2014 Volume 20 Issue 5 Pages 927-938
Soybeans and the processed food, soybean milk, have important health functions. A cheese-like food produced from soybean milk resembles casein protein in its mouthfeel and physical properties. The intracellular soybean milk curdling enzyme-producing Saccharomyces bayanus SCY003 strain was screened. The curdling enzyme is an intracellular protease, a 45 kDa monomer protein that degrades soy protein subunits, produced by this strain. The optimum temperature and pH of enzyme activity are 50°C and pH 7.5, respectively. The protease degrades β-conglycinin and parts of glycinin in soy protein to α’, α and β subunits. The rheological characteristics of the resultant curd were assessed, and revealed that elasticity differed from curd produced by glucono δ-lactone, which is used in tofu production. A new soy protein food having health-supporting functions will be developed using this protease.
Soybeans contain 30 – 40% protein and 20% crude lipid. Their proteins provide rich nutrition; digestibility-corrected amino acid (PDCAAS) values of isolated soybean protein are equivalent to those of animal meat protein (Endres 2001). Also, isoflavone-containing soybeans provide multiple effects (Larkin et al. 2008). In fact, soybeans and its product, soybean milk, have many health-supporting functions. Soybean milk is increasingly consumed throughout the world for its health benefits and because, in contrast to bovine milk, it contains no cholesterol. Yogurt-like foods and cheese-like foods made from soybeans are consumed by people who are concerned about health issues or allergies related to bovine milk. Tofu, soybean curd from soybean milk, resembles cheese or yogurt made from milk curd of cows or other mammals. Tofu is eaten throughout Asia and is produced with a combination of manganese dichloride (MgCl2) and calcium dichloride (CaCl2) as Nigari, or more recently glucono δ-lactone (GDL) has been added during commercial production. However, the cheese-like mouthfeel and physical properties are not identical to those of the protein casein. Unlike casein in bovine milk, enzymatic curdling of soybean milk produces poor flavor and texture. For that reason, the commercial use of enzymes such as bromelain, ficin, and papain for the curdling of soybean milk has not been successful (Arima et al. 1968; Khan et al. 1979; Nouani et al. 2009). Thus, it is not yet a viable alternative to dairy foods (Park et al. 1985; Park et al. 1989; Qua et al., 1981).
The objectives of this investigation were to screen and identify specific food yeast strains (Saccharomyces sp.) that produce a soybean milk curdling enzyme and to purify the enzyme using chromatographic procedures.
Materials and chemicals The materials and chemicals used were as follows: soybean milk was purchased as a commercial product, and modified soybean milk was obtained from Kibun Food Chemifa Co., Ltd. Yeast extract (Cat.# 15838-45, Nacalai Tesque, Kyoto), polypeptone cat# 392-02115, Nihon Pharmaceutical Co., Ltd Tokyo), casein (Cat.# 030-01505, Wako Pure Chemical Industries, Ltd., Osaka) and glucose (Cat.# 049-31165, Wako Pure Chemical Industries, Ltd.) were purchased and used for cultivation of yeast. Chemicals without an indicated vendor were purchased from Wako Pure Chemical Industries, Ltd., while other specific chemicals are indicated by a vendor in the text.
Preliminary screening of soybean milk curdling enzyme-producing yeasts The process for screening soybean milk curdling yeasts is presented in Fig. 1. The yeast strains (1345 strains) stored in the laboratory were screened using soybean milk agar plate medium containing 5.0% soybean milk (modified soybean milk), 0.1% glucose, and 1.5% agar. The glucose and agar mixture was sterilized at 115°C for 10 min. Then, sterilized purchased commercial soybean milk was added aseptically. The strains were inoculated by streaking onto the plate surface, and then incubated at 30°C for 7 days. After cultivation, the diameter of clear zones was measured using calipers. Yeast strains that produced a clear zone were selected.
Screening for soybean curdling yeasts.
Secondary screening of curdling soybean milk enzyme-producing yeasts The screened strains (103 strains) were pre-cultured in GYP (1.0% glucose, 0.5% yeast extract, and 0.5% polypeptone) medium at 30°C for 1 day. The cultivated strains were inoculated to the soybean milk medium containing 90% commercial soybean milk, 0.1% glucose, and 100 mM phosphate buffer (pH 7.0). The glucose and 100 mM phosphate buffer mixture solution were sterilized at 121°C for 15 min. Then, the purchased soybean milk was added to them aseptically. The soybean milk medium was incubated at 30°C for 24 hr. When curdling occurred, the whey pH was measured by pH meter (Horiba Ltd. Kyoto). The production of yeast curd and higher pH (higher than pH 5.90) were screened for in the secondary screening.
