2021 Volume 27 Issue 3 Pages 463-471
Poly-gamma-glutamic acid (γ-PGA) is a major constituent of the fermented soybean product “natto”. We developed a high-accuracy method for quantifying γ-PGA content in natto rich in γ-PGA by lyophilizing and powdering the natto and optimizing protein and peptide removal, ethanol precipitation, degradation by hydrochloric acid, and quantification as glutamic acid using high-performance liquid chromatography. In this method, protein and peptide contaminants were removed from natto samples by trichloroacetic acid precipitation. The γ-PGA content of both γ-PGA-rich and γ-PGA-poor natto was measurable using this method. A single-laboratory validation test showed a repeatability relative standard deviation of 1.9 %–4.3 % and intermediate relative standard deviation of 1.8 %–3.9 %. The recovery of γ-PGA from natto samples spiked with three different amounts of γ-PGA was 98.5 %–101.0 %. These results show that our developed method has high accuracy and repeatability. Therefore, this method serves as a practical quantification protocol for determining the γ-PGA content of natto.
Poly-gamma-glutamic acid (γ-PGA) is a polymer produced by bacteria of the genus Bacillus including B. licheniformis,B. amyloliquefaciens,B. anthracis, and especially B. subtilis var. natto (Kimura et al., 2009; Gao et al., 2019). γ-PGA is a component of natto, a Japanese traditional food made from steamed soybeans fermented with B. subtilis, and is comprised of glutamic acid residues bound at the gamma position. This polymer contributes to the stickiness, slimy texture, and viscosity of natto.
The quality of natto varies depending on the bacterial strain used for fermentation. Natto bacteria have been bred for improved natto quality (Nagai, 1999) by increasing umami content (Nagai et al., 1994) and decreasing odor intensity of branched chain fatty acids (Takemura et al., 2000). The quantity of γ-PGA produced can be altered by various breeding procedures (Nagai, 1999). Different manufacturing conditions and soybean cultivars also affect the quality of natto (Matsumoto, 1995; Weng et al., 2010; Escamilla et al., 2019). Collectively, these variations necessitate the development of methods that can accurately quantify a wide range of γ-PGA contents (Bovarnick, 1942; Bovarnick et al., 1954; Fujii, 1963; Housewright and Thorne, 1950; Matsumoto and Iwahara, 1961; Yamaguchi et al., 1995).
The cetyltrimethylammonium bromide (CET) method, based on complex formation between γ-PGA and CET, has been developed as a simple and rapid spectrophotometric quantification method for determining the γ-PGA content of natto (Kanno and Takamatsu, 1995). However, γ-PGA is easily contaminated by proteins derived from soybeans during extraction from natto. Using the CET method, we observed high variability in the γ-PGA content of samples containing large amounts of protein or peptides, as with natto.
High-performance liquid chromatography (HPLC) can also be used to determine the γ-PGA content of natto by quantifying the amounts of D- and L-glutamic acids (Goto and Kunioka, 1992), which are produced after acid hydrolysis. Unlike the CET method, this method is applicable to a wider measurement range. However, removal of contaminants is important because proteins and peptides derived from soybeans may lead to inaccurate results. Use of dialysis has been reported to be effective for the purification of γ-PGA samples (Goto and Kunioka, 1992); however, it is time consuming, making it unsuitable for processing many samples or samples with high concentrations of γ-PGA.
In the present study, we 1) established a simple method for extracting and purifying γ-PGA, which is suitable for processing multiple samples with a wide range of γ-PGA content, 2) optimized a method for hydrochloric acid hydrolysis of extracted γ-PGA, 3) quantified glutamic acid content using HPLC (Nagai et al., 1997), and 4) determined the reproducibility and accuracy of this method.
