96 automatic DNA isolation system (QIAGEN) as described previously (Taguchi-Shiobara et al. 2019). RIL populations derived from a cross between ‘Satonohohoemi’ and ‘Fukuibuki’ were descended from a single F₅ plant and the leaves for DNA extraction were collected from the F₅ plant. For two progeny lines derived from crosses between ‘Satonohohoemi’ and ‘Sachiyutaka’, and between ‘Karikei-920’ and ‘Karikei-898’ in the F₅ generation, the leaves for DNA extraction were collected from three individuals out of 25 per line in the field in 2018 and bulked these into a single one.

Two sets of ARMS-PCRs and a set of CAPS primers are described in Table 1. ARMS-PCR was performed in a 10 μL reaction volume using GoTaq Green Master Mix according to the manufacturer’s instructions (Promega KK, Tokyo, Japan). Samples were amplified with an initial incubation for 1 min at 94°C, then 40 cycles of 30 s at 94°C, 30 s at 63°C, and 1 min at 72°C. PCR for CAPS was performed in a 10 μL reaction volume using LA Taq Hot Start Version according to the manufacturer’s instructions (TAKARA, Shiga, Japan). Amplifications involved an initial incubation for 2 min at 95°C, then 33 cycles of 1 min at 92°C, 1 min at 58°C, and 1 min at 68°C, followed by a final incubation for 5 min at 72°C. Next, 2.5 μL of PCR product was digested with 1 U of Bst1107 I (TAKARA, Shiga, Japan) at 37°C overnight. DNA fragments were separated using TAE agarose gel electrophoresis or the LabChip GX electrophoresis system (PerkinElmer, MA, US).

Table 1. Primers used for ARMS-PCR and CAPS

Primer sequence	Description	Purpose
AGCAGTGGCACTCAAGGTAAG	A forward outer primer	For detecting polymorphism B of Glyma03g03360 by ARMS-PCR
TACTTCCTCCAAGGCCTACCa	A reverse outer primer
CTACCGCGACTGCTATGTATAC	A forward primer having a mismatch for B
CAAAGATGAAATCAATGGTGCCC	A reverse primer having a mismatch for A, Aʹ, and Bʹ
GCCACACAATCCTAGTGCC	A forward outer primer	For detecting polymorphism Bʹ of Glyma03g03360 by ARMS-PCR
CCTTTCCCTTTGCCTTTATTATTCTTTG	A reverse outer primer
CAAACAAGTGCTACAATAGCCT	A forward primer having a mismatch for A, Aʹ, and B
TGGCACAATCACTAAGGGC	A reverse primer having a mismatch for Bʹ
TTTTCAAAAAGTAATCCAAGATTTTTC	A forward primer upstream of the SNP of B	For detecting polymorphism B of Glyma03g03360 by CAPS
TACTTCCTCCAAGGCCTACCa	A reverse primer downstream of the SNP of B

^a Same DNA sequence.

DNA sequencing

DNA fragments were amplified by PCR using primers designed for CAPS (Table 1) and genomic DNA as a template. PCR was performed under the same conditions as the CAPS method. For DNA sequence analysis, DNA fragments were purified with a MonoFas DNA purification kit I (GL Sciences Inc., Tokyo, Japan) following electrophoresis on agarose gel. Sequencing was performed by Eurofin Genomics Inc. (Tokyo, Japan).

Measurements of cooked bean hardness

Cooked beans were prepared, and cotyledon breaking stress was measured as described previously (Hirata et al. 2014, Toda et al. 2015). Briefly, 30 seeds were soaked in 100 mL of deionized water at 20°C for 22 ± 2 h. Imbibed seeds were boiled for 10 min in 200 mL of deionized water in a 500 mL tall glass beaker (HARIO Ltd., Tokyo, Japan) using a hot stirrer (IKA RCT basic, IKA Japan K.K., Osaka, Japan). After cooking, the seeds were cooled in water at room temperature and the water was drained. The seed coat and embryo axis were removed from the boiled seeds; then, the two cotyledons were separated. Puncture strength was measured using a Rheoner RE-3305S texture analyzer (Yamaden Corp., Tokyo, Japan) fitted with a cylindrical probe (3 mm outside diameter) and a 2 kg capacity load cell. The adaxial (flat) side of a piece of cotyledon was placed facing down on the analyzer stage, and the hardness was measured. The probe was operated at a distance of 2.45 mm with a speed of 1 mm/s. Puncture strength was determined as maximum force (pascals). The average score for more than 40 samples was used as an index of cooked bean hardness per sample.