Yeast identification Isolated yeasts were classified taxonomically and identified according to methods described in earlier studies (Kartzman and Fell 1998). The yeasts were inoculated into a yeast nitrogen base medium (Cat.# 239210; Difco Laboratories, Detroit, MI, USA) supplemented with 0.5% of each carbon source: sugar or organic acid as glucose, galactose, sucrose, maltose, raffinose, trehalose, lactose, melibiose, cellobiose, melezitose, starch, d-xylose, l-arabinose, d-ribose, l-rhamnose, erythritol, d-mannitol, salicin, inositol, dulcitol, ethanol, d-sorbitol, disodium succinate, and trisodium citrate. The yeast was inoculated onto yeast carbon base medium (Cat.# 239110; Difco Laboratories) with added sodium nitrate solution. Sugars were purchased from Wako Pure Chemical Industries, Ltd. The primers used for amplification and sequencing of 18S rRNA-encoding genes were those described by Suh and Nakase (1995). The PCR products were sequenced using a kit (ABI Prism Big Dye Terminator Cycle Sequencing Ready Reaction; Applied Biosystems, Foster City, CA, USA). Analyses of DNA sequence reactions were performed using a sequencer (3130; Applied Biosystems).
Yeast cultivation Glucose-Yeast extract (GY) base medium (final concentrations: 1% glucose, 0.5% yeast extract) was sterilized at 121°C for 15 min, and commercial soybean milk (1%) or sterilized (110°C, 10 min) casein solution (final concentration 1.0%) was added to the GY base medium under aseptic conditions for the preparation of GY-soybean milk medium or GY-casein medium. Furthermore, recaptured yeast cells were inoculated with GY-soybean milk medium, GY-casein, or 0.67% Difco Yeast Nitrogen Base containing 1% glucose. Cells were cultivated at 30°C for 6 days.
Preparation of the enzyme solution The screened yeast, S. bayanus SCY003 was inoculated into 1 L of GY-casein (final concentrations: 1% glucose, 0.5% yeast extract, 1.0% casein) and cultivated at 30°C for 6 days. The cultured medium was centrifuged at 4000 x g for 10 min. The precipitated yeast was suspended in 10 mL of phosphate-citrate buffer (50 mM, pH 7.0, containing 0.5% NaCl, 2 mM 2-mercaptoethanol, Cat.#131-14572, Wako Pure Chemical Industries). The yeast was subsequently crushed by mortar and pestle within glass beads (BZ-02; AZ One, Osaka). The mixture was centrifuged at 10000 x g for 10 min, and the supernatant was used as the intracellular crude enzyme solution. The supernatant was used to assay activity directly as the extracellular crude enzyme solution.
Soybean milk curdling test and curdling activity Soybean milk curdling test/activity was conducted using the method described (Arima et al. 1967) for modified soybean milk from bovine milk-curdling activity, as described below. Soybean milk (1 mL) was reacted with a 0.02 mL enzyme solution at 40°C for 1 hr in a 1.5 mL micro-tube. The mixture was centrifuged at 400 × g for 10 min. The supernatant was removed gently using a Pasteur pipette. The weight of the precipitate was measured using a chemical balance (New Classic ML; Mettler Toledo International Inc., Schwerzenbach, Switzerland). One curdling unit expressed the ratio of curdling (%) from 1 mL of soybean milk at 40°C for 1 hr.
Assay of proteolytic activity Proteolytic activity was measured in duplicate using a commercial kit (Pierce fluorescent protease assay kit; Thermo Fisher Scientific, Waltham, MA, USA). The protocol was modified as follows. Three types of proteases—acid, neutral, and alkaline—were assayed in three different buffers (50 mM phosphate-citrate pH 3.0 as acid protease; phosphate-NaOH buffer pH 6.0 as neutral protease; pH 8.0 as alkaline protease), with fluorescein isothiocyanate-labeled casein (FTC). FTC solution (100 µL) and the prepared enzyme extractions (20 µL) were mixed and incubated for 1 h at 40°C. Fluorescence was measured using a 485 nm excitation wave and a 535 nm emission wave (Genios; Tecan Group Ltd., Männedorf, Switzerland). The proteolytic activity was expressed as one unit being equal to the amount of trypsin (1 ng · mL−1) proteolysis in the FTC solution.
Peptidase Peptidase (carboxypeptidase) activity was determined by the increase in ninhydrin after hydrolysis of benzyloxycarbonyl-glutamyl-tyrosine (pH 7.5) at 40°C according to Ichishima (1972). One unit of peptidase activity was defined as the amount of enzyme necessary to liberate 1 mM of tyrosine per minute at pH 7.5 and 40°C.