Preparation of samples for γ-PGA content analysis As test samples, we used γ-PGA-rich, -average, and -poor natto, which was provided by Takano-foods Co., Ltd. (Ibaraki, Japan) and fermented using their own B. subtilis or B. subtilis IBARAKI lst-1 strains (Kubo and Nakagawa, 2018), according to a previously reported method (Kubo et al., 2011). After fermentation, natto samples were lyophilized and powdered using a mixer to increase sample homogeneity and extraction efficiency. The decrease in water weight during lyophilization was calculated and used for correction of the analysis. A food-additive material grade γ-PGA standard (purity > 70 %) was purchased from Meiji Food Materia Co., Ltd. (Tokyo, Japan). The γ-PGA standard was dissolved in water to a concentration of 1.6 mg/mL.
Scheme for γ-PGA quantification Our protocol for quantifying γ-PGA involved extracting γ-PGA from the lyophilized and powdered natto (Step 1), performing ethanol precipitation and recovery of γ-PGA by centrifugation (Step 2), removing proteins and peptides from the γ-PGA solution (Step 3), subjecting γ-PGA to hydrochloric acid degradation (Step 4), and determining glutamic acid content by HPLC (Step 5) (Fig. 3). We noted that when γ-PGA is recovered from natto samples, more sample-derived peptide and protein contaminants may be present than when it is directly recovered from a GSP plate containing phytone peptone, sodium glutamate and glucose agar (Nagai et al., 1997). Therefore, Steps 3 and 4 were additionally examined.
Optimized method for quantifying γ-PGA content in natto.
Extraction of γ-PGA from lyophilized and powdered natto (Step 1) Approximately 1 g of lyophilized and powdered natto was weighed and crude γ-PGA was extracted three times using 15 mL of distilled water. The extracts were then mixed together and 10 mL of this solution was dispensed into a 50 mL centrifugation tube for use in the following procedures.
Ethanol precipitation and recovery of γ-PGA (Step 2) In this step, 1 mL of 3 M sodium acetate (pH 5.2) (Nakashima, 1972) or 1 mL of saturated sodium chloride (about 5.4 M) (Sawa et al., 1973) was added to 10 mL of γ-PGA solution. A 2.5-fold volume of 99.5 % ethanol was added to the γ-PGA solution and mixed well. Then, crude γ-PGA was collected as a pellet by centrifugation at 9 700 × g, 4 °C for 10 min (7780, KUBOTA Co., Ltd.).
Removal of protein and peptide contaminants from the extracted γ-PGA solution (Step 3) In order to remove protein and peptide contaminants, four purification strategies using the γ-PGA standard and γ-PGA-rich natto were assessed: chloroform extraction and trichloroacetic acid (TCA) precipitation (Ch+/TCA+), chloroform extraction (Ch+/TCA−), TCA precipitation (Ch−/TCA+), or neither procedure (Ch−/TCA−). Chloroform extraction was performed by adding equal volumes of chloroform and γ-PGA solution, shaking the resulting mixture vigorously, and centrifuging at 9 700 × g, 4 °C for 10 min. For TCA precipitation, the γ-PGA pellet was dissolved in 10 % (w/v) TCA solution, incubated on ice for 30 min, and then the insoluble precipitate was removed by centrifugation at 21 000 × g at 4 °C for 10 min. The aqueous layer was subjected to the degradation step.
Degradation of γ-PGA by hydrochloric acid (Step 4) One milliliter of the sample solution and 1 mL of 6 M hydrochloric acid were poured into a degradation tube (yielding a final concentration of 3 M), which was then sealed with degassing. Degradation at 110 °C was assessed at 1, 4, 6, 8, and 24 h using a dry thermo unit (DTU-28, TAITEC Co., Ltd.).
Determination of glutamic acid amount by HPLC (Step 5) After hydrochloric acid degradation, 100 µL of the sample was mixed on ice with 30 µL of 10 M sodium hydroxide and 870 µL of distilled water. The neutralized solution was mixed with an equal amount of 0.2 M sodium citrate (pH 2.2) buffer and centrifuged at 15 000 × g at 4 °C for 5 min. The supernatant was analyzed by HPLC (Prominence series, SHIMADZU Co., Ltd.).