The 100-seed weight and the water absorption ratio were calculated as described previously (Toda et al. 2015).

Calcium measurements

Seed calcium concentrations were determined by inductively coupled plasma-mass spectrometry (ICP-MS; Agilent 7700x, Agilent Technologies, Tokyo, Japan) according to a previously described method with slight modification (Takagi et al. 2015). More than 5 g of seeds were used for analysis. Seeds were milled into flour using a vibrating mill (TI-100, Cosmic Mechanical Technology Co. Ltd., Fukushima, Japan). Soybean flour was preliminarily dried at 70°C for 2 h. Then, 20 mg ground samples were transferred into 15 mL metal-free centrifuge tubes (INA·OPTIKA, Osaka, Japan) and digested in 0.3 mL concentrated nitric acid (1.38) at 95°C for 2 h. The digested volumes were then brought up to 10 mL with ultrapure water. ICP-MS was performed using 5 μg/L ¹¹⁵In as an internal standard. The average concentration of each sample from two measurements was converted into the amounts of mineral elements (mg kg^–1 dry weight) in each sample. Moisture contents were calculated by weighing flour samples before and after heating at 105°C for 24 h.

Statistical analysis

Statistical analyses were performed using the SPSS software package (version 22, IBM Japan, Tokyo, Japan). Data from parental lines, the RIL population, and the progeny lines were used (n = 65).

Results and Discussion

DNA fragment pattern obtained by ARMS-PCR

The ARMS-PCR primer sets used to detect the B and Bʹ polymorphisms of Glyma03g03360 are described in Table 1. Positions of primers for ARMS-PCR are explained in Supplemental Figs. 1 and 2. Polymorphism B-specific DNA fragments of 143 bp were detected using B-type cultivars, while larger DNA fragments of 252 bp were detected using A-, Aʹ-, or Bʹ-type cultivars by ARMS-PCR to detect polymorphism B (Fig. 1a). Polymorphism Bʹ-specific DNA fragments of 271 bp were detected using Bʹ-type cultivars, while larger DNA fragments of 353 bp were detected using A-, Aʹ- or B-type cultivars by ARMS-PCR to detect polymorphism Bʹ (Fig. 1b). The results of ARMS-PCRs using mixed genomic DNA from different cultivars indicated that these methods can detect heterozygous genotypes (Fig. 2).

Fig. 1.

Amplification refractory mutation system-polymerase chain reaction (ARMS-PCR) analysis of 21 soybean cultivars. (a) ARMS-PCR to detect polymorphism B of Glyma03g03360. An arrow indicates the common bands produced from the genomic DNA of all cultivars. An arrowhead indicates bands specific for A-, Aʹ-, and Bʹ-type cultivars. A double arrowhead indicates the band specific for B-type cultivars. (b) ARMS-PCR to detect polymorphism Bʹ of Glyma03g03360. An arrow indicates the common bands produced from the genomic DNA of all cultivars. An arrowhead indicates the bands specific for A-, Aʹ-, and B-type cultivars. A double arrowhead indicates the band specific for Bʹ-type cultivars. Information on lanes, cultivars and genotypes is shown on the right. M: DNA marker (ϕX174 DNA digested with HaeIII).

Fig. 2.

