Yan Zhang, Sam K.C. Chang, and Zhisheng Liu
Abstract: Isoflavones impart health benefits and their overall content and profile in foods are greatly influenced at each step during processing. In this study, 2 soybean varieties (Prosoy and black soybean) were processed with 3 different grinding (ambient, cold, and hot grinding) and heating methods (traditional stove cooking, 1-phase UHT, and 2-phase UHT) for soymilk making. The results showed after cold, ambient, and hot grinding, the total isoflavones were 3917, 5013, and 5949 nmol/g for Prosoy; the total isoflavones were 4073, 3966, and 4284 nmol/g for black soybean. Grinding could significantly increase isoflavone extraction. The grinding process had a destructive effect on isoflavones and this effect varied with grinding temperature. Different heating methods had different effects on different isoflavone forms. Two soybean varieties showed distinct patterns with respect to the change of isoflavone profile during processing. Keywords: degradation, grinding, isoflavone, soymilk, UHT
Introduction Isoflavones, due to an array of health-promoting functions, are attracting more attention (Cohen and others 2000; Brouns 2002; Jenkins and others 2003). However, the isoflavone profile is greatly affected by the many processing steps that soy products undergo (Mahungu and others 1999; Kao and others 2004; Xu and Chang 2009). Soymilk, as a major soy product, is not well accepted because of its characteristic beany flavor (MacLeod and others 1988). In our previous study, it was proved that proper combinations of grinding and heating methods could effectively reduce some soy off-flavor components to desirable levels (Zhang and others 2012). However, it is still unknown how isoflavone profile is affected by these treatments. Even though the heating effect on soymilk isoflavones has been investigated extensively, there is scant literature dealing with ultra-high temperature (UHT) exposure. Most similar research has been done with soymilk under or equal to boiling temperature (Jackson and others 2002; Nufer and others 2009). Even at UHT temperature, most studies were conducted in model systems and/or at lab scale (Xu and others 2002; Chien and others 2005; Huang and others 2006). Hot and cold grindings have been studied widely to limit the activity of lipoxygenases, which are mainly responsible for the formation of typical offflavors in soymilk (Mizutani and Hashimoto 2004; Sun and others 2010). However, the effects of grinding methods on the extraction efficiency of isoflavones, and the heating influence when in conjunction with different grinding methods, have not yet been reported. Prabhakaran and Perera (2006) made a comparison between hot and ambient grinding without a soaking step. However, only β-glucosides and aglycones were investigated in their study. Protein can form complexes with isoflavones through various interactions and the association is dependent on protein content and the protein denaturation state (Boye 1999; Nufer and others 2009). As a result, the isoflavone extraction efficiency might be
affected by the grinding methods, which have been shown to greatly influence protein recovery and lead to some denaturation, when hot grinding was applied (Zhang and others 2012). In our current study, we used the Microthermics Direct/Indirect Steam Injection Processor (DIP; Microthermics, Inc., Raleigh, N.C., U.S.A.) to mimic the actual scenario of soymilk processing as reported by Yuan and Chang (2008). Two UHT treatments and 1 traditional stove cooking were chosen on the basis of sterilization value for trypsin inhibitor destruction. The 2-phase UHT method is commonly used in the soymilk industry. The objectives of this work were to determine how grinding methods would affect isoflavone extraction and how isoflavone profiles would change under traditional and UHT heating processes in conjunction with different grinding methods.
Materials and Methods Materials Two varieties of soybeans (Glycine max) were used in this study: Prosoy (harvested in 2009) and black soybean (harvested in 2006) were obtained from Sinner Brothers and Bresnahan Co. grown in Casselton, N.Dak., U.S.A. The soybeans were stored in a cool and dry air-conditioned room prior to use. All processing methods were replicated 3 times. Chemicals Daidzein, malonyldaidzin, malonylgenistin, malonylglycitin, and acetylglycitin were purchased from Nacalai USA (San Diego, Calif., U.S.A.). Glycitin, daidzin, genistein, and genistin were purchased from LC Laboratories (Woburn, Mass., U.S.A.). Glycitein was purchased from Sigma-Aldrich (St. Louis, Mo., U.S.A.). High Performance Liquid Chromatography (HPLC) grade methanol and acetonitrile were purchased from VWR International (West Chester, Pa., U.S.A.).
MS 20142017 Submitted 12/9/2014, Accepted 2/9/2015. Authors Zhang, Soymilk preparation Chang, and Liu are with Dept. of Food Science, Nutrition and Health Promotion, Soymilk was prepared as reported in our previous study (Zhang Mississippi State Univ, Starkville, MS 39759, U.S.A. Direct inquiries to author Teixeira (E-mail: [email protected]). and others 2012). In brief, 300 g soybeans were soaked in 1500 mL
cold water (4 °C for cold grinding) or room-temperature water
R C 2015 Institute of Food Technologists
doi: 10.1111/1750-3841.12839 Further reproduction without permission is prohibited
Vol. 80, Nr. 5, 2015 r Journal of Food Science C983
C: Food Chemistry
Isoflavone Profile in Soymilk as Affected by Soybean Variety, Grinding, and Heat-Processing Methods
Grinding and heat-processing methods . . .
C: Food Chemistry
(20 °C for ambient and hot grinding) for 16 h. The hydrated beans were drained and ground with cold water (4 °C), water at 20 °C and 80.5 °C with a bean-to-water ratio of 1:10 (w/w). A portion each of soaked beans and soaking water was freezedried and stored at −20 °C for further analysis. The remaining soaked beans were ground at 10,000 rpm with a New Hartford 1-gallon blender (model CB-2-10; Conn., U.S.A.). To maintain a constant temperature while grinding, the blender was covered with insulating foam (1-in. thickness). For cold grinding, ice (−2.5 °C) and precooled (4 °C) water in the proper ratio (1:3.3) were used to result in an initial grinding temperature of about of 2 °C. For hot grinding, prewarming was done by rinsing the blender twice with boiling water and, immediately, soaked soybeans and boiling water were added. The initial grinding temperatures were recorded as the temperatures at 10 s after grinding; and, in fact, the final temperatures after 3-min grinding were about 2 °C higher than the target temperatures. All these temperatures were determined in preliminary tests. After grinding, the soymilk was manually filtered through a piece of muslin cloth with consistent pressing.
