アルカリ処理とLacticaseibacillus casei発酵の併用によるおから中のイソフラボン配糖体のアグリコンへの変換率の向上

Abstract

To increase conversion rate of isoflavone glucosides to aglycones in okara through solid-state fermentation with lactic acid bacteria (LAB), we investigated the effects of alkali treatment to okara on hydrolyzation of isoflavone malonylglucosides. After incubation of okara in a solid state at various pH adjusted with sodium hydroxide solution, malonyl groups of malonylglucosides were most effectively hydrolyzed at pH 11.0. As a result of the alkali treatment, the proportion of β-glucosides in the total isoflavones in okara increased from 32.1% to 86.4%. After fermentation of alkali-treated okara with Lacticaseibacillus casei, 93.3% of the total isoflavones were converted to aglycones, which were 1.8-fold higher than those of untreated okara. Moreover, similar hydrolysis of malonylglucosides was observed when potassium carbonate solution was used. Thus, the combination of alkali treatment and fermentation with LAB could be one of the useful practical methods to increase the isoflavone aglycones contents in okara as a food material at low energy costs.

Translated Abstract

豆腐製造時の副産物であるおからは食品素材としての有効利用が求められている．おからの有用成分を高める方法の一つとして，乳酸菌（Lacticaseibacillus casei）によるイソフラボン配糖体からアグリコンへの変換を行った．特に，発酵の前処理として，温和なアルカリ処理によるマロニル配糖体のグルコシド配糖体への加水分解を検討した．おからをpH11.0で24時間保持することにより，全イソフラボンの86.4％がグルコシド配糖体となった．水酸化ナトリウムによりアルカリ処理を行った場合，おからの乳酸発酵後のアグリコン含量は全イソフラボンの93.3％であり，アルカリ処理を行わない場合の1.8倍に増大した．同様の結果は炭酸カリウムによるアルカリ処理によっても認められた．アルカリ前処理と乳酸菌による固体発酵は，おからのイソフラボンアグリコン含量を高めるための低コストで簡便な方法であると考えられた．

1. Introduction

Okara, a byproduct of tofu manufacturing, is produced when soymilk is separated by the filtration of ground and heated soybeans. Since raw okara putrefies rapidly after production in tofu manufacturing factories, it is rarely used, except as feed for livestock. To prevent the putrefaction of raw okara, we developed a method that combines solid fermentation with lactic acid bacteria (LAB) and semi-anaerobic incubation [1]. This system does not require additional heat sterilization, capital investment, or energy. Furthermore, the fermentation of okara with LAB should contribute to increased shelf life and enhanced functionality.

Isoflavones, which are specific flavonoid compounds in soybeans, are one of the useful dietary components in okara. Isoflavones consist of three types of aglycones: daidzein, glycitein, and genistein. These aglycones and their derivatives are classified into β-glucosides (daidzin, glycitin, and genistin), acetyl-β-glucosides (acetyldaidzin, acetylglycitin, and acetylgenistin), and malonylglucosides (malonyldaidzin, malonylglycitin, and malonylgenistin). In the human body, isoflavone glucosides are hydrolyzed to glucose and aglycones. Isoflavone aglycones are absorbed faster and in greater amounts than glucosides in humans. Isoflavone aglycones are structurally similar to the female hormone estrogen and play important roles in regulating the effects of estrogen [2, 3]. Okara could be a good source of isoflavones due to 12-40% of the total isoflavone content in the soybean remaining in okara after tofu manufacturing [4]. In both soybean seeds and okara, most of isoflavones are malonylglucoside and β-glucoside forms, and aglycones rarely exist [5, 6]. Therefore, isoflavone-aglycone-rich soy products have been developed for enhancing the prevention of several diseases.

In the case of soymilk, isoflavone glucosides are hydrolyzed to glucose and aglycones during fermentation with LAB with β-glucosidase activity [7-9]. Otiento et al. suggested that isoflavone malonylglucosides remaining in soymilk were not hydrolyzed by β-glucosidases of LAB [8]. Lim et al. indicated that LAB exhibiting weak positive esterase activity in addition to β-glucosidase activity reduced malonylglucosides, but the reduction rate of malonylglucosides was much lower than that of β-glucosides [10]. Therefore, to convert the isoflavone glucosides to aglycones in okara by LAB more effectively, malonylglucosides must be hydrolyzed before fermentation.