ARMS-PCR analysis using mixed genomic DNAs. (a) ARMS-PCR to detect polymorphism B of Glyma03g03360. An arrow indicates the common bands produced from the genomic DNA of all cultivars. An arrowhead indicates the bands specific for A-, Aʹ-, and Bʹ-type cultivars. A double arrowhead indicates the band specific for B-type cultivars. (b) ARMS-PCR to detect polymorphism Bʹ of Glyma03g03360. An arrow indicates the common bands produced from the genomic DNA of all cultivars. An arrowhead indicates the bands specific for A-, Aʹ-, and B-type cultivars. A double arrowhead indicates the band specific for Bʹ-type cultivars. Information on lanes, genomic DNA, and genotypes is shown on the right. A total of 25 ng DNA template was used for all lanes. The amount of DNA was calculated from the absorbance at 260 nm. M: DNA marker (ϕX174 DNA digested with HaeIII).

Genotyping soybeans using ARMS-PCR and CAPS

To verify the usefulness of ARMS-PCR, Glyma03g03360 genotype and cooked bean hardness were analyzed using the soybean RIL population and progeny lines described above. Genotypes of 284 lines were determined by ARMS-PCR and CAPS. As only Shuurei and Tamadaikoku were determined as Bʹ genotypes, no breeding lines derived from a Bʹ-type cultivar were obtained. Using ARMS-PCR, 113 lines were determined as A or Aʹ, 134 lines were determined as B, and 37 lines were determined as heterozygous. Seven of 284 lines presented different genotype results between ARMS-PCR and CAPS. These seven lines were found to carry polymorphism A by ARMS-PCR and as heterozygous by CAPS (Supplemental Table 2). These lines were found to carry polymorphism A by sequencing analysis. Using the CAPS method, PCR products from A- or Aʹ-type lines were digested by Bst1107 I; however, this restriction site was lost by a single base substitution in B-type lines. Seven polymorphism A lines were incorrectly identified as heterozygous due to the incomplete digestion of Bst1107 I in the CAPS method. It has been pointed out that incomplete enzyme digestion is a problem inherent to the CAPS for genotyping (Duraisingh et al. 1998, Johnson et al. 2004).

A prerequisite of ARMS-PCR is the absence of 3ʹ-exonucleolyric proofreading activity associated with the DNA polymerase (Newton et al. 1989). Another requirement in the application of ARMS-PCR is that 3ʹ-OH terminal mismatched primers are refractory to extension. Single mismatches have been reported to instigate a broad variety of effects, including those that have minor (eg, A–C, C–A, T–G, G–T) and severe (eg, A–A, G–A, A–G, C–C) impacts on PCR amplification (Stadhouders et al. 2010). The use of ARMS-PCR to detect polymorphism B of Glyma03g03360 showed high credibility, probably because it meets the conditions described above. C–C mismatches occur between A-, Aʹ-, or Bʹ-type genomes and the primer “CAAAGATGAAATCAATGGTGCCC”, and between the B-type genome and the primer “CTACCGCGACTGCTATGTATAC” at the 3ʹ end (Table 1, Supplemental Fig. 1).

Use of ARMS-PCR to detect polymorphism Bʹ of Glyma03g03360, results in a T–G mismatch between A-, Aʹ- or B-type genomes and the primer “CAAACAAGTGCTACAATAGCCT”, which has a minor effect. In contrast, G–G and A–C mismatches occur between the Bʹ-type genome and the primer “TGGCACAATCACTAAGGGC” at the two bases of the 3ʹ end (Table 1, Supplemental Fig. 2). A G–G mismatch has been reported to have a major refractory effect on PCR amplification (Stadhouders et al. 2010). In addition, the Tm value on the calculation is severely reduced due to a single base deletion between the Bʹ-type genome with the primer “TGGCACAATCACTAAGGGC”. For those reasons, it is expected that ARMS-PCR can be used to accurately detect polymorphism Bʹ.

Relationship between Glyma03g03360 genotype and cooked bean hardness, and the effect of calcium

Cooked bean hardness was examined using the soybean RIL population and progeny lines described above and their parental cultivars, ‘Satonohohoemi’, ‘Fukuibuki’, ‘Sachiyutaka’, ‘Karikei-920’, and ‘Karikei-898’. Calcium contents were also analyzed to investigate their effects on cooked bean hardness.

