Effect of Two-step Heat Treatment Processes on the Formation of Protein Particles and Oil Droplets in Soymilk

Abstract

Unheated soymilk (raw soymilk), which was prepared in the laboratory, was heated sequentially at two different temperatures. The proportion of protein particles in the soymilk sample subjected to a low-to-high two-step heating procedure (heating at 70 or 80 °C, then at 100 °C) was almost twice that of a high-to-low two-step heating procedure (heating at 100 °C, then at 70 or 80 °C). The size distribution of oil droplets in the soymilk for two-step heat treatment exhibited a polymodal distribution irrespective of the temperature combinations. However, the low-to-high two-step procedure resulted in an increase in the ratio of aggregated or coalesced oil droplets, in contrast to that for the high-to-low two-step procedure. When using the 7S (β-conglycinin) and 11S (glycinin) fractions separately, each fraction showed increased aggregation after two-step heat treatment, but a 1:1 mixture of the two fractions did not.

Introduction

Soymilk, a milky liquid made from soybean, is an important beverage and is also used as a raw material for the production of food products such as tofu (soybean curd) and yuba (sheets of soymilk skin). More than 300 000 kL of soymilk beverages are produced every year in Japan according to data from the Japan Soymilk Association (i). Soymilk has beneficial qualities because it contains many bioactive components such as proteins, isoflavones, and saponins (Kang et al., 2010). Soy proteins are reported to lower serum lipid and cholesterol levels (Chalvon-Demersay et al., 2017), and isoflavones and saponins exhibit estrogen-like activity (Zaheer and Akhtar, 2017) and inhibitory action against contact hypersensitivity (Nagano et al., 2018), respectively. Soymilk itself is also reported to reduce contact hypersensitivity (Nagano et al., 2019).

Soymilk beverages are manufactured by grinding imbibed soybean seeds, heating the resulting slurry and removing the insoluble residue (okara). Ono and colleagues (Ono et al., 1991; Shun-Tang et al., 1997; Chen and Ono, 2014) have described the composition and structure of soymilk and its changes due to heating. Soymilk contains oil droplets and protein particles. The oil droplets are coated with oleosin proteins derived from cotyledon oil bodies in soybean seeds. The protein particles consist of aggregates of subunits derived from mainly glycinin and β-conglycinin, which are major protein constituents in soybean (Giri and Mangaraj, 2012; Peng et al., 2016). Furthermore, the formation of the soymilk structures is dependent on the temperature of the heat treatment process. Raw soymilk, which has not been heated, contains emulsions where many of the proteins such as glycinin and β-conglycinin adhere to the oil body surfaces. By heating to temperatures over 65 °C, most of the proteins bound to the body surfaces are gradually liberated (Shun-Tang et al., 1997). Furthermore, β-conglycinin and glycinin are reported to denature at around 70–75 °C and around 90 °C, respectively (Zhang et al., 2004). With heat treatment, most of the released denatured proteins are reported to form protein particles (Ono, 2008; Chen and Ono, 2014; Yan et al., 2016). The oil bodies containing reduced amounts of globulins can be floated and hence separated from the liquid by ultracentrifugation. However, it has also been reported that the protein particles formed by heat treatment consist of core peptides and surface peptides. The core peptides include basic polypeptides of glycinin and β subunit of β-conglycinin, and the surface peptides include mainly acidic polypeptides of glycinin and α and α′ subunits of β-conglycinin (Ren et al., 2009; Chen and Ono, 2014). These changes in the proteins and oil bodies are believed to be dependent on the temperature of the heating process.

In the production of soymilk beverages, heating processes are repeated many times, for instance, at the milling, cooking, and sterilization stages. In addition, the products are heated sequentially at different temperatures, for example, in the sterilization process, soymilk may be preheated at a temperature lower than the boiling point (e.g., 80 °C) and then heated to over 100 °C. The heated soymilk is also cooled sequentially.

Liu et al. (2004) has reported that two-step heating, where raw soymilk is heated and kept at two separate temperatures, namely, 75 °C and then 95 °C, caused an increase in sample viscosity and an increased Young's modulus in the case of tofu. The authors speculated that selective thermal denaturation of the two types of globulins modified the aggregations of the denatured soybean proteins. The protein particles have also been reported to affect the formation of tofu (Guo and Ono, 2005). Therefore, the aim of this study was to clarify the effect of two-step heating on aggregate formation of soy globulins and oil droplets derived from the oil bodies.