Enzyme protein assay method Protein was assayed using the Bradford method (1976). Coomassie Brilliant Blue solution (containing 0.1 g Coomassie Brilliant Blue G (Cat.# B0770; Sigma-Aldrich Corp. St. Louis, Mo, USA), 50 mL ethanol, and 100 mL phosphoric acid in 200 mL) was added to the sample solution. Then, the mixture was measured using a spectrophotometer (595 nm, V-630; Jasco Inc., Tokyo). A standard curve was generated using bovine serum albumin (Cat. # 05482; Sigma-Aldrich Corp.) at 10 – 100 [g · mL−1.
Chromatography of proteolysis enzymeIon-exchange chromatography The enzyme was precipitated to between 30 – 40% saturation of ammonium sulfate. The precipitate was re-dissolved in 5 mL of 50 mM phosphate buffer, pH 8.0, containing 1 mM dithiothreitol (DTT, Cat. #D0632; Sigma-Aldrich Corp.). The solution was dialyzed overnight at 4°C using 1 L of 50 mM phosphate-NaOH buffer, pH 8.0, and 2 mM DTT. Ion-exchange chromatography was applied to the enzyme solution using a DEAE column (25 mm × 300 mm; Macro-Prep support; Bio-Rad Laboratories Inc., Hercules, CA, USA). The protein was eluted using a 0 – 1 M NaCl linear gradient at a flow rate of 2 mL min−1. The enzyme solution was fractionated at 4 mL per fraction.
Gel filtration chromatography Fractions containing the highest level of activity were pooled and re-precipitated using 80% saturation of ammonium sulfate. The precipitate was re-dissolved in 2 mL of 50 mM phosphate-NaOH, pH 8.0. Then, the molecular weight of the enzyme was analyzed using gel filtration chromatography (10 mm × 350 mm, P-100 gel; Bio-Rad Laboratories Inc.) with various molecular weight standards (myosin, 200 kDa; serum albumin, 66.2 kDa; ovalbumin, 45.0 kDa and trypsin inhibitor, 21.5 kDa). Subsequently, phosphate-NaOH, 50 mM, pH 8.0 at 0.1 mL min−1 was used as the eluent. The enzyme solution was fractionated at 2 mL per fraction.
Effect of metal ions and inhibitor The following metal ion solutions (1 mM) were added to each enzyme-substrate solution: KCl, NaCl, MgSO4, ZnSO4, CuSO4, CoSO4, FeCl2, CaCl2, HgCl2 and iodoacetamide (ICH2COOH, Cat.# I4386; Sigma-Aldrich Corp.). The following inhibitors (1 mM) were added to the solution: EDTA (Cat.# E5134; Sigma-Aldrich Corp.), azide, sodium azide (Cat.# 195-11092; Wako Pure Chemical Industries Ltd.); NBS, N-bromosuccinimide (Cat.#021-07232; Wako Pure Chemical Industries Ltd.); NEM, N-ethylmaleimide (Cat.# 058-02061; Wako Pure Chemical Industries Ltd.); and EGTA, O,O-bis(2-aminoethyl) ethylene-glycol-N,N,N',N'-tetra-acetic acid (Cat.# 348-01311; Wako Pure Chemical Industries Ltd.).
Optimum pH and temperature of protease activity Optimum pH was assayed at pH 4.0 – 8.0 using 50 mM phosphate-citric buffer or phosphate-NaOH. To determine the optimum pH of enzyme activity, 0.1 mL of FTC solution, which was dissolved in each phosphate buffer (50 mM, pH 4.0 – 8.0), and 20 µL of enzyme solution were reacted at 40°C for 60 min. To determine the optimum temperature of enzyme activity, FTC solution at pH 7.5 (0.1 mL) and 20 µL of enzyme solution were reacted at 15 – 70°C for 60 min. After reaction, fluorescence was measured using a 485 nm excitation wave and a 535 nm emission wave. The proteolytic activity was expressed as one unit being equal to the amount of trypsin (1 ng · mL−1) proteolysis in the FTC solution.