Traditional stove heating of soymilk The method reported by Yuan and Chang (2007) was employed. In brief, l L soymilk was put in a small pot, which was placed in a larger pot with boiling water on a stove. After the soymilk reached 90 °C, the small pot was switched to the stove surface and heated to boiling, from which point the soymilk was maintained boiling with continual stirring for 20 min. Then the small pot was cooled in an ice bath to room temperature. Soymilk was freeze-dried and stored at −20 °C for later analysis. UHT processing of soymilk In this study, Microthermics Direct/Indirect Steam Injection Processor (DIP, Microthermics, Inc., Raleigh, N.C., U.S.A.) was used as reported in our previous study (Zhang and others 2012). In brief, 2 sets of heating temperature and time combinations were applied: 140 °C/5 s (one-phase UHT, F0 = 6.62); 120 °C/80 s + 140 °C/4 s (2-phase UHT, F0 = 6.35). These 2 UHT methods had the similar sterilization power, F0 , calculated on the basis of Z = 10 for bacterial spore sterilization. In the heating tube, the heating medium (steam) was in direct contact with soymilk. The Microthermics Processor was equipped with a vacuum chamber (50 kPa), which was originally designed for the purpose to cool the heated soymilk and to remove the condensed water from the injected steam. However, during the vacuum evaporation process, some volatile compounds were also removed. Extraction of isoflavones Extraction of isoflavones was conducted according to the method of Xu and Chang (2009) with a slight modification. In brief, freeze-dried soymilk and soaking water samples were pulverized with pestle and mortar to fine powder. Freeze-dried soybeans and okara were ground with a coffee blender and passed through a 60-mesh screen. About 1 g of sample powder was accurately weighed into a 15-mL centrifuge tube. Five milliliters of acetonitrile, 4.75 mL DDW, and 0.25 mL 0.5 mg/g 6-hydroxyflavone (in 80% methanol) were added and vortexed. The mixture was shaken in a shaker for 2 h at room temperature and centrifuged at 5500 rpm for 20 min in a centrifuge (Beckman Coulter Ltd., Palo Alto, Calif., U.S.A.). The supernatant was transferred to a 125-mL flat-bottom flask and evaporated at 35 °C until dryness in a rotary evaporator. The residue was dissolved in 5 mL of 80% methanol and filtered through a 0.2-μm syringe filter into vials and were kept at −80 °C until HPLC analyzed. C984 Journal of Food Science r Vol. 80, Nr. 5, 2015
HPLC analysis of isoflavones HPLC analysis was run according to the method of Xu and Chang (2009) with some modification. An Agilent Technologies 1200 series system equipped with a YMC-pack ODS-AM-303 C18 reversed-phase column (250 mm × 4.6 mm i.d., 5 μm) was used. The UV detector was set at 262 nm and column temperature was set at 35 °C. Mobile phase A was 0.1% glacial acetic acid in water, mobile phase B was 0.1% glacial acetic acid in acetonitrile. Initially, B was set 15% for 5 min. Then B was increased to 29% up to 36 min, increased to 35% to 44 min, then increased to 50% up to 46 min, and held until 56 min. Then B was recycled to 15% until 58 min and held for 60 min. Peaks of isoflavones were identified by comparing the retention times of the authentic standards and were subsequently quantified by preparing a calibration curve of standards. Individual isoflavones were expressed as μg/g of dry material; and for the purposes of comparison, total isoflavones were expressed as nmol/g of dry material on the basis of their individual molecular weights. Statistical analysis Soymilk processing was conducted in triplicate and the following HPLC analyses were completed in duplicate. The data were subjected to analysis of variance with SAS 9.1 package (SAS 2005). Significant differences among variables were determined by Duncan’s multiple range test (α = 0.05). Data were expressed as means ± SD (n = 6).
Results and Discussion Effect of grinding methods on isoflavone profile and content in soymilk As presented in Figure 1, significant differences (P < 0.05) were found in individual isoflavone types and total insoflavones among the 3 grinding methods. However, the 2 varieties showed different response to the grinding methods. For Prosoy soymilk, in most cases, except aglycones, hot grinding generated significantly (P < 0.05) higher individual and total isoflavones than the other 2 grinding methods. Meanwhile, ambient grinding yielded significantly (P < 0.05) higher level of isoflavones than cold grinding. For example, in raw Prosoy soymilk, the total isoflavone content from cold, ambient, and hot grinding methods were 3917, 5013, and 5949 nmol/g, respectively. For soymilk from black soybeans, in most cases, ambient grinding produced the lowest isoflavone content and hot grinding and cold grinding exhibited similar efficiency. For example, in raw soymilk from black soybeans, the total isoflavone content from cold, ambient, and hot grinding methods were 4073, 3966, and 4284 nmol/g, respectively. The distinct behaviors of 2 varieties in response to grinding methods suggest different compositions and structures. Aglycones exhibited a distinct pattern in responding to grinding methods; that is, for both varieties, ambient grinding resulted in the highest level of aglycones and hot grinding generated significantly (P < 0.05) higher aglycones than cold grinding. The lower level of aglycones from hot grinding can be explained by the instant inactivation of β-glucosidase during hot grinding and this phenomenon was also observed by Prabhakaran and Perera (2006). Until now, we only found one paper comparing hot grinding (95 °C) and cold grinding (45 °C) with regard to isoflavone availability in soymilk (Prabhakaran and Perera 2006). In their study, the soaking step was omitted, but hot grinding still showed much higher isoflavone-producing efficiency compared with cold grinding in terms of aglycones and β-glucosides. Wang and Murphy (1996)
found that heating slurry after grinding at 95 °C for 7 min prior to pressing could result in about 12% loss in okara and almost 90% of the isoflavones present in soymilk. The 90% isoflavone recovery was very high compared with our results and those of others. Using soy protein isolate, Malaypally and Ismail (2010) also found that heating could significantly facilitate the recovery of isoflavones in soymilk. Therefore, on the basis of combined results, we suppose that heating before separation between insoluble okara and soymilk might advance the recovery of soflavones in soymilk. And this is perhaps the main reason for the relatively higher isoflavone content from hot grinding as observed in our study. In fact, different recoveries resulted mainly from different solubilities of different isoflavone forms responding to different grinding temperatures. Jackson and others (2002) found in ambient grinding that the percentage contents of aglycones, βglucosides, and acetylglucosides in okara were higher than those in soymilk, but malonylglucosides showed an opposite trend. As shown in Figure 2, our results showed that the distributions of each single isoflavone form in okara and soymilk are greatly different for different grinding methods. For example, in raw Prosoy soymilk, the most abundant isoflavone form, the combined content of malonylglucosides in soymilk and okara from cold
A
1400
grinding was 5100 nmol/g, much higher than that from hot grinding of 4777 nmol/g (data not shown). However, only 56% of malonylglucosides remained in soymilk from cold grinding in comparison with 82% from hot grinding. Consequently, hot grinding yielded much higher total isoflavone compared with cold grinding. Apart from the temperature-induced effect on isoflavone recovery in soymilk, Wang and Murphy (1996) attributed isoflavone in soymilk to the association between isoflavones and soluble proteins in soymilk. However, our previous study demonstrated that the protein recovery from hot grinding was lower than that from ambient grinding and a little higher than that from cold grinding (Zhang and others 2012). The difference of total isoflavone was likely due to different solubility of different individual isoflavones in soymilk under different grinding temperatures.
Effect of heating methods on isoflavone profile and content Except for one-phase UHT, all other heating processes significantly (P < 0.05) reduced the contents of malonylglucosides with stove cooking reducing the most (Figure 1). Our result was in agreement with the report by Xu and Chang (2009) who also found that stove cooking reduced malonylglucosides
D 50
a
a
45
1200
40 1000
b
800
µg/g
a
600 400
a
c
b
bb
30
c
b
a
c
a
b
35
a
b
a
a
a
a
b
µg/g
a
ab
25 20 10 5
a
a
15
c
a
c
c
b
a
b
c
a a
b
a a
b
c b
b
c
b
b
a
0 200
Raw
SC
Raw
SC
OP
TP
Raw
Prosoy Cold grinding Ambient grinding
B
OP
TP
Raw
SC
Prosoy
0
a
2500
b
2000
b
c 1500
a
µg/g
Cold grinding
TP
Ambient grinding
Hot grinding
Hot grinding
a a b
c
OP
TP
Black soybean
a a b
SC
OP
Black soybean
E b b b
1000
7000
a
a a a
a a a
5000
nmol/g
4000
a
a
6000
a
a
c
a
b
b
b
a
c
b
c
a
ab b
a
bb
a
a a a
b
ab a
3000 500
2000 1000
0 Raw
SC
OP
TP
Prosoy Cold grinding
C
Raw
OP
TP
0 Raw
Black soybean Hot grinding
Ambient grinding
SC
OP
TP
Prosoy Cold grinding Ambient grinding
Raw
SC
OP
TP
Black soybean Hot grinding
250
a
150
a
100
b 50
a
a
200
µg/g
SC
b
a
ab
b
c
b
b
c
c
a b
b
a
c
c
a b
b c
0 Raw
SC
OP
TP
Prosoy Cold grinding
Raw
SC
OP
TP
Black soybean Ambient grinding
Hot grinding
Figure 1–Effect of grinding methods, heating methods, and variety on (A) β-glucosides, (B) malonylglucosides, (C) aglycones, (D) acetylglycitin, and (E) total isoflavones (SC, stove cooking; OP, one-phase UHT; TP, two-phase UHT). Bar data are expressed as mean ± standard deviation (n = 3) on dry weight basis. Values marked with different letter above the bars are significantly different within the same heating treatment and variety group (p < 0.05). Effect of grinding methods, heating methods, and variety on (A) β-glucosides, (B) malonylglucosides, (C) aglycones, (D) acetylglycitin, and (E) total isoflavones (SC, stove cooking; OP, one-phase UHT; TP, 2-phase UHT). Vol. 80, Nr. 5, 2015 r Journal of Food Science C985
C: Food Chemistry
Grinding and heat-processing methods . . .