Materials and Methods

Preparation of unheated soymilk (raw soymilk)    Soybean seeds (variety, Miyagishirome; 25 g) were soaked in water overnight at 4 °C. The soaked seeds, which were imbibed with 30 mL of water, were then milled at room temperature for 8 min at 10 000 rpm with 170 mL water in a blender (Teraoka Co., Osaka, Japan) and filtered using 5-ply gauze to provide raw soymilk. In this way, the soymilk was prepared using one part whole soybean and 8 parts water.

Heating of raw soymilk    A raw soymilk sample (30 mL) was placed in a plastic centrifuge tube and heated at two different temperatures. First, the sample was heated in a water bath at 70 or 80 °C for 15 min and then immediately heated in a boiling water bath for 15 min (low-to-high two-step heating); the temperature of the raw soymilk samples reached 95 °C after about 10 min of heating in boiling water. The reverse heating sequence (high-to-low) was also performed. Following each heating procedure, the soymilk was rapidly cooled in an ice bath.

Fractionation of soybean globulin proteins and heating    The 7S (mainly β-conglycinin) and 11S (mostly glycinin) globulins were extracted from whole soybean seeds and fractionated according to the method of Nagano et al. (1992). Each lyophilized sample or mixture of 7S and 11S was dispersed in 0.03 mol/L NaCl solution (10 mg/mL) and the pH was adjusted to 7.6. The prepared samples were heated in the same way as for the soymilk samples.

Centrifugation    The soymilk samples heated at various temperatures were centrifuged at 1 500 × g for 10 min using a refrigerated centrifuge (RG21; Koki Holdings Co., Ltd., Tokyo, Japan). The supernatant was removed by gently turning the tube upside down and the resulting precipitate was weighed to calculate precipitation as a fresh weight percentage. Furthermore, the supernatant was ultracentrifuged at 152 000 × g for 30 min using an ultracentrifuge (Optima max; Beckman Coulter, Inc., Brea, CA), as previously described, with slight modification (Ono et al., 1996). The fresh weight of the resulting precipitate was weighed.

Particle size distribution of soymilk    The soymilk samples were applied to a laser diffraction particle size analyzer (LS 13 320; Beckman Coulter, Inc.) to measure the particle size distribution of the samples. The measurement range was 0.04–2000 µm. Measurements were made in triplicate.

Statistical analysis    Data were subjected to one-way ANOVA. Significant differences among averages were determined by Bonferroni's multiple comparison using EXCEL Statics ver. 7.0 (ESMI Co., Ltd., Tokyo, Japan).

Results and Discussion

Effects of two-step heat treatment    In the factory, soymilk is typically heated at several different temperatures in a sequential manner. Thus, to simulate the effect of heating soymilk in a factory setting, a simple two-step heating procedure was adopted whereby raw soymilk samples were simply heated at two different temperatures in a water bath.

The soymilk samples heated under the one- or two-step procedures were then subjected to centrifugation at 1 500 × g; the resulting precipitation levels are shown in Fig. 1. In the case of one-step heating at 100 °C, the precipitation level amounted to about 1.56%. For the soymilk samples heated first to 70 °C and then to 100 °C (i.e., 70 °C→100 °C) or first to 80 °C and then to 100 °C (i.e., 80 °C→100 °C), the precipitation levels were 1.67 and 1.74%, respectively. In other words, the levels of precipitation obtained for the low-to-high two-step heating procedures were nearly equivalent to that for the one-step procedure. For comparison, reversal of the heating temperatures (high-to-low) was investigated. Sequential heating procedures for the soymilk were performed, first at 100 °C and then at 70 °C (i.e., 100 °C→70 °C) or 80 °C (i.e., 100 °C→80 °C), and the precipitation levels were 1.49 and 1.44%, respectively. These values were nearly equivalent to that obtained for one-step heating. Given that there was no clear difference among the different procedures, two-step heating was considered to have no substantial effect on the level of precipitation of soymilk relative to that for one-step heating.

Fig. 1.

Precipitation of soymilk after heating at two temperatures; the precipitate was obtained by centrifugation (1 500 × g).

Different letters indicate significant differences (p < 0.05, n ≥ 9).