Alkali hydrolysis is one of the methods for converting malonylglucosides to β-glucosides, which are mainly used for sample preparation in HPLC analysis to determine the amount of isoflavones [11-13]. The hydrolysis of malonylglucosides under alkaline conditions may be a useful practical method without energy costs or additional capital investment. This study investigated the mild alkali treatment for hydrolyzing malonylglucosides to β-glucosides in okara and converting to aglycones by LAB to produce aglycone-rich valuable food materials.

2. Materials and Methods

2.1 Okara

The okara used in this study was obtained from a tofu factory in Aichi, Japan. Five-kilogram samples were placed on a tray immediately after separation from the soymilk, cooled in a refrigeration chamber, and then stored at -20°C prior to the experiment. Okara was thawed in the refrigerator before alkali treatment or fermentation.

2.2 Alkali treatment of okara

The okara which contains about 75% moisture was subjected to alkali treatement in a solid state. Okara (100 g, initial pH 6.8) was adjusted to pH 8.0, 9.0, 10.0, 11.0, and 12.0 using 1 M NaOH solution. Each okara sample was sealed in a polyethylene bag and incubated at 30°C for 24 h, after which the bag was opened, and the sample was neutralized with 1 M hydrogen chloride (HCl) solution. For alkali treatment with K₂CO₃ solution, 2 M K₂CO₃ solution was used for pH adjustment, and undiluted lactic acid solution was used for neutralization.

2.3 Solid state fermentation of okara with L. casei

Alkali-treated okara was fermented the same way as in a previous study [1]. Lacticaseibacillus casei L-14 (MAFF 401403) obtained from the Genebank Research Center of Genetic Resources (Ibaraki, Japan) was used. The strain was propagated in de Man, Rogosa, and Sharp (MRS) broth (Difco, St. Louis, MO, USA) at 30°C for 24 h. The cultures were harvested by centrifuging at 10,000 × g for 5 min at 4°C, washed twice with sterile 0.6 mM potassium phosphate buffer (pH 7.2) in 0.85% (w/v) NaCl, and resuspended in sterile distilled water at a cell concentration of 8.3 log colony-forming units (CFU)/g. Okara (100 g) was added with 5% (w/w) of the cell suspension of L. casei, containing 8.3 log CFU/g of cells (okara contains 7.0 log CFU/g), mixed uniformly in a polyester bag, and fermented at 35°C.

2.4 Determining pH values

Fermented okara (5.0 g) was mixed with 45 mL of distilled deionized water and shaken at 150 rpm for 10 min. The pH values of the samples were measured using a pH meter (LAQUA F-72; Horiba Ltd., Kyoto, Japan).

2.5 Extraction of isoflavones and HPLC analysis

Each sample was freeze-dried into a powder, and 2.0 g of the powder was mixed with 10 mL of 70% ethanol, sonicated for 30 min, and extracted three times. The final volume was adjusted to 20 mL by adding 70% ethanol.The extract was filtered through a 0.45-μm filter. The isoflavone content was determined using a HPLC system (Prominence HPLC System; Shimadzu Co., Kyoto, Japan) equipped with a Develosil C30-UG-5 column (4.6 × 150 mm ID; Nomura Chemical Co., Ltd., Aichi, Japan). The mobile phases were a gradient of solution A (0.1% phosphate in 15% acetonitrile) and solution B (0.1% phosphate in 70% acetonitrile) at a ratio of 63:37, v/v from 10 to 60 min of running time at a flow rate of 1.0 mL/min. The absorbance of the isoflavones was measured at 254 nm using a photodiode array detector. The three acetyl glucosides were not observed because they are present in trace amounts in okara and were undetectable. The content of each isoflavone in the sample was calculated using a linear plot of the standard isoflavones. The isoflavone content in okara is expressed as μmol/100 g of dry weight. The total amount of isoflavone aglycones is calculated by the sum of daidzein, glycitein, and genistein content. The total amount of β-glucosides and malonylglucosides are calculated by the sum of daidzin, glycitin, and genistin content, and by the sum of malonlyldaidzin, malonylglycitin, and malonylgenistin content, respectively.