The hardness of cooked whole beans has shown a significant positive correlation with that of the cooked embryo (cotyledon and embryonic axis, r = 0.976, p < 0.001). A highly positive correlation was observed in the hardness of whole beans (r = 0.874, p = 0.010) and embryos (r = 0.832, p = 0.020) between materials cooked for 10 and 40 min (Yasui et al. 2014). Shorter cooking time is preferred because it allows reducing industrial processing costs (da Silva et al. 2009) and longer cooking time or higher cooking temperature result in a darker color and unfavorable tastes (Mori and Taya 2008, Takemura 2001). For those reasons, cotyledon breaking stress was measured after 10 min cooking for evaluation of cooked bean hardness.

The cooked bean hardness of RILs and progeny lines ranged from 682 to 1734 kPa (Supplemental Table 3). The cooked bean hardness of 26 of 29 A- or Aʹ-type lines was greater than 1100 kPa, and the cooked bean hardness of 26 of 30 B-type lines was less than 1100 kPa.

The independent sample t-test detected a significant difference between cooked bean hardness of A or Aʹ-type lines and that of B-type lines (p < 0.05). Correlation among Glyma03g03360 genotype, calcium content, 100-seed weight, water absorption ratio, and cooked bean hardness were analyzed. Motoki et al. (1999) reported that 100-seed weight and water absorption ratio showed a significant negative correlation with cooked bean hardness (r = –0.478 and –0.416, respectively). Hirota et al. (2005) also reported a similar result (r = –0.582 and –0.680, respectively). Glyma03g03360 genotype, calcium content, and 100-seed weight were significantly related with cooked bean hardness (Table 2). A correlation coefficient of the Glyma03g03360 genotype was negative, because A (or Aʹ) and B genotypes were assigned as 1 and 2, respectively, for analysis. The value –0.73 indicated that the contribution ratio of the Glyma03g03360 genotype to the cooked bean hardness was 0.52. It was close to the contributions of Glyma03g03360 using the RIL population derived from a cross between two Japanese soybean cultivars, ‘Natto-shoryu’ and ‘Hyoukei-kuro 3’ in 2010 and 2011, which was previously reported (Hirata et al. 2014).

Table 2. Correlation among Glyma03g03360 genotype, Ca content, 100-seed weight, water absorption ratio, and cooked bean hardness

	Ca content	100-seed weight	Water absorption ratio	Hardness of cooked bean
Genotype	–0.08	0.15	–0.03	–0.73**
Ca content	—	–0.48**	–0.03	0.40**
100-seed weight	—	—	–0.39**	–0.25*
Water absorption ratio	—	—	—	–0.05

n = 65. * significant at 0.05 level. ** significant at 0.01 level.

A multiple linear regression model of the cooked bean hardness was obtained using Glyma03g03360 genotype, calcium content, and 100-seed weight as three independent variables (Table 3). Collinearity was also checked using the tolerance values and the variance inflation factors (VIFs). Small tolerance values (much below 0.1) or large VIF values (above 10) indicate high collinearity (Hair et al. 1995). In the present study, tolerance values were more than 0.7 and VIF values were less than 2 (Table 3). Standardized partial regression coefficients of Glyma03g03360 genotype, calcium content, and 100-seed weight were –0.702, 0.350, and 0.025, respectively, indicating high contribution of the Glyma03g03360 genotype. Adjusted R-squared of the multiple linear regression was 0.62 (p < 0.001), which was not reduced when 100-seed weight was removed. We concluded that Glyma03g03360 genotype and calcium content mainly affect the cooked bean hardness among factors examined in this study.

Table 3. Summary of regression analysis on relationship between Glyma03g03360 genotype, calcium content, and 100-seed weight, and cooked bean hardness

Independent variable	Standardized coefficient beta	Tolerance value	VIF
Glyma03g03360 genotype	–0.702	0.977	1.024
Ca content	0.350	0.771	1.297
100-seed weight	0.025	0.758	1.319

n = 65.