Ultracentrifugation (152 000 × g, 30 min) was performed next in an attempt to study the effect of two-step heating on the formation of protein particles in soymilk (Ono et al., 1991; Shun-Tang et al., 1997). Following heat treatment of the soymilk samples as outlined above, the precipitates formed after ultracentrifugation (152 000 × g, 30 min) were weighed (Fig. 2). The amounts of precipitation after low-to-high two-step heating of the raw soymilk were 15.1% (70 °C→100 °C) and 16.8% (80 °C→100 °C), respectively. The values obtained were roughly twice that for one-step heating (8.2%; 100 °C). For the high-to-low temperature combinations (100 °C→70 °C and 100 °C→80 °C), the values obtained were almost equivalent to that for one step heating. Thus, protein particle formation was considered to be affected by the low-to-high two-step heating procedure, but hardly by the total heating time (30 min for two-step procedures). From these data, it can be inferred that the low-to-high two-step heating procedures accelerate the formation of protein particles in the heat-treated samples. It has been reported that soymilk prepared by sequential heat treatment at 100 °C and then 70 °C (100 °C→70 °C) showed an increase in precipitation after storage for about 2 weeks (Shimoyamada et al., 2008). The data obtained in the present study were obtained immediately after heating. Clearly, formation of insoluble aggregates may be dependent on storage conditions and time duration.

Fig. 2.

Protein particle fraction of soymilk after heating at two temperatures; the particle fractions was obtained by ultracentrifugation (152 000 × g).

Different letters mean significant differences (p < 0.05, n ≥ 10).

Next, the size distribution of oil body droplets was measured by a particle size analyzer. The soymilk samples heated by the one-step procedure showed a monomodal distribution with the peak centered on 0.4 µm (Fig. 3A), which is in good agreement with an earlier study (Shimoyamada et al., 2010). This distribution is considered to reflect a monomeric oil body. In contrast, all soymilk samples heated by the two-step treatments, regardless of the temperature combinations, exhibited polymodal distributions. Both of the soymilk samples heated under the high-to-low procedures (100 °C→70 °C and 100 °C→80 °C) had a major peak at around 0.3 µm and additional peaks at 2.0 and 9.0 µm. The distribution above 1 µm is considered to be aggregates of oil body droplets (Toda et al., 2007; Shimoyamada et al., 2010). The data of the present study suggest that prolonged heating alone accelerates the aggregate formation. Liu and Chang (2007) reported that prolonged heating of soymilk increased the sample viscosity. It is suspected, therefore, that the increase in sample viscosity probably relates to aggregate formation as observed in the present study. Furthermore, soymilk samples heated under the low-to-high two-step procedures (70 °C→100 °C or 80 °C →100 °C) showed drastic changes in particle-size distribution. That is, the main distribution ranged from 1 to 10 µm with a small peak at around 0.3 µm (Fig. 3B). The Z-average size of oil bodies in soymilk is reported to increase at the beginning of heating (0–4 min at 100 °C and 0–8 min for 70–90 °C) and then remains almost constant during further heating (Yan et al., 2016). The authors reported that the remaining proteins bound to the oil bodies after 0–8 min of heating were inhibited from further coalescence. According to the results of the present study, the first treatment at 100 °C is considered to have increased the size of oil bodies to the maximum extent because the soymilk was subjected to heating for about 5 min after the target temperature was reached. Subsequently, the second heating at 70 or 80 °C may have had little effect on the size of oil bodies. Conversely, for the low-to-high two-step procedure, heat treatment at 70 or 80 °C may have had a minimal effect on the growth of oil bodies, so further heating at 100 °C was necessary to accelerate their growth. This result implies that the combination and the temperature order during heat treatment in the factory as well as the laboratory is very important for the quality of soymilk products.

Fig. 3.

Particle size distributions for soymilk samples after heating at two temperatures.

A, dotted line, one-step heating (100 °C); solid line, two-step heating (100 °C→70 °C); broken line, two-step heating (100 °C→80 °C).

B, dotted line, one-step heating (100 °C); solid line, two-step heating (70 °C→100 °C); broken line, two-step heating (80 °C→100 °C).