2.6 Nutrient Composition

The nutritional composition of okara was analyzed using standard methods [14]. Moisture content was determined using the atmospheric heating and drying method at 105°C. Protein content was analyzed by the Kjeldahl method with a nitrogen-to-protein conversion factor of 5.71. Lipid content was measured using the chloroform-methanol extraction method. Dietary fiber was determined by the Prosky method. Ash content was analyzed by direct ashing at 550°C, and sodium content was determined after dry digestion of the sample using atomic absorption spectrophotometry. Available carbohydrate content was calculated by subtracting the amounts of moisture, protein, lipid, dietary fiber, and ash from 100. Each data was expressed as the mean of two measurements.

3. Results and Discussion

3.1 Changes in the isoflavone composition of okara after alkali treatment

The composition of isoflavone β-glucosides, malonylglucosides, and aglycones in okara before alkali treatment is shown in Table 1 (Control). The contents of β-glucosides, malonylglucosides, and aglycones forms were 100.5 μmol/100 g, 191.2 μmol/100 g and 22.4 μmol/100 g, respectively which were 32.0%, 60.9%, and 7.1% of total isoflavones. The results were consistent with those of previous study [5,6].

Table 1 Effect of alkali treatment (NaOH) at different pH on the hydrolyzation of isoflavones malonylglucosides in okara.

Category	Isoflavone	(μmol/100 g)
		pH 6.8 (Control)	pH 8.0	pH 9.0	pH 10.0	pH 11.0	pH 12.0
β-Glucosides	Daidzin	34.7±1.0	37.1±2.5	43.9±2.6	64.5±5.7	81.7±7.9	9.4±1.1
	Glycitin	2.1±0.2	2.4±0.1	2.9±0.1	4.4±0.1	5.8±0.2	7.3±1.7
	Genistin	63.7±1.5	68.7±4.4	80.2±4.6	119.4±5.6	156.1±7.4	135.3±11.2
Malonylglucosides	Malonyldaidzin	74.0±3.4	54.8±9.7	46.8±8.2	25.5±4.4	5.5±0.9	n.d.
	Malonylglycitin	4.6±0.6	5.0±0.1	4.1±0.0	2.5±0.7	n.d.	n.d.
	Malonylgenistin	112.6±5.0	100.5±2.8	87.3±3.0	50.9±1.8	13.7±0.7	n.d.
Isoflavone aglycones	Daidzein	9.2±0.4	8.8±0.0	8.7±0.2	8.6±0.2	8.3±0.1	7.5±0.7
	Glycitein	n.d.	n.d.	n.d.	n.d.	n.d.	n.d.
	Genistein	13.2±0.9	12.1±0.2	12.0±0.2	11.6±0.0	11.0±0.2	9.3±0.6
	Total	313.2±6.2	289.5±0.1	286.0±2.9	287.4±8.8	282.1±15.5	168.8±15.1

Data were the means ± standard deviation of three independent experiments.

n.d., not detectable.

In this study, okara was subjected to alkali treatment to cleave ester bonds in the malonyl groups of isoflavone malonylglucosides. The contents of malonylglucosides and β-glucosides in okara are shown in Fig. 1, which describes the data at incubation pH values of 8.0 to 12.0. The pH of okara used in this study was 6.8. The malonylglucosides content decreased as pH increased and was mostly undetectable at 12.0. In contrast, β-glucoside levels increased with increasing pH values to 11.0, then they rapidly decreased at pH 12.0. The decrease in malonylglucosides levels and the increase in β-glucosides levels changed linearly as pH increased from pH 8.0 to 11.0. At　pH 12.0, β-glucosides levels remarkably decreased, though isoflavone aglycones were not detected. From these results, isoflavone malonylglucosides were most efficiently converted to β-glucosides at pH 11.0.