For HPLC analysis, the post-column orthophthalaldehyde derivatization method was used. HPLC conditions were as follows: column, Shim-pack Amino-Na; guard column, Shim-pack ISC-30/S0504 Na; mobile phase, Na-type amino acid mobile phase kit [AA-MA(Na), AA-MB(Na), AA-MC(Na)]; and reaction solution, amino acid reaction reagent kit (AA-RA, AA-RB) (all from SHIMADZU). The flow rate was 0.4 mL/min and the column temperature was 60 °C. The gradient method analysis program was used according to the user manual, and detection was performed using a fluorescence detector set at 350 nm as the excitation wavelength and 450 nm as the detection wavelength. Quantitative estimation was performed using the absolute calibration curve method.
Calculation of γ-PGA content from glutamic acid content The γ-PGA content (mg/100 g natto) was calculated using the following equation:
 |
Statistical analyses The significance of differences in γ-PGA content was determined by the Tukey–Kramer method using Ekuseru–Toukei 2015 (Social Survey Research Information Co., Ltd.). p < 0.05 was considered statistically significant.
Single-laboratory validation
Linearity of the calibration curve Reagent grade glutamic acid standard (FUJIFILM Wako Pure Chemical Co., Ltd., Osaka, Japan) was dissolved in 0.1 M hydrochloric acid and diluted to 1.56, 3.12, 6.25, 12.5, 25, 50, and 100 µg/mL solutions. The regression line between the glutamic acid concentration and peak area of glutamic acid standard solution obtained by HPLC analysis was determined, and the correlation coefficient was calculated.
Precision within a laboratory The γ-PGA content of natto samples (poor, average, and rich in γ-PGA content) was determined using the quantitative method established in this study. Triplicate measurements were carried out on three different days. The repeatability of the standard deviation (Sr), repeatability relative standard deviation (RSDr), intermediate standard deviation (Si), and intermediate relative standard deviation (RSDi) were determined by one-way analysis of variance.
Recovery The γ-PGA standard solution was prepared at a concentration of 2.9 mg/mL, and the purity of the solution was confirmed by HPLC analysis. The γ-PGA standard solution was added to the powdered γ-PGA from natto samples of average γ-PGA content at 1.61, 4.82, and 8.84 mL/g dry weight. The amount of spiked γ-PGA (1.61, 4.82, and 8.84 mL/g dry weight) was approximately equivalent to 200, 600, and 1 100 mg/100 g wet weight of natto, respectively.
Spiked and non-spiked samples were extracted as described above, and the γ-PGA content of each type of natto (poor, average, and rich in γ-PGA content) and the recovery were calculated according to the Procedural Manual of the Codex Alimentarius Commission (FAO, 2018).
Optimization of quantification method for γ-PGA content in natto
Sample preparation for γ-PGA content analysis (Step 1) At first, we compared the results of extracting γ-PGA from powdered natto after lyophilization and whole natto without lyophilization. The γ-PGA content extracted from whole natto was 722 ± 24 mg/100 g under Ch−/TCA+ treatment (the most optimal condition), while that extracted from the lyophilized natto powder was 1 019 ± 14 mg/100 g. The amount of γ-PGA extracted from the lyophilized powder was 40 % higher than that from whole natto. Furthermore, the standard deviation was reduced by about half, indicating that the variability was reduced. The results revealed that both extraction efficiency and reproducibility were superior when using lyophilized powder, and it was thought that the degradation by γ-glutamyl transferase (Kimura et al., 2009) would also be suppressed.
Comparison of salts for ethanol precipitation (Step 2) Salt is known to increase the efficiency of ethanol precipitation (Sambrook and Russell, 2001). Hence, its addition was necessary in Step 2 of our experiment. To determine the suitability of various salts, we compared the precipitation of the γ-PGA standard against the γ-PGA extracted from γ-PGA-rich natto using sodium acetate (Nakashima, 1972), sodium chloride (Nagai et al., 1997; Kubo et al., 2013; Sawa et al., 1973), or no salt (as a reference). γ-PGA has a relatively similar chemical structure to that of chondroitin sulfate (CS), and Nakashima (1972) reported that the concentration of sodium acetate required for complete sedimentation of CS in 70 % ethanol is ≥ 0.3 M. Additionally, Sawa et al. (1973) reported that γ-PGA showed good recovery from a culture medium when saturated sodium chloride was added to a 70 % ethanol solution at one-tenth of its volume, and that increasing the alcohol concentration increased the uptake of contaminants. In our study, concentrations of sodium acetate and sodium chloride were 0.3 and 0.54 M, respectively.