Effect of two-step heat treatment on soybean 7S and 11S globulins    Two-step heating was considered to affect protein particle formation and aggregation of oil bodies in the soymilk. To examine the effect of two-step heat treatment on individual proteins, soybean protein was separated into 7S and 11S fractions. The fractions were dispersed in 0.03 mol/L NaCl and then heated under the one-step (100 °C) or two-step procedures (70 °C→100 °C and 100 °C→70 °C). The samples were then subjected to ultracentrifugation (152 000 × g, 30 min) and the obtained precipitates were weighed (Fig. 4). The amounts of precipitate obtained were much lower than those for the soymilk samples, and the differences were considered to be due to the much lower protein concentrations of the globulin dispersions (about 10 g/L) compared with the soymilk samples (about 50 g/L). As for the 7S fraction, heating at 100 °C showed a significantly higher precipitation level than for heating at 70 °C. The high-to-low two-step heating procedure (100 °C →70 °C) gave almost the same level of precipitation as that at 100 °C; however, only the low-to-high two-step heating (70 °C →100 °C) gave a significantly higher level of precipitation than the other high-to-low two-step (100 °C→70 °C) and the one-step heating procedures. A similar trend was observed for the 11S fraction, but the precipitation levels for both the one-step procedure at 100 °C and the low-to-high two-step procedure were higher than those for the 7S fraction, which is similar to the results of a previous paper (Guo et al., 2012).

Fig. 4.

Precipitation of globulin dispersions by ultracentrifugation (152 000 × g, 30 min) after one- or two-step heating.

Open bar, 7S; Hatched bar, 11S; Shaded bar, mixture.

Different letters mean significant differences (p < 0.05, n ≥ 7).

The precipitation levels for a 1:1 mixture of 7S and 11S after heating by the one-step and low-to-high two-step procedures were nearly equivalent and lower than those for either the 7S or 11S fraction alone. This result differed from that shown in Fig. 2. These results may be related to the fact that the globulin concentrations were much lower than that in the soymilk samples and that there were no oil bodies. As another possibility, since Miyagishirome is reported to have a larger amount of 11S than 7S (Nishinari et al., 1991), the data of two-step heating for the soymilk sample may show primarily the effect of 11S.

Selective heat denaturation of soybean globulins, namely β-conglycinin and glycinin, under two-step heating has been reported to cause an increase in the apparent breaking strength and Young's modulus of tofu-gel (Liu et al., 2004). Furthermore, Liu and Chang (2007) reported that two-step heating increases the viscosity of the resulting soymilk. These behaviors of soymilk subjected to two-step heating are essentially equivalent to the data of this study. However, the 7S or 11S globulin fraction did show a similar increase in precipitate formation under two-step heating, although the data were obtained under the condition where the protein concentration was much lower than that in soymilk. Further studies should be conducted without the constraints of selective heat denaturation to clarify the mechanism of two-step heat treatment of soymilk.

From the above data, two-step heating of soymilk, where the sample was sequentially heated first at a low temperature and then at a higher temperature, accelerated the formation of protein particles, and aggregation and/or coalescence of the oil bodies. This finding may be used as a basis to support the design and optimization of heat treatment processes in soymilk production