Grinding and heat-processing methods . . .
C: Food Chemistry
even more than direct and indirect UHT at 143 °C for 60 s. Malonylglucosides are very heat labile and are prone to convert to β-glucosides and acetylglucosides under moist heat (Chien and others 2005; Xu and Chang 2009). After one-phase UHT, 3 forms of malonylglucosides decreased slightly or did not change. Figure 1 shows that the change of malonylglucosides under the same single thermal treatment was different in response to differentgrinding methods. Concomitantly, corresponding increases of β-glucosides were observed with stove cooking increasing the most, and 1-phase UHT increasing slightly or even no change. As for acetylglycitin, all heating methods significantly increased its content as compared with raw soymilk, with stove cooking generating the highest, followed by 2-phase UHT and 1-phase UHT sequentially (Figure 1). This is consistent with other studies (Wang and Murphy 1996; Xu and Chang 2009) and the result also proved that malonylglucosides could also be converted to acetylglucosides under wet heating conditions (Kudou and others 1991). The observed low level of acetylglucosides was mainly due to rapid concurrent degradation (Chien and others 2005). As shown in Figure 1, heating methods also had a great impact on aglycones. Heating methods did not follow definite trends with regard to grinding methods and varieties. As the end products in the interconversion chain, aglycones can be formed and degraded
simultaneously. β-Glucosidase has an optimal temperature of 45 °C and remains active in the pH range of 4.3–7.0 (Matsuura and Obata 1993). In stove cooking, because it took about 8 min to reach boiling, some aglycones could be formed by β-glucosidase–induced hydrolysis from β-glucosides. In soymilk from hot grinding, this hydrolysis was unlikely to occur, because β-glucosidase can be inactivated at 60 °C (Matsuura and Obata 1993). However, heating can also convert β-glucosides to aglycones, but this can only happen over boiling temperatures and daidzin, glycitin, and genistin showed different thermal stabilities (Xu and others 2002). Apart from the conversion from βglucosides, aglycones can undergo auto-decomposition, oxidation, or even Maillard reactions (Davies and others 1998; Ungar and others 2003). In general, heating caused no drastic change on the content of aglycones, which is in agreement with the report of Prabhakaran and Perera (2006). Heating-induced interconversion and degradation can alter the bioactivity of Isoflavones (Singletary and others 2000). As aglycone forms may be absorbed faster and in higher amounts by the human body than their corresponding glucosides, heating conditions should be optimized with consideration of grinding methods to achieve better health results. In the case of total isoflavones, significant differences (P < 0.05) can be found among different heating methods, but they imparted
A 120
Figure 2–Isoflavone percentage distribution in (A) Prosoy soymilk and (B) black soymilk with respect to summed values in soymilk and okara (Din, daidzin; Gly, glycitin; Gin, genistin; MDin, malonyldaidzin; MGly, malonylglycitin; MGin, malonylgenistin; AGly, acetylglycitin; Dein, daidzein; Glein, glycitein; Gein, genistein; Total, total isoflavones.
100 80 60
(%) 40 20 0 Din
Gly
Gin
Cold grinding
MDin MGly MGin AGly Ambient grinding
Dein Glein Gein Total Hot grinding
B 120 100 80
(%)
60 40 20 0 Din
Gly
Gin
Cold grinding
MDin MGly MGin AGly Ambient grinding
C986 Journal of Food Science r Vol. 80, Nr. 5, 2015
Dein Glein Gein Total
Hot grinding
Grinding and heat-processing methods . . .
Prosoy Black soybean a
Dry soybeans
Cold soaked soybeans
Cold soaking water
Ambient soaked soybeans
Ambient soaking water
10425 (7.92)b 9506 (111.1)
10005B (23.89) 8780A (23.85)
47.54B (2.84) 114.4B (3.72)
10702A (50.34) 8608B (43.62)
85.93A (6.20) 204.0A (9.35)
Means with different capital letters in the same row are significantly different (P < 0.05) for different soaking methods. Values in parenthesis are standard deviations.
b
different effects in relation to different grinding methods (Figure 1). For example, in Prosoy soymilk from hot grinding, stove cooking significantly (P < 0.05) increased total isoflavones, whereas the 2 UHT methods significantly (P < 0.05) reduced total isoflavones. It should be noted that the increase of total isoflavones after stove cooking did not mean some new isoflavones were formed during thermal process. It was because of the release of bonded isoflavones from isoflavone–protein complexes. Heating causes denaturation and unfolding of protein, thus disrupting the association between them (Nufer and others 2009). Enzymeaided extraction has proved such isoflavone–protein interaction and showed that, for raw soymilk, the measured isoflavone is somewhat lower than its real value, whereas for heated soymilk, it is very close to the real value (Nufer and others 2009). Hence, in our study, we can only make a comparison of different treatments, but could not measure the loss during the thermal process accurately. This phenomenon was also observed by other researchers and it seemed that the retention of isoflavones was largely dependent on
A
the specific heating methods applied (Xu and Chang 2009). From our results and those of others (Prabhakaran and Perera 2006), the commonly used 2-phase UHT could decrease total isoflavones to some extent, suggesting its degrading effect. The different effects of stove cooking and UHT methods on total isoflavones may be attributed to the protective effect imparted by associated proteins. The thermal stability of isoflavones in response to conversion and degradation was affected by protein content and denaturation state (Malaypally and Ismail 2010). For Prosoy soymilk from ambient grinding, 2-phase UHT did not significantly (P < 0.05) change total isoflavone content, whereas for Prosoy soymilk from hot grinding, 2-phase UHT reduced total isoflavone from 5948 to 5317 nmol/g (Figure 1), which represented 10.6% decrease. The larger loss from hot grinding might be due to its lower protein content and the denaturation state of protein upon heating. The protein contents from ambient and hot grinding were 2.81 and 2.46 g/100 g soymilk, respectively, as reported in our previous work (Zhang and others 2012).
60 40 20 0
(%)
Din
Gly
Gin
MDin
MGly
MGin
AGly
Dein
Glein
Gein
Total
Figure 3–Percentage loss of isoflavones during grinding for (A) Prosoy and (B) black soybean (Din, daidzin; Gly, glycitin; Gin, genistin; MDin, malonyldaidzin; MGly, malonylglycitin; MGin, malonylgenistin; AGly, acetylglycitin; Dein, daidzein; Glein, glycitein; Gein, genistein; Total, total isoflavones).
-20 -40 -60 -80 Cold grinding
B
Ambient grinding
Hot grinding
250 200 150 100
(%) 50 0 Din
Gly
Gin
MDin
MGly
MGin
AGly
Dein
Glein
Gein
Total
-50 -100 Cold grinding
Ambient grinding
Hot grinding Vol. 80, Nr. 5, 2015 r Journal of Food Science C987
C: Food Chemistry
Table 1–Change of total isoflavones during soaking (nmol/g dry bean)a .
Grinding and heat-processing methods . . .