With regard to the γ-PGA standard, ethanol precipitation only occurred after the addition of sodium acetate or sodium chloride to the solution (Fig. 1). Further, no significant differences in the amount of γ-PGA precipitated from the natto sample were observed following the addition of either salt; however, sodium acetate was found to be more effective than sodium chloride in precipitating the γ-PGA standard (Fig. 1). These differences may be due to the pH of the salt solution, as the pKa of γ-PGA's carboxyl group is 4.09 (Tajima and Sukigara, 2011). Hence, acidification with sodium acetate (pH 5.2) may have suppressed the dissociation of carboxyl groups and reduced the carboxyl anions. Therefore, sodium acetate was selected for ethanol precipitation. The final concentration of salt in the solution was 0.3 M.
Effect of sodium acetate or sodium chloride on ethanol precipitation of γ-PGA.
*: No precipitation occurred. The final concentration of sodium acetate and sodium chloride in the sample solutions were 0.3 M and 0.54 M, respectively. Data are expressed as the mean ± SD of three samples. Different letters indicate significant differences in the amount of γ-PGA precipitated between treatments.
Removal of protein and peptide contaminants in the extracted γ-PGA (Step 3) Next, we examined the removal of contaminating proteins and peptides from the γ-PGA solution in Step 3. Some protein precipitation methods use ammonium sulfate, acetone, TCA, or dialysis (Jiang et al., 2004). We chose the TCA precipitation method because it is easy to perform, especially when handling many samples, and avoids the post-processing step before hydrochloric acid degradation. We also used chloroform to remove soluble organic components such as fatty acids. Chloroform has a specific gravity over 1.0, allowing for easy recovery of the aqueous supernatant containing γ-PGA after centrifugation.
The glutamic acid content of the γ-PGA standard with chloroform extraction (Ch+/TCA−), TCA precipitation (Ch−/TCA+), or no treatment (Ch−/TCA−) was 867 ± 21, 858 ± 12, and 866 ± 5 mg/100 g, respectively. As evidenced, its content did not differ remarkably between the procedures, indicating that the γ-PGA standard contained few free amino acids (i.e., peptide and protein contaminants). Table 1 lists the amino acid content in γ-PGA extracted from γ-PGA-rich natto using chloroform extraction and/or TCA precipitation. Contents of all amino acids, except glutamic acid, were significantly lower after chloroform extraction or TCA precipitation. Notably, the latter procedure was the most effective at removing contaminants and no combination effects of TCA precipitation and chloroform extraction were observed. Besides, given that chloroform is toxic (Fawell, 2000) and harms the environment, Ch−/TCA+ treatment was employed for quantifying the γ-PGA content.