Cooked bean hardness within the same genotype group was significantly correlated with calcium content (Fig. 3). The cooked bean hardness of A-type lines with low calcium content, such as SHxFK_060 and SHxFK_065, and the parental cultivar ‘Sachiyutaka’ was less than 1100 kPa, likely due to the low calcium content (Supplemental Table 3). The line K920xK898_14 presented a higher calcium content (2337 mg/kg), which may result in greater cooked bean hardness (1203 kPa), than the other B-type lines (Supplemental Table 3).

Fig. 3.

Correlation between calcium content and the cooked bean hardness of soybean lines with an A or Aʹ genotype (black squares), and a B genotype (white circles), including 6 parental data. Approximate lines, expressions for each group, correlation coefficients, and sample numbers are shown in the graph.

Calcium pretreatment has been shown to improve the firmness of thermally processed vegetables (Siliha et al. 1996, Smout et al. 2005, Stanley et al. 1995). Makabe (2006) reported that cooked seed hardness is positively correlated with the quantity of calcium ions in the cooking solution. Furthermore, endogenous calcium contents were also found to be positively correlated with cooked bean hardness (Saio and Watanabe 1972, Toda et al. 2015). Free carboxyl groups produced by PME activity on the pectin molecules can cross-link by calcium bridges, which provides structural rigidity to the cell wall (Goldberg et al. 1986, Holdaway-Clarke and Hepler 2003). Calcium ions bind to protein or phytate, which may also explain the effect of calcium on cooked bean hardness. The relationship between calcium content and cooked bean hardness obtained using soybeans with polymorphism B (Fig. 3) indicates that other PMEs may affect the cotyledon, or mechanisms other than calcium binding with pectin.

In conclusion, in this study, primers for ARMS-PCR were designed to detect B and Bʹ genotypes of Glyma03g03360. Comparison of the results obtained using ARMS- PCR, CAPS, and sequencing suggests that ARMS-PCR accurately detects B-type genotypes. Cooked bean hardness of RILs and progeny lines was assessed by genotyping Glyma03g03360. RILs or progeny lines with polymorphism A or Aʹ presented cooked bean hardness greater than 1000 kPa, while those with polymorphism B presented cooked bean hardness less than 1300 kPa. Overall, 88% (52 out of 59) of RILs or progeny lines were considered to be at the border of 1100 kPa by genotyping in this study. Previously, we showed that Glyma03g03360 significantly affected cooked bean hardness in many soybean cultivars and F₂ populations (Toda et al. 2015). The ARMS-PCR method established in the present study is expected to be useful for the selection of other breeding lines. Since calcium contents were positively correlated with cooked bean hardness, this may be used for additional selection.

Author Contribution Statement

KT, SK, KH, AK, and MH planned the research. SK and AK developed the breeding lines. KT and YN performed the experiments and analyzed the data. KT prepared the manuscript.

Acknowledgments

The authors are grateful to A. Hishinuma for the supply of soybeans, to Dr. F. Taguchi-Shiobara for the supply of genomic DNAs, and to Dr. J. Tanaka for advice on ARMS-PCR. They would like to thank Y. Tominaga for technical assistance with the experiments. They also would like to thank Editage (www.editage.com) for English language editing. This study was supported by JSPS KAKENHI Grant Numbers JP 19K05985.