SDS-polyacrylamide gel electrophoresis Samples were separated on a 12.5% uniform gel (e-PAGEL E-T12.5L, Atto Corp., Tokyo). Electrophoresis was conducted as described by Laemmli (1970) with the following modifications. Samples (0.01 mL) were added to 0.01 mL of sample buffer (AE-1430 EzApply, Atto Corp., Tokyo) and then heated at 100°C for 3 min. Next, samples (10 µL) were added to each well. The gels were run at 20 mA with 1.5 mm gel thickness. Various molecular weight standards were used, protein standard mixture (Bio-Rad Laboratories, Inc.) or Ez Standard (Atto Corp., Cat.# AE-1440). The gel was stained with 0.25% Coomassie Brilliant Blue R-250 or the silver stain method (Silver Stain II Kit Cat. #291-50301; Wako Pure Chemical Industries Ltd.).
Curdling glycinin and β-conglycinin in soy protein Glycinin and β-conglycinin were extracted from soy protein (Cat#.162-10845; Wako Pure Chemical Industries Ltd.) according the process of Nagano and others (1992). The soy protein (100 g) was suspended in 1500 mL of distilled water at pH 7.5 adjusted with 0.1 M NaOH. Glycinin was separated by isoelectric precipitation at pH 6.4, while β-conglycinin was precipitated at pH 4.8, and the two fractions were freeze-dried (FDU-1110; Tokyo Rikakikai Co., Ltd., Tokyo). Each fraction as glycinin and β-conglycinin (50 mg) fractions was resolved to 1 mL of 50 mM phosphate buffer (pH 7.5). Next, the enzyme solution (0.1 mL) was added, and the mixtures were incubated at 40°C for 16 hr. After reaction, soybean milk curdling activity was assayed according to the preceding method.
Dissolution of curdled soybean milk by chemical solution Soybean milk (1.0 g) was poured into a glass vessel (32 mm inner diameter, 45 mm height) and 0.1 mL of enzyme solution was added to the soybean milk. The mixture was incubated at 40°C for 4 hr. Then, 0.1 mL of glucono δ-lactone (GDL, 3.0% solution) was added to the soybean milk (1.0 g) and incubated at 80°C for 1 hr. To each of the two curdled soybean milk samples was added 9 mL of chemical solution as 2% SDS solution, 4 M urea solution, and 10 mM 2-mercaptoethanol. Samples were held for 1 hr at room temperature and then centrifuged at 4000 × g for 10 min. Protein in the supernatant was assayed using the Lowry method (Lowry et al. 1951).
Rheological characteristics of enzyme-curdled soybean milk Soybean milk (10 mL) with added 0.01 mL of anti-foam (KM-72F; Shin-Etsu Chemical Co., Ltd. Tokyo) was poured into a glass cup (32 mm inner diameter, 45 mm height). Then, 1.5 mL of enzyme solution was added to the soybean milk, and the mixture was incubated at 40°C for 4 hr. As a control sample, GDL (0.3%) was added to soybean milk (10 mL) with 0.01 mL of anti-foam (KM-72F) added. Then, the mixture was incubated at 80°C for 1 hr. The curdled soybean milk samples were held at room temperature for 30 min. The rupture strength of the curdled soybean milk in the cups was measured directly using a creep meter (RE-3305; Yamaden Co., Ltd., Tokyo) with a 16 mm diameter plunger compressing 1 mm sec−1 at 80.0%.
Screening of yeasts producing soybean milk curdling enzyme Soybean milk-curdling yeasts were subjected to preliminary screening by the presence of a clarified zone in the soybean milk - glucose plate agar medium. The process of screening yeasts for soybean milk curdling is presented in Fig. 1. Results show that 1242 of 1345 yeast strains produced no clear zone on the plate medium. Several yeast strains (42 strains) produced less than 1 mm clear zone, 57 yeast strains produced 1.0 – 5.0 mm, and 4 yeast strains produced greater than 6 mm (data not shown). Soybean milk curdling by the yeast and the pH of the curdled soybean milk was measured (Table 1). Curdled soybean milk was not produced completely by 67 yeast strains, while 20 yeast strains produced curdled soybean milk at pH greater than 5.5. In particular, three yeasts curdled at pH greater than 5.90. Curdling occurred at pH 5.90 (SCY001), pH 6.05 (SCY002), and pH 6.38 (SCY003). The curdling activity of strain SCY003 was the highest among the strains; therefore, strain SCY003 was selected for further experiments. The morphology and physiological characteristics of SCY003 were observed. Cells are 1.5 – 6.5 mm and ovaloid. The .yeast, which buds by multi-budding reproduction, does not form pseudomycelia or pellicles in the liquid medium (data not shown), instead forming ascospores; therefore, it was identified as Saccharomyces sp. The glucose, galactose, sucrose, maltose, and raffinose in the medium as the carbon source were fermented by strain SCY003 (data not shown). The yeast also grew in and assimilated dulcitol, erythritol, ethanol, galactose, maltose, mannitol, melibiose, melezitose, raffinose, trehalose, D-sorbitol, and sucrose in the medium as carbon sources (data not shown). The yeasts grew in a vitamin-free medium. Furthermore, the strain did not grow in 0.01% cycloheximide, which agreed with S. bayanus (Kurtzman and Fell 1998). The 18S rRNA coding DNA was sequenced. Homology was assessed using the Basic Local Alignment Search Tool (BLAST; http://www.ncbi.nlm.nih.gov/BLAST/). The yeast showed 99% homology with S. bayanus (Accession No. AY046227). It was identified as S. bayanus through homology research and phenotypic testing.