Fig. 1

Effect of alkali treatment on hydrolyzation of isoflavone malonylglucosides at different pH. Data were expressed as mean ± standard deviations from triplicate experiments.

Table 1 shows the each isoflavone content in okara with alkali treatment. The total isoflavone content was not significantly different according to the pH value, except at pH 12.0. The decrease in malonyldaidzin and the increase in daidzin, as well as the decrease in malonylgenistin and the increase in genistin, were approximately equivalent under varying pH conditions. The levels of β-glucosides, malonylglucosides, and aglycones formed in okara after alkali treatment at pH 11.0, which is the optimal condition in this study were 86.4%, 6.8%, and 6.8% of total isoflavones. Incidentally, increases in the levels of aglycones were not observed at all pH values examined. From these results, the increase in isoflavone glucoside content was likely a result of hydrolyzation of malonyl group of malonyl glucosides.

At pH 12.0 more significant decrease was observed in daidzin than genistin among β-glucosides, which suggests that daidzin would become more unstable by alkalinity than genistin. Details of the reaction mechanism are unknown, although it is possible that the heterocyclic compound of isoflavone could be decomposed at pH 12.0.

For the HPLC analysis of isoflavones, the ester bonds in the malonyl groups of isoflavone malonylglucosides were cleaved in methanol or acetonitrile extract solutions using 2 M NaOH at room temperature for 10 min [13]. In this study, the amount of 1 M NaOH required to adjust the pH to 11.0 was approximately 8.0 mL per 100 g of okara. Thus, the isoflavone malonylglucosides were hydrolyzed in notably lower concentrations of NaOH for a significantly longer incubation time than in the sample preparation process. The glycosidic bond between glucose and aglycone was not affected by alkaline conditions or the sample preparation process. Similar alkaline conditions have been reported for soy slurry production. Rickert et al. reported that approximately 50% of malonyldaidzin and 57% of malonylgenistin were hydrolyzed into daidzin and genistin, respectively, at pH 10.5 for 30 min in a soy slurry [15]. In this study, approximately 54% of malonyldaidzin and 56% of malonylgenistin in okara were converted to daidzin and genistin, respectively, after 2 h at pH 11.0 (data not shown). The conversion rate of malonylglucosides to β-glucosides in okara was similar to that in soy slurry.

Furthermore, the low-concentration alkali treatment applied in this study had minimal effects on the nutritional composition of okara (Table 2). While slight increases in moisture, ash, and sodium content were observed due to the addition of the NaOH solution, the overall nutritional composition of the original okara remained largely unchanged. Although no significant decrease in protein or lipid content was observed, it is possible that the structure of individual compounds was affected by alkali treatment without undergoing significant degradation. Taken together, alkali treatment may serve as a simple and effective pretreatment method for the fermentation of okara.

Table 2 Nutrient composition of alkali-treated okara.

Nutrient	Control	Alkali-treated okara
Moisture (g/100g)	76.7	79.4
Protein (g/100g)	6.3	5.6
Lipid (g/100g)	2.8	2.7
Available carbohydrate (g/100g)	2.0	1.2
Dietary fiber (g/100g)	11.4	10.0
Ash (g/100g)	0.9	1.1
Sodium (mg/100g)	36	850