Literature Cited

•  Bosch,  M.,  A.Y.  Cheung and  P.K.  Hepler (2005) Pectin methylesterase, a regulator of pollen tube growth. Plant Physiol. 138: 1334–1346.
•  da Silva,  J.B.,  M.C.  Carrão-Panizzi and  S.H.  Prudêncio (2009) Chemical and physical composition of grain‐type and food‐type soybean for food processing. Pesqui. Agropecu. Bras. 44: 777–784.
•  Duraisingh,  M.T.,  J.  Curtis and  D.C.  Warhurst (1998) Plasmodium falciparum: Detection of polymorphisms in the dihydrofolate reductase and dihydropteroate synthetase genes by PCR and restriction digestion. Exp. Parasitol. 89: 1–8.
•  Goldberg,  R.,  C.  Morvan and  J.C.  Roland (1986) Composition, properties and localization of pectins in young and mature cells of the mung bean hypocotyl. Plant Cell Physiol. 27: 417–429.
•  Hacia,  J.G. (1999) Resequencing and mutational analysis using oligonucleotide microarrays. Nat. Genet. 21: 42–47.
• Hair, J.F., R.E. Anderson, R.L. Tatham and W.C. Black (1995) Multivariate data analysis with readings, 4th edn. Prentice Hall, New Jersey, p. 127.
•  Hirata,  K.,  R.  Masuda,  Y.  Tsubokura,  T.  Yasui,  T.  Yamada,  K.  Takahashi,  T.  Nagaya,  T.  Sayama,  M.  Ishimoto and  M.  Hajika (2014) Identification of quantitative trait loci associated with boiled seed hardness in soybean. Breed. Sci. 64: 362–370.
•  Hirota,  T.,  K.  Takahata,  T.  Ogawa,  M.  Iwai and  Y.  Inoue (2005) Quality of soybean seeds grown in Hyogo prefecture. Bull. Hyogo Prefect. Technol. Cent. Agric., For. Fish., Agric. Sect. 53: 6–12 (in Japanese).
•  Holdaway-Clarke,  T.L. and  P.K.  Hepler (2003) Control of pollen tube growth: role of ion gradients and fluxes. New Phytol. 159: 539–563.
•  Isobe,  S.,  A.  Kaga,  A.  Tezuka and  G.  Ishikawa (2016) Current SNP genotyping methods for genetics and molecular breeding. Breed. Res. 18: 21–26 (in Japanese).
•  Jimenez-Lopez,  J.C.,  S.O.  Kotchoni,  M.I.  Rodríguez-García and  J.D.  Alché (2012) Structure and functional features of olive pollen pectin methylesterase using homology modeling and molecular docking methods. J. Mol. Model. 18: 4965–4984.
•  Johnson,  V.J.,  B.  Yucesoy and  M.I.  Luster (2004) Genotyping of single nucleotide polymorphisms in cytokine genes using real-time PCR allelic discrimination technology. Cytokine 27: 135–141.
•  Jungerius,  B.J.,  A.T.  Veenendaal,  B.A.  van Oost,  M.F.W.  te Pas and  M.A.M.  Groenen (2003) Typing single-nucleotide polymorphisms using a gel-based sequencer. Mol. Biotechnol. 25: 283–287.
•  Li,  S.,  H.N.  Liu,  Y.Y.  Jia,  Y.  Deng,  L.M.  Zhang,  Z.X.  Lu and  N.  He (2012) A novel SNPs detection method based on gold magnetic nanoparticles array and single base extension. Theranostics 2: 967–975.
•  Makabe,  Y. (2006) Effect of salts on cooked beans. Bull. Soc. Sea Water Sci., Jpn. 60: 342–347 (in Japanese with English abstract).
• Mori, A. and A. Taya (2008) Steamed soybean production method and natto using the same. JP4955640B2.
•  Motoki,  S.,  N.  Yamada,  N.  Tanaka,  M.  Takamatsu and  N.  Takahashi (1999) Studies on selection of soybean for high processing suitability in bean paste (miso). Hokuriku Crop Sci. 34: 118–119 (in Japanese).
•  Newton,  C.R.,  A.  Graham,  L.E.  Heptinstall,  S.J.  Powell,  C.  Summers,  N.  Kalsheker,  J.C.  Smith and  A.F.  Markham (1989) Analysis of any point mutation in DNA. The amplification refractory mutation system (ARMS). Nucleic Acids Res. 17: 2503–2516.