| pH | Soy milk curdling | ||
|---|---|---|---|
| (++) | (+) | (−) | |
| > 6.50 | 0 | 0 | 7 |
| 6.49 − 5.90 | 1 | 2 | 43 |
| 5.89 − 5.50 | 0 | 17 | 17 |
| 5.49 − 5.00 | 2 | 11 | 0 |
| <4.99 | 3 | 0 | 0 |
Initial pH was 6 .70 . ++, coagulated very well; +, coagulated; −, non-coagulated
Purification of a protease as the soybean milk curdling enzyme The precipitation ratio from soybean milk was assayed to optimize the curdling activity method (Fig. 2). Commercial soy milk has dispersion stability due to the presence of oleosomes or the formation of aggregate soy proteins (Waschatko et al. 2012; Shimoyamada et al. 2008). Therefore, the precipitate was not produced from commercial soybean milk by low centrifugal gravity (400 x g). However, precipitation ratios were increased with the enzyme reaction period (R2= 0.9978).
Enzyme activity of soybean milk curdling.
S. bayanus SCY003 was cultured in three types of media, each with a different nitrogen source. The highest extracellular curdling activity (66.7 U · mL−1 · hr−1) was found in the GY-casein medium, followed by the GY-soybean milk medium (35.4 U · mL−1 · hr−1) (data not shown). No activity was observed in the yeast nitrogen-based medium. The intracellular curdling activity was 66.7 U mL−1 · hr−1, and the extracellular curdling activity was 0.8 U mL−1 · hr−1. Moreover, in a comparison of intracellular and extracellular proteolytic activities, the intracellular activity (535 U · mL−1 · min−1) was higher than that of the extracellular activity (12.6 U · mL−1 · min−1) (data not shown). In generally, yeast does not secrete enzymes. However, in this yeast strain, slight extracellular protease activity was found. In the case of nitrogen in the medium, it is thought that extracellular enzymes will be secreted for the uptake of the nitrogen source as amino acids are produced from proteins dissolved in the medium.
After cell homogenization and enzyme extraction, the enzyme protein was analyzed using ion-exchange chromatography and size-exclusion chromatography. The results are shown in Figs. 3 and 4. In ion-exchange chromatography, one main curdling activity peak was identified. The peak (fraction (fr.) 49) agreed with both protein and proteolytic activity determinations. The protein was eluted using buffer solution containing approx. 0.5 M NaCl. A larger peak of proteolytic activity was found around fr. 25. A small peak was found at fr. 49. The fraction showing proteolytic activity (fr. 25) did not correspond to the curdling activity. Therefore, it is considered that the former fractions are attributable to intense proteolytic enzymes and that the latter fractions are attributable to soybean milk curding enzymes.
Ion-exchange chromatography of soybean milk curdling enzyme.
Gel filtration chromatography of soybean milk curdling enzyme.
After re-precipitation, the sample was analyzed using gel filtration chromatography. A peak was found at fr. 11 – 14. The fractions agreed with soybean milk curdling activity, proteolytic activity and protein determination. No other fraction was associated with the activities. The chromatogram results demonstrate that the protease and soybean milk curdling enzyme have related activity. Maddox and Hough (1970) reported that proteolytic enzymes in Saccharomyces sp. were isolated in four fractions by ion-exchange chromatography. However, in this chromatogram, only two peaks of proteolytic enzymes were detected clearly.
Park et al. (1985) reported on microorganisms of soybean milk curd, and identified (Park et al. 1987; 1989) these as Bacillus sp. and that a neutral protease exhibited soybean milk curdling activity. Murata et al. (1987a and 1987b) reported on several proteases of vegetables and of microbial origin, excluding some kinds of curd protease, capable of curdling soybean milk. Yasuda et al. (1999 and 2002) also reported that a serine protease from Bacillus sp. curdled soy protein and analyzed the mechanism responsible for soybean milk coagulation. In contrast, few reports have described the curdling of soybean milk by proteases from yeasts such as S. bayanus sp.