| Amino acid content in γ-PGA extracted from γ-PGA-rich natto (mg/100g) | ||||
|---|---|---|---|---|
| Amino acid | Ch−/TCA− | Ch−/TCA+ | Ch+/TCA− | Ch+/TCA+ |
| Asp | 104.2 ± 0.2 | 52.1 ± 0.1 | 78.3 ± 0.2 | 52.2 ± 0.1 |
| Thr | 23.3 ± 0.0 | ND | 23.4 ± 0.1 | ND |
| Ser | 61.7 ± 0.1 | 20.6 ± 0.0 | 41.2 ± 0.1 | 20.6 ± 0.0 |
| Glu | 1381.6 ± 39.7 | 1161.7 ± 13.0 | 1381.3 ± 46.2 | 1164.2 ± 14.8 |
| Pro | 67.6 ± 0.1 | 22.5 ± 0.0 | 45.2 ± 0.1 | 22.6 ± 0.1 |
| Gly | 58.7 ± 0.1 | 29.4 ± 0.1 | 44.2 ± 0.1 | 29.5 ± 0.1 |
| Ala | 34.9 ± 0.1 | ND | 17.5 ± 0.1 | 17.5 ± 0.0 |
| Val | 22.9 ± 0.0 | ND | ND | ND |
| Cys | ND | ND | ND | ND |
| Met | ND | ND | ND | ND |
| Ile | ND | ND | ND | ND |
| Leu | 25.7 ± 0.0 | ND | 25.7 ± 0.1 | ND |
| Tyr | ND | ND | ND | ND |
| Phe | 129.3 ± 0.2 | 32.3 ± 0.1 | 64.8 ± 0.2 | 32.4 ± 0.1 |
| His | 30.4 ± 0.0 | 30.4 ± 0.1 | 30.4 ± 0.1 | 30.4 ± 0.1 |
| Lys | 85.8 ± 0.1 | 28.6 ± 0.1 | 57.4 ± 0.2 | 28.7 ± 0.1 |
| Arg | ND | ND | ND | ND |
| Total free amino acid | 2025.9 ± 39.3a | 1377.6 ± 13.0c | 1746.5 ± 45.4b | 1398.1 ± 15.1c |
| Other than glutamic acid | 644.5 ± 0.9a | 215.9 ± 0.5b | 428.1 ± 1.3c | 233.9 ± 0.6d |
| γ-PGA content (calculated) | 1212.5 ± 42.7 | 1019.5 ± 14.0 | 1140.1 ± 23.6 | 1021.7 ± 15.9 |
The four treatments of (Ch+/TCA+), (Ch+/TCA−), (Ch−/TCA+), and (Ch−/TCA−) indicate chloroform extraction and trichloroacetic acid (TCA) precipitation, chloroform extraction, TCA precipitation, and neither procedure, respectively. Data are expressed as mean ± SD of three samples. ND indicates it was not detected. Different letters indicate significant differences in amino acid contents between treatments.
Identification of the optimal γ-PGA degradation time (Step 4) The time required for γ-PGA degradation by hydrochloric acid was investigated using the γ-PGA standard and γ-PGA-rich natto to determine the optimal analytical condition. In previous reports, γ-PGA samples have been treated with 6 M hydrochloric acid (final concentration) at 110–150 °C for over 3 h (Goto and Kunioka, 1992; Nagai et al., 1997). However, excessive acid concentration and decomposition time may lead to false increases in glutamic acid due to the degradation of residual soybean-derived proteins and peptides in the γ-PGA solution. First, we confirmed that the degradation of γ-PGA proceeded in 3 M hydrochloric acid at 110 °C. By analyzing the γ-PGA content in the γ-PGA standard solution and γ-PGA-rich natto after 1, 4, 6, 8, and 24 h, we determined the necessary and sufficient time for the completion of γ-PGA degradation (Fig. 2). Results revealed that the amount of γ-PGA in the γ-PGA standard solution did not increase significantly after 4 h. The γ-PGA content in γ-PGA-rich natto after 1 and 4 h reached 1 045 ± 23 and 1 207 ± 12 mg/100 g, respectively. Concentrations of γ-PGA in γ-PGA-rich natto continued to increase steadily after 4 h, and by 24 h, the content was significantly greater than that observed at 4 h. These findings indicate that the degradation of γ-PGA in 3 M hydrochloric acid at 110 °C is complete at 4 h, and any subsequent increases in the amount of γ-PGA (calculated from the amount of glutamic acid) are believed to be due to other proteins and peptides.
Effect of degradation time on γ-PGA content determination from γ-PGA standard solution and γ-PGA-rich natto.
Data are expressed as the mean ± SD of three samples. Different letters indicate significant differences in the amount of γ-PGA detected between different degradation times.
Determination of glutamic acid content by HPLC (Step 5) Following γ-PGA degradation, the solution was neutralized with sodium hydroxide and diluted two-fold in a 0.2 M sodium citrate (pH 2.2) buffer. Next, the γ-PGA contents following treatment with centrifugation (15 000 × g at 4 °C for 5 min) or without centrifugation, respectively, were determined to be 756 ± 2.8 and 760 ± 1.5 mg/100 g in the γ-PGA standard and 1 019 ± 14 and 1 003 ± 28 mg/100 g in the γ-PGA-rich natto. Centrifugation of the γ-PGA solution prior to HPLC did not affect the γ-PGA content in any of the samples and removed precipitate when the solution became cloudy, suggesting that treatment with centrifugation is more ideal. Following this procedure, contents of amino acids, including glutamic acid, in the γ-PGA extract from γ-PGA-rich natto were measured by HPLC.