3.2. Production of isoflavone aglycones in okara using a combination of alkali treatment with L. casei fermentation

The effect of alkali treatment on the production of isoflavone aglycones in okara during fermentation with L. casei was investigated. Fig. 2 (A) shows the changes in the pH of okara, with or without alkali pretreatment during fermentation. The pH of okara sharply reduced to under 5.0 after 16 h of fermentation, and then remained stable until 72 h of fermentation whether or not alkali treatment is applied. The slight differences in the pH values between the two samples were likely due to the neutralization conditions. Fig. 2 (B) shows the changes in isoflavone aglycones during fermentation with and without alkali pretreatment. Without alkali pretreatment, aglycones levels increased to 96.7 μmol/100 g after 24 h of fermentation, then mostly reached the peak point after 48 h of fermentation. Finally, the isoflavone aglycones content was 145.9 μmol/100 g after 72 h of fermentation. In contrast, with alkali pretreatment, aglycones levels significantly increased to 203.4 μmol/100 g after 24 h of fermentation, then mostly reached the peak point after 48 h of fermentation as well. Finally, the content of isoflavone aglycones was 259.0 μmol/100 g after 72 h fermentation, which were approximately 1.8-fold higher than the levels in okara without alkali pretreatment. The conversion rate of isoflavone aglycones was approximately two-fold faster after 24 h of fermentation in okara treated with alkali than without alkali pretreatment. This is probably because the content of β-glucosides before fermentation, as a substrate of β-glucosidase, was approximately 2.4-fold higher after alkali pretreatment (Table 1). Based on these results, the increase in β-glucoside levels by alkali pretreatment allowed the production of isoflavone aglycones more efficiently.

Fig. 2

Changes in pH (A) and isoflavone aglycone (B) content during fermentation with L. casei with or without alkali pretreatment. Data were expressed as mean ± standard deviations from triplicate experiments.

The composition of isoflavone β-glucosides, malonylglucosides, and aglycones in okara after 72 h fermentation, with or without alkali pretreatment, is shown in Table 3. Without alkali pretreatment, most β-glucosides in okara were converted to aglycones during fermentation, as compared to unfermented okara (Table 1 and Table 3). That is, daidzin and genistin levels decreased from 34.7 and 64.7 μmol/100 g to 5.2 and 4.1 μmol/100 g, respectively, and daidzein and genistein levels increased from 9.2 and 13.2 μmol/100 g to 55.2 and 87.1 μmol/100 g, respectively. The aglycones content was 49.6% of the total isoflavones after fermentation. In addition, as the levels of malonyldaidzin and malonylgenistin decreased from 74.0 and 112.6 μmol/100 g to 54.9 and 81.3 μmol/100 g, respectively, a part of malonylglucosides are likely to be converted to aglycones through β-glucosides. The degradation rate of malonylglucosides in this study was calculated to 26.9% which was almost the same as previous study about an isoflavone standard fermented by L. plantarum [10]. From these results, most of the β-glucosides in okara were converted to aglycones at a high efficiency by L. casei, while malonylglucosides seemed to be hydrolyzed much more slowly.

Table 3 Changes in the isoflavone contents in alkali-treated okara (NaOH) after fermentation with L. casei.

Category	Isoflavones	Fermentation of untreated okara (μmol/100 g)	Fermentation of alkali-treated okara (NaOH, μmol/100 g)
β-Glucosides	Daidzin	5.2±0.5	6.3±0.1
	Glycitin	n.d.	tr.
	Genistin	4.1±0.6	tr.
Malonylglucosides	Malonyldaidzin	54.9±1.3	2.9±0.1
	Malonylglycitin	3.5±0.5	n.d.
	Malonylgenistin	81.3±3.8	7.3±0.2
Isoflavone aglycones	Daidzein	55.2±0.5	94.8±0.0
	Glycitein	3.6±0.6	8.2±0.1
	Genistein	87.1±1.4	156.0±0.2
	Total	294.9±0.4	277.5±1.2

Data were expressed as mean ± standard deviations from triplicate experiments.

n.d., not detectable, tr, minimum of the linear range.

With alkali pretreatment, a similar trend was observed as in the case without alkali pretreatment. (Table 1 and Table 3). However, the higher content of β-glucosides in alkali-pretreated okara resulted in a greater production of aglycones after fermentation. that is, daidzin and genistin levels decreased from 81.7 and 156.1 μmol/100 g to 6.3 μmol/100 g and undetectable levels, respectively, and daidzein and genistein levels significantly increased from 8.3 and 11.0 μmol/100 g to 94.8 and 156.0 μmol/100 g, respectively during fermentation. Finally, the aglycones content was 93.3% of the total isoflavones after fermentation of okara with alkali pretreatment.