C: Food Chemistry
Loss of isoflavones during soaking As Table 1 shows, either for Prosoy or black soybean, the loss in soaking water is significantly (P < 0.05) higher in ambient soaking than cold soaking. However, the loss is negligible compared with total isoflavone content in soybeans. Jackson and others (2002) reported a 4% loss of total isoflavones during soaking. However, their result was based on the difference between dry soybeans and soaked soybeans and expressed in terms of weight instead of mole. Consequently, a higher soaking loss ensued with the interconversion from high-molecular weight compounds to lowmolecular weight compounds during soaking. In agreement with the result of Kao and others (2004), our study also showed that the soaking process increased aglycone forms and decreased other forms (data not shown), and ambient soaking exhibited greater change of individual isoflavone compounds compared with cold soaking. What is interesting is that acetylglycitin could not be found in both soaked soybeans and soaking water even though it was present in dry soybeans. This might be due to the conversion to β-glucosides or aglycones (Kao and others 2004). Loss during grinding If total isoflavones of raw soymilk and okara are summed, the value would be obviously much lower than the total isoflavones of soaked beans (Figure 3). This clearly means that there must be some destruction of isoflavones during grinding. As Figure 3 shows, for Prosoy, the losses after ambient grinding, cold grinding, and hot grinding were 31.8%, 29.4%, and 30.4%, respectively; for black soybeans, the losses are 27.1%, 16.4%, and 33.4%, respectively. However, not all forms of isoflavones followed the decreasing trend during grinding, in which β-glucosides, acetylglycitin, malonylgenistin, and malonyldaidzin decreased, but malonylglycitin increased. Aglycones, except for those after hot grinding, also exhibited increasing patterns. The different change patterns of individual isoflavones demonstrated that, during grinding, not only destruction occurred, but conversion also took place simultaneously. The above results (Figure 3) demonstrate that, although cold grinding yielded the lowest total isoflavone content in soymilk, it destroyed total isoflavones the least. As for the loss of isoflavones during grinding, to the best of our knowledge, only one paper has mentioned it and attributed it to grinding with boiling water added during grinding (Jackson and others 2002). However, from the results of our study, grinding at lower temperature could also cause loss, and the loss was temperature dependent. The extent of losses also varied for different soybean varieties. The mechanism underling the loss from grinding needs further study.
Conclusions
publication as a journal article, number J-12434 of the Mississippi Agricultural and Forestry Experiment Station, Mississippi State University.
References Boy J. 1999. Protein-polyphenol interactions in fruits juices. J Agric Food Chem 3:85–107. Brouns F. 2002. Soya isoflavones: a new and promising ingredient for the health foods sector. Food Res Int 35(2–3):187–93. Chien JT, Hsieh HC, Kao TH, Chen BH. 2005. Kinetic model for studying the conversion and degradation of isoflavones during heating. Food Chem 91(3):425–34. Cohen LA, Zhao Z, Pittman B, Scimeca JA. 2000. Effect of intact and isoflavone-depleted soy protein on NMU-induced rat mammary tumorigenesis. Carcinogenesis 21(5):929–35. Davies CGA, Netto FM, Glassenap N, Gallaher CM, Labuza TP, Gallaher DD. 1998. Indication of the Maillard Reaction during storage of protein isolates. J Agric Food Chem 46(7):2485–89. Huang H, Liang H, Kwok K-C. 2006. Effect of thermal processing on genistein, daidzein and glycitein content in soymilk. J Sci Food Agr 86(7):1110–4. Jackson CJC, Dini JP, Lavandier C, Rupasinghe HPV, Faulkner H, Poysa V, Buzzell D, DeGrandis S. 2002. Effects of processing on the content and composition of isoflavones during manufacturing of soy beverage and tofu. Process Biochem 37(10):1117–23. Jenkins DJ, Kendall CW, D’Costa MA, Jackson CJ, Vidgen E, Singer W, Silverman JA, Koumbridis G, Honey J, Rao AV, Fleshner N, Klotz L. 2003. Soy consumption and phytoestrogens: effect on serum prostate specific antigen when blood lipids and oxidized low-density lipoprotein are reduced in hyperlipidemic men. J Urology 169(2):507–11. Kao TH, Lu YF, Hsieh HC, Chen BH. 2004. Stability of isoflavone glucosides during processing of soymilk and tofu. Food Res Int 37(9):891–900. Kudou S, Fleury Y, Welti D, Magnolato D, Uchida T, Kitamura K, Okubo K. 1991. Malonyl isoflavone glycosides in soybean seeds (glycine max MERRILL). Agr Biol Chem Tokyo 55(9):2227–33. MacLeod G, Ames J Betz NL. 1988. Soy flavor and its improvement. Crit Rev Food Sci 27(4):219–400. Mahungu SM, Diaz-Mercado S, Li J, Schwenk M, Singletary K, Faller J. 1999. Stability of isoflavones during extrusion processing of corn/soy mixture. J Agric Food Chem 47(1):279– 84. Malaypally SP, Ismail B. 2010. Effect of protein content and denaturation on the extractability and stability of isoflavones in different soy systems. J Agric Food Chem 58(16):8958–65. Matsuura M, Obata A. 1993. β-Glucosidases from soybeans hydrolyze daidzin and genistin. J Food Sci 58(1):144–47. Mizutani T, Hashimoto H. 2004. Effect of grinding temperature on hydroperoxide and off-flavor contents during soymilk manufacturing process. J Food Sci 69(3):SNQ112–6. Nufer KR, Ismail B, Hayes KD. 2009. The effects of processing and extraction conditions on content, profile, and stability of isoflavones in a soymilk system. J Agric Food Chem 57(4):1213–18. Prabhakaran MP, Perera CO. 2006. Effect of extraction methods and UHT treatment conditions on the level of isoflavones during soymilk manufacture. Food Chem 99(2):231–7. Singletary K, Faller J, Li JY Mahungu S. 2000. Effect of extrusion on isoflavone content and antiproliferative bioactivity of soy/corn mixtures. J Agric Food Chem 48(8):3566–71. Sun C, Cadwallader KR, Kim H. 2010. Comparison of key aroma components between soymilks prepared by cold and hot grinding methods. ACS Symposium Series. Oxford University Press, Cary, NC. p. 361–73. Ungar Y, Osundahunsi OF, Shimoni E. 2003. Thermal stability of genistein and daidzein and its effect on their antioxidant activity. J Agric Food Chem 51(15):4394–99. Wang H-J, Murphy PA. 1996. Mass balance study of isoflavones during soybean processing. J Agric Food Chem 44(8):2377–83. Xu B, Chang SKC. 2009. Isoflavones, flavan-3-ols, phenolic acids, total phenolic profiles, and antioxidant capacities of soy milk as affected by ultrahigh-temperature and traditional processing methods. J Agric Food Chem 57(11):4706–17. Xu Z, Wu Q, Godber JS. 2002. Stabilities of daidzin, glycitin, genistin, and generation of derivatives during heating. J Agric Food Chem 50(25):7402–6. Yuan SH, Chang SKC. 2007. Selected odor compounds in cooked soymilk as affected by soybean materials and direct steaming injection. J Food Sci 72(7):s481–6. Yuan SH, Chang SKC, Liu ZS, Xu BJ. 2008. Elimination of trypsin inhibitor activity and beany flavor in soy milk by consecutive blanching and Ultrahigh-Temperature (UHT) Processing. J Agric Food Chem 56:7957–63. Zhang Y, Guo S, Liu Z, Chang SKC. 2012. Off-flavor related volatiles in soymilk as affected by soybean variety, grinding, and heat processing methods. J Agric Food Chem 60(30):7457– 62.