Calculation of γ-PGA content from glutamic acid content The glutamic acid (MW = 147.13) content observed before degradation was subtracted from that observed after degradation in order to exclude it from the analyses. Following this, the obtained value was used to determine the γ-PGA content (Eq. 1). For our calculations, we noted that one molecule of water (MW = 18.015) is eliminated from one peptide bond formation during glutamate polymerization. Figure 3 shows the optimized method for quantifying γ-PGA content in natto.
Single-laboratory validation
Linearity of the calibration curve Regression analysis of the glutamic acid standard curve revealed a correlation coefficient ≥ 0.999, and the curve showed high linearity in the range of 1.56–100 µg/mL.
Precision within a laboratory The mean values of γ-PGA content in natto (poor, average, or rich in γ-PGA content) were 183, 589, and 1 058 mg/100 g, respectively (Table 2). The RSDr of these natto samples was 4.3 %, 2.1 %, and 1.9 %, and the RSDi was 3.9 %, 2.3 %, and 1.8 %, respectively. The actual RSDr values were compared to those predicted by the Horwitz equation. The predicted RSDr (PRSDr) of these natto samples (poor, average, and rich in γ-PGA content) were 2.6 %, 2.2 %, and 2.0 %, respectively. All Horwitz ratio values for samples poor, average, and rich in γ-PGA content were < 2. Thus, both the repeatability and intermediate precision of natto γ-PGA quantification within a laboratory were good.
| γ-PGA content (mg/100g) | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| Sample | γ-PGA-poor natto | γ-PGA-average natto | γ-PGA-rich natto | ||||||
| Day 1 | Day 2 | Day 3 | Day 1 | Day 2 | Day 3 | Day 1 | Day 2 | Day 3 | |
| 1 | 194 | 197 | 179 | 588 | 563 | 616 | 1080 | 1080 | 1058 |
| 2 | 179 | 179 | 179 | 588 | 588 | 589 | 1054 | 1031 | 1056 |
| 3 | 179 | 179 | 179 | 589 | 588 | 589 | 1055 | 1031 | 1080 |
| Mean (mg/100g) | 183 | 589 | 1058 | ||||||
| Sr (mg/100g) | 7.9 | 12.4 | 20.1 | ||||||
| RSDr (%) | 4.3 | 2.1 | 1.9 | ||||||
| Si (%) | 7.2 | 13.5 | 18.9 | ||||||
| RSDi (%) | 3.9 | 2.3 | 1.8 | ||||||
| PRSDr (%) | 2.6 | 2.2 | 2.0 | ||||||
| HorRat | 1.5 | 1.1 | 0.9 | ||||||
Sr, repeatability standard deviation; RSDr, repeatability relative standard deviation; Si, intermediate standard deviation; RSDi, intermediate relative standard deviation; PRSDr, predicted relative standard deviation. PRSDr was calculated as C−0.1505. Day 1, 2, and 3 indicate triplicate measurements were carried out on three different days.
Recovery The mean recovery of spiked γ-PGA at 200, 600, and 1 100 mg/100 g was 98.5 %–101.0 % (Table 3). RSDr values were 2.1 %, 2.2 %, and 2.2 %, and RSDi values were 6.7 %, 2.4 %, and 3.4 %, respectively. According to the single-laboratory validation of the Codex Alimentarius Commission procedural manual (FAO, 2018), the recovery limit of the spiked analyte was 97 %–103 %. These results indicate that the trueness of the developed method for quantifying the γ-PGA content of natto sample is good.