•  Pirhayati,  M.,  N.  Soltanizadeh and  M.  Kadivar (2011) Chemical and microstructural evaluation of ‘hard-to-cook’ phenomenon in legumes (pinto bean and small-type lentil). Int. J. Food Sci. Technol. 46: 1884–1890.
•  Saio,  K. and  T.  Watanabe (1972) Distribution of inorganic constituents (calcium, phosphorus and magnesium) in soybean seed and their changes in food processing. Nippon Shokuhin Kogyo Gakkaishi 19: 219–222 (in Japanese with English abstract).
•  Sajjaanantakul,  T.,  J.P.  Van Buren and  D.L.  Downing (1989) Effect of ester content on heat degradation of chelater-soluble carrot pectin. J. Food Sci. 54: 1272–1277.
•  Siliha,  H.,  W.  Jahn and  K.  Gierschner (1996) Effect of a new canning process on cell wall pectic substances, calcium retention and texture of canned carrots. Prog. in Biotechnol. 14: 495–508.
•  Smout,  C.,  D.N.  Sila,  T.S.  Vu,  A.M.L.  Van Loey and  M.E.G.  Hendrickx (2005) Effect of preheating and calcium pre-treatment on pectin structure and thermal texture degradation: a case study on carrots. J. Food Eng. 67: 419–425.
•  Stadhouders,  R.,  S.D.  Pas,  J.  Anber,  J.  Voermans,  T.H.  Mes and  M.  Schutten (2010) The effect of primer-template mismatches on the detection and quantification of nucleic acids using the 5′ nuclease assay. J. Mol. Diagn. 12: 109–117.
•  Stanley,  D.W.,  M.C.  Bourne,  A.P.  Stone and  W.V.  Wismer (1995) Low temperature blanching effects on chemistry, firmness and structure of canned green beans and carrots. J. Food Sci. 60: 327–333.
•  Suzuki,  M.,  K.  Fujino,  Y.  Nakamoto,  M.  Ishimoto and  H.  Funatsuki (2010) Fine mapping and development of DNA markers for the qPDH1 locus associated with pod dehiscence in soybean. Mol. Breed. 25: 407–418.
•  Taguchi-Shiobara,  F.,  K.  Fujii,  T.  Sayama,  K.  Hirata,  S.  Kato,  A.  Kikuchi,  K.  Takahashi,  M.  Iwahashi,  C.  Ikeda,  K.  Kosuge et al. (2019) Mapping versatile QTL for soybean downy mildew resistance. Theor. Appl. Genet. 132: 959–968.
•  Taira,  H. (1990) Quality of soybeans for processed foods in Japan. Jpn. Agric. Res. Q. 24: 224–230.
•  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.
•  Takemura,  H. (2001) Isolation of branched short-chain fatty acids non-producing Natto bacteria and its application to production of Natto with light smells. Jpn. J. Taste Smell Res. 8: 185–192 (in Japanese).
•  Toda,  K.,  K.  Hirata,  R.  Masuda,  T.  Yasui,  T.  Yamada,  K.  Takahashi,  T.  Nagaya and  M.  Hajika (2015) Relationship between mutations of the pectin methylesterase gene in soybean and the hardness of cooked beans. J. Agric. Food Chem. 63: 8870–8878.
•  Yamada,  T.,  M.  Hajika,  N.  Yamada,  K.  Hirata,  A.  Okabe,  N.  Oki,  K.  Takahashi,  K.  Seki,  K.  Okano,  Y.  Fujita et al. (2012) Effects on flowering and seed yield of dominant alleles at maturity loci E2 and E3 in a Japanese cultivar, Enrei. Breed. Sci. 61: 653–660.
•  Yasui,  T.,  T.  Sasaki,  K.  Kohyama and  M.  Hajika (2014) Variation in firmness of whole beans, embryos, and testas of cooked soybean (Glycine max) cultivars. Cereal Chem. 91: 419–424.
•  Yoshioka,  K.,  M.  Sekine,  M.  Suzuki and  K.  Otobe (2009) Influence of soaking and steaming conditions on the hardness of steamed soybeans and fermented soybeans (natto). Nippon Shokuhin Kagaku Kogaku Kaishi 56: 40–47.
•  Zhang,  B.,  P.  Chen,  C.Y.  Chen,  D.  Wang,  A.  Shi,  A.  Hou and  T.  Ishibashi (2008) Quantitative trait loci mapping of seed hardness in soybean. Crop Sci. 48: 1341–1349.