The content of each isoflavone form was greatly influenced by grinding, heating, and variety. Interconversion, degradation, leaching, and heat-induced release were all involved in the whole process. This study provides a foundation for the soymilk industry Supporting Information to optimize its processing conditions. Additional Supporting Information may be found in the online version of this article at the publisher’s website:
Acknowledgments
ND Soybean Council, NDSU-AES, MSU-MAFES, USDA- Table S1. Effect of grinding methods, heating methods, and vaARS SCA 58-6402-2729 contributed funding. Approved for riety on isoflavone profile in soymilk.
C988 Journal of Food Science r Vol. 80, Nr. 5, 2015
Abstract: Isoflavones impart health benefits and their overall content and profile in foods are greatly influenced at each step during processing. In this study, 2 soybean varieties (Prosoy and black soybean) were processed with 3 different grinding (ambient, cold, and hot grinding) and heating methods (traditional stove cooking, 1-phase UHT, and 2-phase UHT) for soymilk making. The results showed after cold, ambient, and hot grinding, the total isoflavones were 3917, 5013, and 5949 nmol/g for Prosoy; the total isoflavones were 4073, 3966, and 4284 nmol/g for black soybean. Grinding could significantly increase isoflavone extraction. The grinding process had a destructive effect on isoflavones and this effect varied with grinding temperature. Different heating methods had different effects on different isoflavone forms. Two soybean varieties showed distinct patterns with respect to the change of isoflavone profile during processing. Keywords: degradation, grinding, isoflavone, soymilk, UHT
Introduction Isoflavones, due to an array of health-promoting functions, are attracting more attention (Cohen and others 2000; Brouns 2002; Jenkins and others 2003). However, the isoflavone profile is greatly affected by the many processing steps that soy products undergo (Mahungu and others 1999; Kao and others 2004; Xu and Chang 2009). Soymilk, as a major soy product, is not well accepted because of its characteristic beany flavor (MacLeod and others 1988). In our previous study, it was proved that proper combinations of grinding and heating methods could effectively reduce some soy off-flavor components to desirable levels (Zhang and others 2012). However, it is still unknown how isoflavone profile is affected by these treatments. Even though the heating effect on soymilk isoflavones has been investigated extensively, there is scant literature dealing with ultra-high temperature (UHT) exposure. Most similar research has been done with soymilk under or equal to boiling temperature (Jackson and others 2002; Nufer and others 2009). Even at UHT temperature, most studies were conducted in model systems and/or at lab scale (Xu and others 2002; Chien and others 2005; Huang and others 2006). Hot and cold grindings have been studied widely to limit the activity of lipoxygenases, which are mainly responsible for the formation of typical offflavors in soymilk (Mizutani and Hashimoto 2004; Sun and others 2010). However, the effects of grinding methods on the extraction efficiency of isoflavones, and the heating influence when in conjunction with different grinding methods, have not yet been reported. Prabhakaran and Perera (2006) made a comparison between hot and ambient grinding without a soaking step. However, only β-glucosides and aglycones were investigated in their study. Protein can form complexes with isoflavones through various interactions and the association is dependent on protein content and the protein denaturation state (Boye 1999; Nufer and others 2009). As a result, the isoflavone extraction efficiency might be
affected by the grinding methods, which have been shown to greatly influence protein recovery and lead to some denaturation, when hot grinding was applied (Zhang and others 2012). In our current study, we used the Microthermics Direct/Indirect Steam Injection Processor (DIP; Microthermics, Inc., Raleigh, N.C., U.S.A.) to mimic the actual scenario of soymilk processing as reported by Yuan and Chang (2008). Two UHT treatments and 1 traditional stove cooking were chosen on the basis of sterilization value for trypsin inhibitor destruction. The 2-phase UHT method is commonly used in the soymilk industry. The objectives of this work were to determine how grinding methods would affect isoflavone extraction and how isoflavone profiles would change under traditional and UHT heating processes in conjunction with different grinding methods.
Materials and Methods Materials Two varieties of soybeans (Glycine max) were used in this study: Prosoy (harvested in 2009) and black soybean (harvested in 2006) were obtained from Sinner Brothers and Bresnahan Co. grown in Casselton, N.Dak., U.S.A. The soybeans were stored in a cool and dry air-conditioned room prior to use. All processing methods were replicated 3 times. Chemicals Daidzein, malonyldaidzin, malonylgenistin, malonylglycitin, and acetylglycitin were purchased from Nacalai USA (San Diego, Calif., U.S.A.). Glycitin, daidzin, genistein, and genistin were purchased from LC Laboratories (Woburn, Mass., U.S.A.). Glycitein was purchased from Sigma-Aldrich (St. Louis, Mo., U.S.A.). High Performance Liquid Chromatography (HPLC) grade methanol and acetonitrile were purchased from VWR International (West Chester, Pa., U.S.A.).
MS 20142017 Submitted 12/9/2014, Accepted 2/9/2015. Authors Zhang, Soymilk preparation Chang, and Liu are with Dept. of Food Science, Nutrition and Health Promotion, Soymilk was prepared as reported in our previous study (Zhang Mississippi State Univ, Starkville, MS 39759, U.S.A. Direct inquiries to author Teixeira (E-mail: [email protected]). and others 2012). In brief, 300 g soybeans were soaked in 1500 mL
cold water (4 °C for cold grinding) or room-temperature water
R C 2015 Institute of Food Technologists
doi: 10.1111/1750-3841.12839 Further reproduction without permission is prohibited
Vol. 80, Nr. 5, 2015 r Journal of Food Science C983
C: Food Chemistry
Isoflavone Profile in Soymilk as Affected by Soybean Variety, Grinding, and Heat-Processing Methods
Grinding and heat-processing methods . . .
C: Food Chemistry
(20 °C for ambient and hot grinding) for 16 h. The hydrated beans were drained and ground with cold water (4 °C), water at 20 °C and 80.5 °C with a bean-to-water ratio of 1:10 (w/w). A portion each of soaked beans and soaking water was freezedried and stored at −20 °C for further analysis. The remaining soaked beans were ground at 10,000 rpm with a New Hartford 1-gallon blender (model CB-2-10; Conn., U.S.A.). To maintain a constant temperature while grinding, the blender was covered with insulating foam (1-in. thickness). For cold grinding, ice (−2.5 °C) and precooled (4 °C) water in the proper ratio (1:3.3) were used to result in an initial grinding temperature of about of 2 °C. For hot grinding, prewarming was done by rinsing the blender twice with boiling water and, immediately, soaked soybeans and boiling water were added. The initial grinding temperatures were recorded as the temperatures at 10 s after grinding; and, in fact, the final temperatures after 3-min grinding were about 2 °C higher than the target temperatures. All these temperatures were determined in preliminary tests. After grinding, the soymilk was manually filtered through a piece of muslin cloth with consistent pressing.