In this study, the amount of NaOH required to hydrolyze malonylglucosides in okara was 8.0 mmol per 100 g of raw okara, which was much less than required for sample preparation for the HPLC analysis. However, considering their practical use as food additives, alkali chemicals other than NaOH are preferred due to the health hazards associated with NaOH. Potassium carbonate (K₂CO₃), commonly used as a pH-raising agent or neutralizer in food products, was selected as a possible alternative to NaOH for treating okara. Therefore, we investigated whether the hydrolysis of malonylglucosides and the subsequent conversion to aglycones through fermentation could occur under K₂CO₃ treatment, similar to NaOH treatment.

Table 4 showed the changes isoflavone composition of okara incubated at pH 10.0 adjusted with 2 M K₂CO₃ solution and fermented with L. casei. The pH of okara treated with 2 M K₂CO₃ was 10.8 as maximal. However, under these conditions, not only malonylglucosides but also β-glucosides were hydrolyzed, resulting in a decrease in total isoflavones to 102.3 μmol/100 g. In comparison, at pH 10.0, the total isoflavone content was 248.7 μmol/100 g, which represents 19.8% decrease compared to the content before alkali treatment. At pH 10.0, the levels of β-glucosides, malonylglucosides, and aglycones formed in okara after alkali treatment at pH 10.0 with K₂CO₃ were 76.5%, 10.9%, and 12.6% of total isoflavones. After fermentation, daidzein and genistein levels were 103.1 and 107.8 μmol/100 g, respectively. The aglycones content was 88.3% of the total isoflavones after fermentation of okara with alkali pretreatment using K₂CO₃.

Table 4 Changes in the isoflavone contents in alkali-treated okara (K₂CO₃) after fermentation with L. casei.

Category	Isoflavones	Control (μmol/100 g)	Alkali-treated okara (K2CO3, μmol/100 g)	Fermentation of alkali-treated okara (K2CO3, μmol/100 g)
β-Glucosides	Daidzin	37.9±3.4	81.1±2.9	8.4±1.6
	Glycitin	2.1±0.1	6.3±0.1	n.d.
	Genistin	52.0±3.4	102.7±3.4	tr.
Malonylglucosides	Malonyldaidzin	71.1±4.6	9.6±0.6	7.2±0.9
	Malonylglycitin	5.0±0.2	n.d.	n.d.
	Malonylgenistin	97.8±1.3	17.6±1.1	13.3±1.3
Isoflavone aglycones	Daidzein	19.7±0.8	17.2±0.1	103.1±9.8
	Glycitein	n.d.	n.d.	6.1±0.7
	Genistein	17.6±0.4	14.2±0.4	107.8±7.3
	Total	310.1±12.6	248.7±8.8	245.8±21.4

Data were expressed as mean ± standard deviations from triplicate experiments.

n.d., not detectable, tr, minimum of the linear range.

From these results, the hydrolysis of malonylglucosides was suggested to be influenced not only by pH but also by the type of alkali chemical and the amount used. The observed differences were presumed to be partly attributed to variations in dissociation properties of the alkali chemicals. Although further investigation is needed to determine the optimum conditions, K₂CO₃ was shown to be a safer and effective alternative for alkali pretreatment of okara in preparation for fermentation with LAB. Hydrolysis of isoflavone malonylglucosides in okara by alkali pretreatment and fermentation with LAB is one of the effective methods for increasing isoflavone aglycones in okara. By preventing putrefaction and increasing the levels of aglycones, okara can be used as a functional food in conjunction with dietary fiber and LAB.

4. Conclusion

The present study reports a novel method to produce isoflavone aglycones in okara by fermentation with L. casei combined with alkali pretreatment. As the isoflavone malonylglucosides in okara are insufficiently converted to aglycones by L. casei, malonylglucosides are required for hydrolysis to β-glucosides before fermentation. Alkaline treatment was found to be a useful method for cleaving the ester bonds of malonylglucosides. Fermentation with L. casei combined with alkali pretreatment converted 93.3% of the total isoflavones in okara to aglycones, which may make okara a more valuable food material with high amounts of isoflavone aglycones and dietary fiber.