| Fortified (mg/100g) | Recovery (%) | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| 200 | 600 | 1100 | |||||||
| Day 1 | Day 2 | Day 3 | Day 1 | Day 2 | Day 3 | Day 1 | Day 2 | Day 3 | |
| 1 | 103.6 | 92.4 | 104.3 | 100.9 | 103.7 | 98.6 | 100.5 | 103.8 | 101.4 |
| 2 | 100.0 | 92.7 | 100.6 | 103.5 | 103.4 | 100.5 | 93.6 | 102.3 | 101.3 |
| 3 | 100.4 | 88.5 | 103.9 | 96.6 | 101.2 | 100.8 | 98.8 | 103.7 | 102.8 |
| Mean (%) | 98.5 | 101.0 | 100.9 | ||||||
| Sr (mg/100g) | 2.1 | 2.3 | 2.2 | ||||||
| RSDr (%) | 2.1 | 2.2 | 2.2 | ||||||
| Si (%) | 6.6 | 2.4 | 3.4 | ||||||
| RSDi (%) | 6.7 | 2.4 | 3.4 | ||||||
Sr, repeatability standard deviation; RSDr, repeatability relative standard deviation; Si, intermediate standard deviation; RSDi, intermediate relative standard deviation. Day 1, 2, and 3 indicate triplicate measurements were carried out on three different days.
Comparison with CET method Although the CET method is simple and has demonstrated a γ-PGA recovery rate of 96.5 %–102.5 %, removal of protein and peptide contaminants is necessary when the natto extracts contain a large amount of contaminants (Kanno and Takamatsu, 1995). The γ-PGA content in commercial natto fermented in polystyrene paper is reported to be 162–549 mg/100 g (Kanno and Takamatsu, 1995). The CET method accurately quantifies samples containing low to average amounts of γ-PGA, but this method might not be suitable for γ-PGA-rich natto. In this study, to obtain a clean and optimized γ-PGA yield, the amount of glutamic acid was measured after hydrochloric acid hydrolysis via HPLC and a 98.5 %–101.0 % recovery was observed from samples spiked with different γ-PGA concentrations (Table 3). Taken together, these results indicate that our quantification method is superior to the CET method because it is more accurate and does not require special reagents, instruments, or devices. Notably, complementary use of this method with the CET method may additionally broaden the measurable range of γ-PGA content when the sample type is taken into consideration.
Application of our developed method Natto is known to possess strong alkaline protease activity, resulting in the decomposition of proteins into peptides and amino acids (Kada et al., 2013). The alkaline protease continues to degrade soybean proteins even after fermentation is completed and the natto is stored in cold storage. The γ-PGA extracted from such natto contains a large amount of peptides and amino acids, but our developed method can eliminate these contaminants. Therefore, it is suggested that this method can be applied to natto with more advanced degradation of soybean proteins due to the passage of time after production. The quantification method we developed seems to be applicable to protein-rich samples such as “hikiwari natto”. Hikiwari natto is made from crushed soybean that is not enveloped by the hull, making the contents more soluble than that of ordinary natto made from whole soybean.
In addition to natto, fermented foods produced by B. subtilis and containing γ-PGA include Cheongguk-jang, Kinema, and Tua-Nao (Ratha and Jhon, 2019; Inatsu et al., 2002). The γ-PGA quantification method developed in this study is expected to be applicable to the evaluation of γ-PGA content in all of these foods.
To quantify the content of γ-PGA in natto, we optimized a method for measuring the amount of glutamic acid produced by acid hydrolysis of γ-PGA extract using HPLC. A validation test showed that the RSDr and RSDi were 1.9 %–4.3 % and 1.8 %–3.9 %, respectively. The recovery of γ-PGA from natto samples spiked with three different amounts of γ-PGA was 98.5 %–101.0 %. These results demonstrate that this method has high accuracy, repeatability, and performance within the range specified by the Codex Alimentarius Commission procedural manual. Furthermore, our method does not require special reagents, instruments, or devices. Therefore, this method serves as a useful quantification procedure for determining the γ-PGA content of natto.
Acknowledgements This work was supported by a grant from a commissioned project study on “Research development for discovering of regional agricultural products and foods having a beneficial impact on health,” Ministry of Agriculture, Fishery and Forestry, Japan (JPJ005336).
Conflict of Interest There is no conflict of interest.