Traditional stove heating of soymilk The method reported by Yuan and Chang (2007) was employed. In brief, l L soymilk was put in a small pot, which was placed in a larger pot with boiling water on a stove. After the soymilk reached 90 °C, the small pot was switched to the stove surface and heated to boiling, from which point the soymilk was maintained boiling with continual stirring for 20 min. Then the small pot was cooled in an ice bath to room temperature. Soymilk was freeze-dried and stored at −20 °C for later analysis. UHT processing of soymilk In this study, Microthermics Direct/Indirect Steam Injection Processor (DIP, Microthermics, Inc., Raleigh, N.C., U.S.A.) was used as reported in our previous study (Zhang and others 2012). In brief, 2 sets of heating temperature and time combinations were applied: 140 °C/5 s (one-phase UHT, F0 = 6.62); 120 °C/80 s + 140 °C/4 s (2-phase UHT, F0 = 6.35). These 2 UHT methods had the similar sterilization power, F0 , calculated on the basis of Z = 10 for bacterial spore sterilization. In the heating tube, the heating medium (steam) was in direct contact with soymilk. The Microthermics Processor was equipped with a vacuum chamber (50 kPa), which was originally designed for the purpose to cool the heated soymilk and to remove the condensed water from the injected steam. However, during the vacuum evaporation process, some volatile compounds were also removed. Extraction of isoflavones Extraction of isoflavones was conducted according to the method of Xu and Chang (2009) with a slight modification. In brief, freeze-dried soymilk and soaking water samples were pulverized with pestle and mortar to fine powder. Freeze-dried soybeans and okara were ground with a coffee blender and passed through a 60-mesh screen. About 1 g of sample powder was accurately weighed into a 15-mL centrifuge tube. Five milliliters of acetonitrile, 4.75 mL DDW, and 0.25 mL 0.5 mg/g 6-hydroxyflavone (in 80% methanol) were added and vortexed. The mixture was shaken in a shaker for 2 h at room temperature and centrifuged at 5500 rpm for 20 min in a centrifuge (Beckman Coulter Ltd., Palo Alto, Calif., U.S.A.). The supernatant was transferred to a 125-mL flat-bottom flask and evaporated at 35 °C until dryness in a rotary evaporator. The residue was dissolved in 5 mL of 80% methanol and filtered through a 0.2-μm syringe filter into vials and were kept at −80 °C until HPLC analyzed. C984 Journal of Food Science r Vol. 80, Nr. 5, 2015
HPLC analysis of isoflavones HPLC analysis was run according to the method of Xu and Chang (2009) with some modification. An Agilent Technologies 1200 series system equipped with a YMC-pack ODS-AM-303 C18 reversed-phase column (250 mm × 4.6 mm i.d., 5 μm) was used. The UV detector was set at 262 nm and column temperature was set at 35 °C. Mobile phase A was 0.1% glacial acetic acid in water, mobile phase B was 0.1% glacial acetic acid in acetonitrile. Initially, B was set 15% for 5 min. Then B was increased to 29% up to 36 min, increased to 35% to 44 min, then increased to 50% up to 46 min, and held until 56 min. Then B was recycled to 15% until 58 min and held for 60 min. Peaks of isoflavones were identified by comparing the retention times of the authentic standards and were subsequently quantified by preparing a calibration curve of standards. Individual isoflavones were expressed as μg/g of dry material; and for the purposes of comparison, total isoflavones were expressed as nmol/g of dry material on the basis of their individual molecular weights. Statistical analysis Soymilk processing was conducted in triplicate and the following HPLC analyses were completed in duplicate. The data were subjected to analysis of variance with SAS 9.1 package (SAS 2005). Significant differences among variables were determined by Duncan’s multiple range test (α = 0.05). Data were expressed as means ± SD (n = 6).
Results and Discussion Effect of grinding methods on isoflavone profile and content in soymilk As presented in Figure 1, significant differences (P < 0.05) were found in individual isoflavone types and total insoflavones among the 3 grinding methods. However, the 2 varieties showed different response to the grinding methods. For Prosoy soymilk, in most cases, except aglycones, hot grinding generated significantly (P < 0.05) higher individual and total isoflavones than the other 2 grinding methods. Meanwhile, ambient grinding yielded significantly (P < 0.05) higher level of isoflavones than cold grinding. For example, in raw Prosoy soymilk, the total isoflavone content from cold, ambient, and hot grinding methods were 3917, 5013, and 5949 nmol/g, respectively. For soymilk from black soybeans, in most cases, ambient grinding produced the lowest isoflavone content and hot grinding and cold grinding exhibited similar efficiency. For example, in raw soymilk from black soybeans, the total isoflavone content from cold, ambient, and hot grinding methods were 4073, 3966, and 4284 nmol/g, respectively. The distinct behaviors of 2 varieties in response to grinding methods suggest different compositions and structures. Aglycones exhibited a distinct pattern in responding to grinding methods; that is, for both varieties, ambient grinding resulted in the highest level of aglycones and hot grinding generated significantly (P < 0.05) higher aglycones than cold grinding. The lower level of aglycones from hot grinding can be explained by the instant inactivation of β-glucosidase during hot grinding and this phenomenon was also observed by Prabhakaran and Perera (2006). Until now, we only found one paper comparing hot grinding (95 °C) and cold grinding (45 °C) with regard to isoflavone availability in soymilk (Prabhakaran and Perera 2006). In their study, the soaking step was omitted, but hot grinding still showed much higher isoflavone-producing efficiency compared with cold grinding in terms of aglycones and β-glucosides. Wang and Murphy (1996)
found that heating slurry after grinding at 95 °C for 7 min prior to pressing could result in about 12% loss in okara and almost 90% of the isoflavones present in soymilk. The 90% isoflavone recovery was very high compared with our results and those of others. Using soy protein isolate, Malaypally and Ismail (2010) also found that heating could significantly facilitate the recovery of isoflavones in soymilk. Therefore, on the basis of combined results, we suppose that heating before separation between insoluble okara and soymilk might advance the recovery of soflavones in soymilk. And this is perhaps the main reason for the relatively higher isoflavone content from hot grinding as observed in our study. In fact, different recoveries resulted mainly from different solubilities of different isoflavone forms responding to different grinding temperatures. Jackson and others (2002) found in ambient grinding that the percentage contents of aglycones, βglucosides, and acetylglucosides in okara were higher than those in soymilk, but malonylglucosides showed an opposite trend. As shown in Figure 2, our results showed that the distributions of each single isoflavone form in okara and soymilk are greatly different for different grinding methods. For example, in raw Prosoy soymilk, the most abundant isoflavone form, the combined content of malonylglucosides in soymilk and okara from cold
A
1400
grinding was 5100 nmol/g, much higher than that from hot grinding of 4777 nmol/g (data not shown). However, only 56% of malonylglucosides remained in soymilk from cold grinding in comparison with 82% from hot grinding. Consequently, hot grinding yielded much higher total isoflavone compared with cold grinding. Apart from the temperature-induced effect on isoflavone recovery in soymilk, Wang and Murphy (1996) attributed isoflavone in soymilk to the association between isoflavones and soluble proteins in soymilk. However, our previous study demonstrated that the protein recovery from hot grinding was lower than that from ambient grinding and a little higher than that from cold grinding (Zhang and others 2012). The difference of total isoflavone was likely due to different solubility of different individual isoflavones in soymilk under different grinding temperatures.
Effect of heating methods on isoflavone profile and content Except for one-phase UHT, all other heating processes significantly (P < 0.05) reduced the contents of malonylglucosides with stove cooking reducing the most (Figure 1). Our result was in agreement with the report by Xu and Chang (2009) who also found that stove cooking reduced malonylglucosides
D 50
a
a
45
1200
40 1000
b
800
µg/g
a
600 400
a
c
b
bb
30
c
b
a
c
a
b
35
a
b
a
a
a
a
b
µg/g
a
ab
25 20 10 5
a
a
15
c
a
c
c
b
a
b
c
a a
b
a a
b
c b
b
c
b
b
a
0 200
Raw
SC
Raw
SC
OP
TP
Raw
Prosoy Cold grinding Ambient grinding
B
OP
TP
Raw
SC
Prosoy
0
a
2500
b
2000
b
c 1500
a
µg/g
Cold grinding
TP
Ambient grinding
Hot grinding
Hot grinding
a a b
c
OP
TP
Black soybean
a a b
SC
OP
Black soybean
E b b b
1000
7000
a
a a a
a a a
5000
nmol/g
4000
a
a
6000
a
a
c
a
b
b
b
a
c
b
c
a
ab b
a
bb
a
a a a
b
ab a
3000 500
2000 1000
0 Raw
SC
OP
TP
Prosoy Cold grinding
C
Raw
OP
TP
0 Raw
Black soybean Hot grinding
Ambient grinding
SC
OP
TP
Prosoy Cold grinding Ambient grinding
Raw
SC
OP
TP
Black soybean Hot grinding
250
a
150
a
100
b 50
a
a
200
µg/g
SC
b
a
ab
b
c
b
b
c
c
a b
b
a
c
c
a b
b c
0 Raw
SC
OP
TP
Prosoy Cold grinding
Raw
SC
OP
TP
Black soybean Ambient grinding
Hot grinding
Figure 1–Effect of grinding methods, heating methods, and variety on (A) β-glucosides, (B) malonylglucosides, (C) aglycones, (D) acetylglycitin, and (E) total isoflavones (SC, stove cooking; OP, one-phase UHT; TP, two-phase UHT). Bar data are expressed as mean ± standard deviation (n = 3) on dry weight basis. Values marked with different letter above the bars are significantly different within the same heating treatment and variety group (p < 0.05). Effect of grinding methods, heating methods, and variety on (A) β-glucosides, (B) malonylglucosides, (C) aglycones, (D) acetylglycitin, and (E) total isoflavones (SC, stove cooking; OP, one-phase UHT; TP, 2-phase UHT). Vol. 80, Nr. 5, 2015 r Journal of Food Science C985
C: Food Chemistry
Grinding and heat-processing methods . . .