References

• 1)  M. Hiwatashi, S. Kano, T. Kato; “Development of a preservation method for okara using lactic acid fermentation” (in Japanese). Nippon Shokuhin Kagaku Kogaku Kaishi, 62, 572-578. (2015).
• 2)  T. Izumi, M. K. Piskula, S. Osawa, A. Obata, K. Tobe, M. Saito, S. Kataoka, Y. Kubota, M. Kikuchi; Soy isoflavone aglycones are absorbed faster and in higher amounts than their glucosides in humans. J. Nutr., 130, 1695–1699 (2000).
• 3)  M. Kano, T. Takayanagi, K. Harada, S. Sawada, F. Ishikawa; Bioavailability of isoflavones after ingestion of soy beverages in healthy adults. J. Nutr., 136, 2291–2296 (2006).
• 4)  A. Colletti, A. Attrpvio, L. Boffa, S. Mantegna, G. Cravotto; Valorization of by-products from soybean (Glycine max (L.) Merr.) processing. Molecules, 25, 2129-2162 (2020).
• 5)  H. J. Wang, P. A. Murphy; Mass balance study of isoflavones during soybean processing. J. Agric. Food Chem., 44, 2377-2383 (1996).
• 6)  T. H. Kao, Y. F. Lu, H-C. Hsieh, B. H. Chen; Stability of isoflavone glucosides during processing of soymilk and tofu. Food Res. Int., 37, 891-900 (2004).
• 7)  H. L. Chien, H. Y. Huang, C. C. Chou; Transformation of isoflavone phytoestrogens during the fermentation of soymilk with lactic acid bacteria and bifidobacteria. Food Microbiol., 23, 772-778 (2006).
• 8)  J. A. Marazza, M. S. Garro, G. S. de Giori; Aglycone production by Lactobacillus rhamnosus CRL981 during soymilk fermentation. Food Microbiol. 26, 333-339 (2009).
• 9)  D. O. Otieno, J. F. Ashton, N. P. Shah; Evaluation of enzymic potential for biotransformation of isoflavone phytoestrogen in soymilk by Bifidobacterium animals, Lactobacillus acidophilus and Lactobacillus casei. Food Res. Int., 39, 394-407 (2006).
• 10)  Y. J. Lim, B. Lim, H. Y. Kim, S. J. Kwon, S. H. Eom; Deglycosylation patterns of isoflavones in soybean extracts inoculated with two enzymatically different strains of lactobacillus species. Enzyme Microb. Technol., 132, 109394-109401 (2020).
• 11)  P. Delmonte, J. Perry, J. I. Rader; Determination of isoflavones in dietary supplements containing soy, Red Clover and kudzu: extraction followed by basic or acid hydrolysis. J. Chromatogr. A, 1107, 59-69 (2006).
• 12)  D. Yuan, Y. Pan, Y. Chen, T. Uno, S. Zhang, Y. Kano; An improved method for basic hydrolysis of isoflavone malonylglucosides and quality evaluation of Chinese soy materials. Chem. Pharm. Bull. 56, 1-6 (2008).
• 13)  AOAC International, AOAC Official Method 2001.10: Isoflavones in Soy and Selected Foods Containing Soy: Extraction, Saponification, and Liquid Chromatography, C45-114-C45-117, AOAC International, Rockville, MD (2001).
• 14)  Y. Funatsu, Y. Nishimura, M. Ishioroshi, K. Uebaba, Y. Nishio, T. Terashima and N. Mafune; “Hypoglycemic effect of refuse-enriched cake on blood glucose level in healthy subjects and its sensory attributes”(in Japanese). Nippon Shokuhin Kagaku Kogaku Kaishi, 55, 367-372. (2008).
• 15)  D. A. Rickert, M. A. Meyer, J. Hu, P. A. Murphy; Effect of extraction pH and temperature on isoflavone and saponin partitioning and profile during soy protein isolate production. J. Food Sci., 69, C623-C631 (2004).