References

•  Chalvon-Demersay,  T.,  Azzout-Marniche,  D.,  Arfsten,  J.,  Egli,  L.,  Gaudichon,  C.,  Karagounis,  L.G., and  Tomé,  D. (2017). A systematic review of the effects of plant compared with animal protein sources on features of metabolic syndrome. J. Nutr., 147, 281-292.
•  Chen,  Y. and  Ono,  T. (2014). Protein particle and soluble protein structure in prepared soymilk. Food Hydrocol., 39, 120-126.
•  Giri,  S.K. and  Mangaraj,  S. (2012). Processing influences on composition and quality attributes of soymilk and its powder. Food Eng. Rev., 4, 149-164.
•  Guo,  S.-T. and  Ono,  T. (2005). The role of composition and content of protein particles in soymilk on tofu curding by glucono-δ-lactone or calcium sulfate. J. Food Sci., 70, C258-C262.
•  Guo,  J.,  Yang,  X.-Q.,  He,  X.-T.,  Wu,  N.-N.,  Wang,  J.-M.,  Gu,  W., and  Zhang,  Y.-Y. (2012). Limited aggregation behavior of β-conglycinin and its terminating effect on glycinin aggregation during heating at pH 7.0. J. Agric. Food Chem., 60, 3782-3791.
•  Kang,  J.,  Badger,  T.M.,  Ronis,  M.J.J., and  Wu,  X. (2010). Non-isoflavone phytochemicals in soy and their health effects. J. Agric. Food Chem., 58, 8119-8133.
•  Liu,  Z.S.,  Chang,  S.K.C.,  Li,  L.T., and  Tatsumi,  E. (2004). Effect of selective thermal denaturation of soybean proteins on soymilk viscosity and tofu's physical properties. Food Res. Int., 37, 815-822.
•  Liu,  Z.-S. and  Chang,  S.K.C. (2007). Soymilk viscosity as influenced by heating methods and soybean varieties. J. Food Process. Preserv., 31, 320-333.
•  Nagano,  T.,  Hirotsuka,  M.,  Mori,  H.,  Kohyama,  K., and  Nishinari,  K. (1992). Dynamic viscoelastic study on the gelation of 7S globulin from soybeans. J. Agric. Food Chem., 40, 941-944.
•  Nagano,  T.,  Katase,  M., and  Tsumura,  K. (2018). Dietary soyasaponin attenuates 2,4-dinitrofluorobenzene-induced contact hypersensitivity via gut microbiota in mice. Clin. Exp. Immunol., 195, 86-95.
•  Nagano,  T.,  Katase,  M., and  Tsumura,  K. (2019). Impact of soymilk consumption on 2,4-dinitrofluorobenzene-induced contact hypersensitivity and gut microbiota in mice. Int. J. Food Sci. Nutr., 70, 579-584.
•  Nishinari,  K.,  Kohyama,  K.,  Zhang,  Y.,  Kitamura,  K.,  Sugimoto,  T.,  Saio,  K., and  Kawamura,  Y. (1991). Rheological study on the effect of the A₅ subunit on the gelation characteristics of soybean proteins. Agric. Biol. Chem., 55, 351-355.
•  Ono,  T. (2008). The mechanism of soymilk and tofu formation from soybean, and the factors affecting the formation. J. Jpn. Soc. Food Sci. Technol. (Nippon Shokuhin Kagaku Kogaku Kaishi), 55, 39-48 (in Japanese).
•  Ono,  T.,  Choi,  M.R.,  Ikeda,  A., and  Odagiri,  S. (1991). Changes in the composition and size distribution of soymilk protein particles by heating. Agric. Biol. Chem., 55, 2291-2297.
•  Ono,  T.,  Takeda,  M., and  Shuntang,  G. (1996). Interaction of protein particles with lipids in soybean milk. Biosci. Biotechnol. Biochem., 60, 1165-1169.
•  Peng,  X.,  Ren,  C., and  Guo,  S. (2016). Particle formation and gelation of soymilk: Effect of heat. Trends Food Sci. Technol., 54, 138-147.
•  Ren,  C.,  Tang,  L.,  Zhang,  M., and  Guo,  S. (2009). Structural characterization of heat-induced protein particles in soy milk. J. Agric. Food Chem., 57, 1921-1926.
•  Shimoyamada,  M.,  Kusaka,  K.,  Mogami,  S.,  Tsuzuki,  K.,  Honda,  Y., and  Asao,  H. (2010). Characterization of soymilk prepared from non-imbibed soybean seeds. Food Sci. Technol. Res., 16, 505-508.
•  Shimoyamada,  M.,  Tsushima,  N.,  Tsuzuki,  K.,  Asao,  H., and  Yamauchi,  R. (2008). Effect of heat treatment on dispersion stability of soymilk and heat denaturation of soymilk protein. Food Sci. Technol. Res., 14, 32-38.
•  Shun-Tang,  G.,  Ono,  T., and  Mikami,  M. (1997). Interaction between protein and lipid in soybean milk at elevated temperature. J. Agric. Food Chem., 45, 4601-4605.
•  Toda,  K.,  Chiba,  K., and  Ono,  T. (2007). Effect of components extracted from okara on the physicochemical properties of soymilk and tofu texture. J. Food Sci., 72, C108-C113.
•  Yan,  Z.,  Zao,  L.,  Kong,  X.,  Hua,  Y., and  Chen,  Y. (2016). Behaviors of particle size and bound proteins of oil bodies in soymilk processing. Food Chem., 194, 881-890.
•  Zaheer,  K. and  Akhtar,  M. H. (2017). An updated review of dietary isoflavones: Nutrition, processing, bioavailability and impacts on human health. Crit. Rev. Food Sci. Nutr., 57, 1280-1293.
•  Zhang,  H.,  Takenaka,  M., and  Isobe,  S. (2004). DSC and electrophoretic studies on soymilk protein denaturation. J. Therm. Anal. Cal., 75, 719-726.
• i) http://tounyu.jp/shared/PDF/database/seisan_researchH31.pdf (Nov. 19, 2019)