Grinding and heat-processing methods . . .
C: Food Chemistry
even more than direct and indirect UHT at 143 °C for 60 s. Malonylglucosides are very heat labile and are prone to convert to β-glucosides and acetylglucosides under moist heat (Chien and others 2005; Xu and Chang 2009). After one-phase UHT, 3 forms of malonylglucosides decreased slightly or did not change. Figure 1 shows that the change of malonylglucosides under the same single thermal treatment was different in response to differentgrinding methods. Concomitantly, corresponding increases of β-glucosides were observed with stove cooking increasing the most, and 1-phase UHT increasing slightly or even no change. As for acetylglycitin, all heating methods significantly increased its content as compared with raw soymilk, with stove cooking generating the highest, followed by 2-phase UHT and 1-phase UHT sequentially (Figure 1). This is consistent with other studies (Wang and Murphy 1996; Xu and Chang 2009) and the result also proved that malonylglucosides could also be converted to acetylglucosides under wet heating conditions (Kudou and others 1991). The observed low level of acetylglucosides was mainly due to rapid concurrent degradation (Chien and others 2005). As shown in Figure 1, heating methods also had a great impact on aglycones. Heating methods did not follow definite trends with regard to grinding methods and varieties. As the end products in the interconversion chain, aglycones can be formed and degraded
simultaneously. β-Glucosidase has an optimal temperature of 45 °C and remains active in the pH range of 4.3–7.0 (Matsuura and Obata 1993). In stove cooking, because it took about 8 min to reach boiling, some aglycones could be formed by β-glucosidase–induced hydrolysis from β-glucosides. In soymilk from hot grinding, this hydrolysis was unlikely to occur, because β-glucosidase can be inactivated at 60 °C (Matsuura and Obata 1993). However, heating can also convert β-glucosides to aglycones, but this can only happen over boiling temperatures and daidzin, glycitin, and genistin showed different thermal stabilities (Xu and others 2002). Apart from the conversion from βglucosides, aglycones can undergo auto-decomposition, oxidation, or even Maillard reactions (Davies and others 1998; Ungar and others 2003). In general, heating caused no drastic change on the content of aglycones, which is in agreement with the report of Prabhakaran and Perera (2006). Heating-induced interconversion and degradation can alter the bioactivity of Isoflavones (Singletary and others 2000). As aglycone forms may be absorbed faster and in higher amounts by the human body than their corresponding glucosides, heating conditions should be optimized with consideration of grinding methods to achieve better health results. In the case of total isoflavones, significant differences (P < 0.05) can be found among different heating methods, but they imparted
A 120
Figure 2–Isoflavone percentage distribution in (A) Prosoy soymilk and (B) black soymilk with respect to summed values in soymilk and okara (Din, daidzin; Gly, glycitin; Gin, genistin; MDin, malonyldaidzin; MGly, malonylglycitin; MGin, malonylgenistin; AGly, acetylglycitin; Dein, daidzein; Glein, glycitein; Gein, genistein; Total, total isoflavones.
100 80 60
(%) 40 20 0 Din
Gly
Gin
Cold grinding
MDin MGly MGin AGly Ambient grinding
Dein Glein Gein Total Hot grinding
B 120 100 80
(%)
60 40 20 0 Din
Gly
Gin
Cold grinding
MDin MGly MGin AGly Ambient grinding
C986 Journal of Food Science r Vol. 80, Nr. 5, 2015
Dein Glein Gein Total
Hot grinding
Grinding and heat-processing methods . . .
Prosoy Black soybean a
Dry soybeans
Cold soaked soybeans
Cold soaking water
Ambient soaked soybeans
Ambient soaking water
10425 (7.92)b 9506 (111.1)
10005B (23.89) 8780A (23.85)
47.54B (2.84) 114.4B (3.72)
10702A (50.34) 8608B (43.62)
85.93A (6.20) 204.0A (9.35)
Means with different capital letters in the same row are significantly different (P < 0.05) for different soaking methods. Values in parenthesis are standard deviations.
b
different effects in relation to different grinding methods (Figure 1). For example, in Prosoy soymilk from hot grinding, stove cooking significantly (P < 0.05) increased total isoflavones, whereas the 2 UHT methods significantly (P < 0.05) reduced total isoflavones. It should be noted that the increase of total isoflavones after stove cooking did not mean some new isoflavones were formed during thermal process. It was because of the release of bonded isoflavones from isoflavone–protein complexes. Heating causes denaturation and unfolding of protein, thus disrupting the association between them (Nufer and others 2009). Enzymeaided extraction has proved such isoflavone–protein interaction and showed that, for raw soymilk, the measured isoflavone is somewhat lower than its real value, whereas for heated soymilk, it is very close to the real value (Nufer and others 2009). Hence, in our study, we can only make a comparison of different treatments, but could not measure the loss during the thermal process accurately. This phenomenon was also observed by other researchers and it seemed that the retention of isoflavones was largely dependent on
A
the specific heating methods applied (Xu and Chang 2009). From our results and those of others (Prabhakaran and Perera 2006), the commonly used 2-phase UHT could decrease total isoflavones to some extent, suggesting its degrading effect. The different effects of stove cooking and UHT methods on total isoflavones may be attributed to the protective effect imparted by associated proteins. The thermal stability of isoflavones in response to conversion and degradation was affected by protein content and denaturation state (Malaypally and Ismail 2010). For Prosoy soymilk from ambient grinding, 2-phase UHT did not significantly (P < 0.05) change total isoflavone content, whereas for Prosoy soymilk from hot grinding, 2-phase UHT reduced total isoflavone from 5948 to 5317 nmol/g (Figure 1), which represented 10.6% decrease. The larger loss from hot grinding might be due to its lower protein content and the denaturation state of protein upon heating. The protein contents from ambient and hot grinding were 2.81 and 2.46 g/100 g soymilk, respectively, as reported in our previous work (Zhang and others 2012).
60 40 20 0
(%)
Din
Gly
Gin
MDin
MGly
MGin
AGly
Dein
Glein
Gein
Total
Figure 3–Percentage loss of isoflavones during grinding for (A) Prosoy and (B) black soybean (Din, daidzin; Gly, glycitin; Gin, genistin; MDin, malonyldaidzin; MGly, malonylglycitin; MGin, malonylgenistin; AGly, acetylglycitin; Dein, daidzein; Glein, glycitein; Gein, genistein; Total, total isoflavones).
-20 -40 -60 -80 Cold grinding
B
Ambient grinding
Hot grinding
250 200 150 100
(%) 50 0 Din
Gly
Gin
MDin
MGly
MGin
AGly
Dein
Glein
Gein
Total
-50 -100 Cold grinding
Ambient grinding
Hot grinding Vol. 80, Nr. 5, 2015 r Journal of Food Science C987
C: Food Chemistry
Table 1–Change of total isoflavones during soaking (nmol/g dry bean)a .