The enzyme solution was analyzed using SDS-PAGE (Fig. 5a). A protein band of approximately 45 kDa was observed, which is the same as for other proteases. The molecular weight of the protease was measured using gel filtration chromatography (Fig. 5b). Results confirmed that the molecular weight was about 45 kDa. The protease was inferred to be a monomer protein. Many researchers have reported on proteases produced by yeasts, i.e., Candida albicans (Remold et al. 1968; Ruchel 1968; Negi et al. 1984), Candida humicola (Ray et al. 1992) and Saccharomyces cerevisiae (Jones 1991). Extracellular proteases produced by yeasts such as Candida sp. are 42 – 45 kDa (Ruchel 1968; Negi et al. 1984), while those of bacteria are 21 kDa (Cowan and Daniel 1982). In contrast, few reports describe the intracellular proteases produced by S. cerevisiae, although many intracellular proteases in the vacuole or other organelles are known to be related to Proteinase A (42 kDa; Jones 1991). The molecular weight of the soy protein curdling protease agreed with those of other yeasts of the Ascomycota. On the other hand, the 49-kDa protease of Mucor sp. (Nouani et al. 2009), which produces curdling in bovine milk, is larger than those produced by these yeasts.
Image of SDS-PAGE of soybean curdling enzyme.
Molecular weight determination of the curdling enzyme using gel filtration chromatography.
Characteristics of soybean milk curdling protease The optimum pH, temperature, and stability of the enzyme are presented in Fig. 6a and b. The optimum pH for protease activity was 7.5, and the optimum temperature was 50°C. A bovine milk curdling protease from Mucor pusillus showed an pH optimum of 5.0. The optimum pH values of many commercially available proteases are 5.9 – 6.7. The curdling activity decreases concomitantly with increasing alkalinity. Park and others (1987) reported that the optimum pH of a soybean milk curdling protease produced by Bacillus was 6.0. Regarding the present enzyme, the protease showed an optimum pH of 7.5, and an optimum temperature of 50°C. The enzyme also curdled soybean milk at pH 7.5 and 50°C. Commercial soybean milk sold in Japan is typically pH 7.0 – 7.2. The pH range agreed with their optimum pH range.
Optimum pH of soybean curdling enzyme.
Optimum temperature of soybean curdling enzyme.
Effects of metal ions and inhibitors on protease are presented in Table 2. Zinc, copper and mercury inhibited protease activity. The amino acid of the active site contains a cysteine residue because of inhibition by mercury (Springham et al. 1999). Magnesium inhibited 50% of the enzyme activity. Soybean contains high quantities of the mineral, magnesium (USDA Nutrient Database). Normally, enzyme activity may be inhibited in soy milk by magnesium ions; however, the enzyme was activated by its presence. It is thought that some metal ion levels were decreased by phytate chelation in the soy milk, allowing for enzyme activation.
| Metal ions & inhibitors | Concentrations | Relative activity (%) |
|---|---|---|
| Na+ | 1 mM | 101.5 |
| K+ | 1 mM | 105.9 |
| Mn2+ | 1 mM | 101.9 |
| Mg2+ | 1 mM | 46.2 |
| Fe2+ | 1 mM | 95.4 |
| Zn2+ | 1 mM | 23.6 |
| Co2+ | 1 mM | 47.9 |
| Cu2+ | 1 mM | 17.9 |
| Ca2+ | 1 mM | 100.9 |
| Iodo Acetate | 1 mM | 74.1 |
| Hg2+ | 1 mM | 34.1 |
| EGTA | 1 mM | 101.6 |
| 10 mM | 62.0 | |
| EDTA | 1 mM | 96.1 |
| 10 mM | 89.8 | |
| NEM | 1 mM | 98.5 |
| Azide | 1 mM | 100.4 |
| NBS | 1 mM | 97.6 |
| Control | - | 100.0 |
Furthermore, EGTA (10 mM) inhibited protease activity (62.0% of relative activity). The activity was not activated by metal ions, but was inactivated by mercury. Soybean milk-curdling enzyme (Park et al. 1987) was inhibited by zinc ions and mercury ions. These results agree with our data related to zinc and mercury. The report of Kim et al. (1984) on the extracellular protease produced by Enterobacter aerogenes agreed with our data. Its residual activity was 18% with mercury, and the protease was not activated by metal ions. Therefore, the protease is not a metalloprotease, a metal-dependent enzyme. Further, the amino acid of the active site contains a cysteine residue.
The mechanism of soybean milk curdling by the intracellular protease was investigated. The curdled soybean milk samples with and without added protease were treated with sample buffer solution, and then examined using SDS-PAGE (Fig. 7). The left lane shows the protein standards. The next lane (0 hr) shows soybean milk protein without protease reaction. The other lanes show soybean milk protein degraded for 4, 8, 12, 16, and 24 hr. Consequently, Lane 0 hr shows the α’ and α subunits of β-conglycinin (approx. 84 – 73 kDa), the A3 acidic subunit (approx. 40 kDa), other acidic subunits as A4, A1a, A1b and A2 (approx. 30–42 kDa) of glycinin, the β subunit (approx. 50 kDa), and basic subunits as B3, B1a, B1b and B4 (approx. 20 kDa) (Thanh et al. 1976, 1978). After the reaction, two bands as α’ and α subunit of β-conglycinin disappeared gradually. Of the glycinin subunits, the A3 acidic subunit band disappeared completely after 4 hr. A4, A1a, A1b and A2 also disappeared to a partial degree, the same as the A3 acidic subunit. Furthermore, peptides smaller than 20 kDa were detected on the gel. Namely, β-conglycinin as α, α’, and part of glycinin as A3 A4, A1a, A1b and A2 were degraded. Soybean protein became loosely curdled with the addition of other proteases from microorganisms or plants. The protein was degraded and low-molecular-weight peptides increased on the gel. Generally, 11S glycinin was related to the formation of a stiffer gel. Ono et al. (2004) reported that hydrophobic bonds and hydrogen bonds were related to the curdling of tofu. Utsumi et al. (1984) reported that the basic subunit and β-subunit formed macro-complexes by heating. The complexes were considered to form cores during tofu coagulation; the complexes were also reported to be wrapped in α, α' and acidic subunits (Ono et al. 2004).
Digestion of soy protein during enzymatic curdling of soybean milk.
According to our data, after the enzymatic curdling of soy milk, α, and α’ subunits were easily cleaved by the protease, whereas the basic subunit and β-conglycinin remained relatively unchanged. It is proposed that surface proteins such as α, and α’ subunits were degraded easily. Some degraded subunits such as α, α’, A3, acidic and basic subunits are regarded as related to the curdling of soybean milk.
The synthesis peptide substrates, Z-glutamyl-tyrosine, and casein, FTC, were reacted with the enzyme. The endo- and exo-peptidase activity were assayed in an effort to identify the mechanism of proteolytic activity. Results showed that the enzyme had 0.14 U · mg−1 protein as peptidase activity and 0.55 U · mg−1 protein as endo-type proteolytic activity (data not shown). Results also showed that the soybean milk curdling enzyme exhibited endo-type proteolytic activity.
Jones et al. (1991) reported three types of proteolytic enzymes in yeasts: cytosolic proteases, vacuolar proteases, and proteases located within the secretory pathway. They belong to aspartic type, serine, or metallo-type proteolytic enzymes with endo- or exo-type activity. Generally, metallo-type enzymes require metal ions such as zinc. The optimal condition for aspartic proteases is an acidic condition. Our enzyme did not require metal ions. Its optimum pH was 7.5, which is weakly alkaline. It is therefore considered to be a serine protease. These results agreed with the serine protease from Bacillus sp. that curdles soy protein (Yasuda et al. 2002). In future studies, we will ascertain the amino acid sequence in substrates cleaved by the enzyme using synthetic substrates.
Mechanisms and characteristics of soybean curdlingTo examine the mechanisms involved in soybean curdling, the curdled soybean milk was dissolved with 2-mercaptoethanol, urea, or SDS solutions. The resultant data show the relative ratio of protein (%). The 100% relative ratio indicates the amount of protein dissolution of non-curdled soybean in each solution (Fig. 8). Enzymatically-produced curd showed 47.3% dissolution in the 2-mercaptoethanol solutions, whereas GDL-produced curd showed 41.6% dissolution. In the urea solution, enzymatically-produced curd showed 93.8% dissolution, whereas GDL showed 40.3% dissolution. Both the enzyme- and GDL-produced curds dissolved with the SDS solution. 2-Mercaptoethanol cleaves disulfide bonds in proteins, urea cleaves hydrogen bonds or hydrophobic bonds, and SDS cleaves hydrogen bonds or hydrophobic bonds. Yasuda et al. (2002) reported that the serine protease from Bacillus sp. curdled soybean milk, generating protein bonds through mutual hydrophobic bonding.
Chemical dissolution of curdled soybean milk
The curdling activity of the enzyme against two protein fractions is shown in Fig. 9. Glycinin and β-conglycinin were isolated according to Nagano and others (1992) and curdling activity was assayed against the two substrates. Consequently, glycinin was curdled strongly (86.9 U · mL−1 · min−1 activity). On the other hand, β-conglycinin was curdled weakly, with an activity of 38.0 U · mL−1 · min−1. The data agreed with other reports (Yasuda et al. 2002; Mohri and Matsushita 1984; Guo and Ono 2005). Mori and others (1984) found that bromelain degrades 11S globulin during curdling, and that the bands representing acidic subunits and most basic subunits disappeared. Guo and Ono (2005) reported that soybean milk, which is rich in glycinin, was curdled strongly. This is in agreement with our data that the glycinin fraction was curdled and its curdling activity was approx. 50%. The enzyme exhibited proteolytic activity toward glycinin and β-conglycinin (Fig. 9). However, the enzyme made glycinin curdle without metal ions or GDL. Generally, glycinin is known to contain more sulfur-containing amino acids than β-conglycinin. According to Fig. 8, soymilk was curdled by hydrogen bonds or hydrophobic bonds. Also, some alkaline proteases such as subtilisin and chymotrypsin cleave hydrophobic amino acid residues. According to Fig. 9, the enzyme produced curd from soymilk mainly by enzymatic reaction against glycinin. Moreover, Choi reported that the α subunit in glycinin contained a hydrophobic sequence (Choi et al. 2005).
Curdling activity against the soy proteins glycinin and β-conglycinin.
It was considered that the α and α'subunits of β-conglycinin, in which the core proteins (mainly basic and β subunits) were wrapped, were degraded first, followed by the mutual recombination of hydrophobic peptides to curdle the soybean milk.
The stress-strain curves of curdled soybean milk are presented in Fig. 10. The vertical axis shows stress (N · m−2), which represents the internal forces of the sample curd pushing back against the strain. The horizontal axis shows the strain of the curd from the plunger of the creep meter. The curd sample can be broken by a force that exceeds the breaking point. The breaking load represents the hardness or softness of the curd sample. The breaking strain represents the resilience of the sample curd; a large value signifies a highly resilient sample. A large brittleness load indicates a brittle curd sample.
Stress-strain curves of curdled soybean milk.
The soybean milk was poured into a glass vessel and then curdled using the respective methods. After curdling, rheological analysis of the sample was conducted using a creep meter. The breaking point of curd curdled by GDL was at 34.7% strain. The breaking stress was 7980 (N · m−2). The brittleness point was 49.3% and 6650 (N · m−2), and the brittleness of the curd produced using GDL was 1327 (N · m−2).
The breaking point of the enzyme-curdled curd was 58.4% strain. The breaking stress was 10,900 (N · m‒2). The brittleness point was 81.2% and 10200 (N · m−2), and the brittleness of the enzyme-produced curd was 727 (N · m−2). The brittleness of the enzyme-produced curd was less than that produced using GDL, and the breaking point was greater than that of the GDL-produced curd. Furthermore, the curd had a sticky and chewy texture (data not shown). The enzyme-produced curd had greater resilience than normal tofu. Thus, the rheological properties of enzyme-produced curd differed from those of tofu.
Guo and Ono (2005) and Toyokawa and co-author (2008) reported on the relationship between the breaking stress of normal tofu and soy milk conditions, i.e., concentration of glycinin, protein or temperature. In future, research will be conducted on the relationship between soy milk conditions and the breaking stress of enzyme-curdled soy milk.
This report revealed that the novel protease from S. bayanus SCY 003 produced a soy product with a distinct texture that can be applied to the production of health foods, anti-milk-allergy foods, and others.
Six important conclusions were reached based on the findings of this study: (1) The curdling enzyme was an intracellular alkaline protease produced by S. bayanus SCY003. (2) The molecular weight of the alkaline protease monomer protein was 45 kDa. (3) The optimum pH and temperature for activity were, respectively, 7.5 and 50°C. (4) The enzyme includes a cysteine residue, as inferred from the fact that the enzyme was inhibited by mercury and copper. (5) The protease degraded α’, α, β subunits in β-conglycinin and part of glycinin in soybean milk, which was curdled by hydrophobic bonds. (6) In the protease-produced curd, brittleness was lower and the breaking point was higher than that of curd produced using GDL. The protease-produced curd exhibited elasticity resembling that of milk-casein cheese.
Acknowledgements We thank A-STEP, Adaptable and Seamless Technology, Transfer Program through Target Driven R&D by Japan Science and Technology Agency (JST) (AS231Z00291E) for financial support.