Grinding and heat-processing methods . . .
C: Food Chemistry
Loss of isoflavones during soaking As Table 1 shows, either for Prosoy or black soybean, the loss in soaking water is significantly (P < 0.05) higher in ambient soaking than cold soaking. However, the loss is negligible compared with total isoflavone content in soybeans. Jackson and others (2002) reported a 4% loss of total isoflavones during soaking. However, their result was based on the difference between dry soybeans and soaked soybeans and expressed in terms of weight instead of mole. Consequently, a higher soaking loss ensued with the interconversion from high-molecular weight compounds to lowmolecular weight compounds during soaking. In agreement with the result of Kao and others (2004), our study also showed that the soaking process increased aglycone forms and decreased other forms (data not shown), and ambient soaking exhibited greater change of individual isoflavone compounds compared with cold soaking. What is interesting is that acetylglycitin could not be found in both soaked soybeans and soaking water even though it was present in dry soybeans. This might be due to the conversion to β-glucosides or aglycones (Kao and others 2004). Loss during grinding If total isoflavones of raw soymilk and okara are summed, the value would be obviously much lower than the total isoflavones of soaked beans (Figure 3). This clearly means that there must be some destruction of isoflavones during grinding. As Figure 3 shows, for Prosoy, the losses after ambient grinding, cold grinding, and hot grinding were 31.8%, 29.4%, and 30.4%, respectively; for black soybeans, the losses are 27.1%, 16.4%, and 33.4%, respectively. However, not all forms of isoflavones followed the decreasing trend during grinding, in which β-glucosides, acetylglycitin, malonylgenistin, and malonyldaidzin decreased, but malonylglycitin increased. Aglycones, except for those after hot grinding, also exhibited increasing patterns. The different change patterns of individual isoflavones demonstrated that, during grinding, not only destruction occurred, but conversion also took place simultaneously. The above results (Figure 3) demonstrate that, although cold grinding yielded the lowest total isoflavone content in soymilk, it destroyed total isoflavones the least. As for the loss of isoflavones during grinding, to the best of our knowledge, only one paper has mentioned it and attributed it to grinding with boiling water added during grinding (Jackson and others 2002). However, from the results of our study, grinding at lower temperature could also cause loss, and the loss was temperature dependent. The extent of losses also varied for different soybean varieties. The mechanism underling the loss from grinding needs further study.
Conclusions
publication as a journal article, number J-12434 of the Mississippi Agricultural and Forestry Experiment Station, Mississippi State University.
References Boy J. 1999. Protein-polyphenol interactions in fruits juices. J Agric Food Chem 3:85–107. Brouns F. 2002. Soya isoflavones: a new and promising ingredient for the health foods sector. Food Res Int 35(2–3):187–93. Chien JT, Hsieh HC, Kao TH, Chen BH. 2005. Kinetic model for studying the conversion and degradation of isoflavones during heating. Food Chem 91(3):425–34. Cohen LA, Zhao Z, Pittman B, Scimeca JA. 2000. Effect of intact and isoflavone-depleted soy protein on NMU-induced rat mammary tumorigenesis. Carcinogenesis 21(5):929–35. Davies CGA, Netto FM, Glassenap N, Gallaher CM, Labuza TP, Gallaher DD. 1998. Indication of the Maillard Reaction during storage of protein isolates. J Agric Food Chem 46(7):2485–89. Huang H, Liang H, Kwok K-C. 2006. Effect of thermal processing on genistein, daidzein and glycitein content in soymilk. J Sci Food Agr 86(7):1110–4. Jackson CJC, Dini JP, Lavandier C, Rupasinghe HPV, Faulkner H, Poysa V, Buzzell D, DeGrandis S. 2002. Effects of processing on the content and composition of isoflavones during manufacturing of soy beverage and tofu. Process Biochem 37(10):1117–23. Jenkins DJ, Kendall CW, D’Costa MA, Jackson CJ, Vidgen E, Singer W, Silverman JA, Koumbridis G, Honey J, Rao AV, Fleshner N, Klotz L. 2003. Soy consumption and phytoestrogens: effect on serum prostate specific antigen when blood lipids and oxidized low-density lipoprotein are reduced in hyperlipidemic men. J Urology 169(2):507–11. Kao TH, Lu YF, Hsieh HC, Chen BH. 2004. Stability of isoflavone glucosides during processing of soymilk and tofu. Food Res Int 37(9):891–900. Kudou S, Fleury Y, Welti D, Magnolato D, Uchida T, Kitamura K, Okubo K. 1991. Malonyl isoflavone glycosides in soybean seeds (glycine max MERRILL). Agr Biol Chem Tokyo 55(9):2227–33. MacLeod G, Ames J Betz NL. 1988. Soy flavor and its improvement. Crit Rev Food Sci 27(4):219–400. Mahungu SM, Diaz-Mercado S, Li J, Schwenk M, Singletary K, Faller J. 1999. Stability of isoflavones during extrusion processing of corn/soy mixture. J Agric Food Chem 47(1):279– 84. Malaypally SP, Ismail B. 2010. Effect of protein content and denaturation on the extractability and stability of isoflavones in different soy systems. J Agric Food Chem 58(16):8958–65. Matsuura M, Obata A. 1993. β-Glucosidases from soybeans hydrolyze daidzin and genistin. J Food Sci 58(1):144–47. Mizutani T, Hashimoto H. 2004. Effect of grinding temperature on hydroperoxide and off-flavor contents during soymilk manufacturing process. J Food Sci 69(3):SNQ112–6. Nufer KR, Ismail B, Hayes KD. 2009. The effects of processing and extraction conditions on content, profile, and stability of isoflavones in a soymilk system. J Agric Food Chem 57(4):1213–18. Prabhakaran MP, Perera CO. 2006. Effect of extraction methods and UHT treatment conditions on the level of isoflavones during soymilk manufacture. Food Chem 99(2):231–7. Singletary K, Faller J, Li JY Mahungu S. 2000. Effect of extrusion on isoflavone content and antiproliferative bioactivity of soy/corn mixtures. J Agric Food Chem 48(8):3566–71. Sun C, Cadwallader KR, Kim H. 2010. Comparison of key aroma components between soymilks prepared by cold and hot grinding methods. ACS Symposium Series. Oxford University Press, Cary, NC. p. 361–73. Ungar Y, Osundahunsi OF, Shimoni E. 2003. Thermal stability of genistein and daidzein and its effect on their antioxidant activity. J Agric Food Chem 51(15):4394–99. Wang H-J, Murphy PA. 1996. Mass balance study of isoflavones during soybean processing. J Agric Food Chem 44(8):2377–83. Xu B, Chang SKC. 2009. Isoflavones, flavan-3-ols, phenolic acids, total phenolic profiles, and antioxidant capacities of soy milk as affected by ultrahigh-temperature and traditional processing methods. J Agric Food Chem 57(11):4706–17. Xu Z, Wu Q, Godber JS. 2002. Stabilities of daidzin, glycitin, genistin, and generation of derivatives during heating. J Agric Food Chem 50(25):7402–6. Yuan SH, Chang SKC. 2007. Selected odor compounds in cooked soymilk as affected by soybean materials and direct steaming injection. J Food Sci 72(7):s481–6. Yuan SH, Chang SKC, Liu ZS, Xu BJ. 2008. Elimination of trypsin inhibitor activity and beany flavor in soy milk by consecutive blanching and Ultrahigh-Temperature (UHT) Processing. J Agric Food Chem 56:7957–63. Zhang Y, Guo S, Liu Z, Chang SKC. 2012. Off-flavor related volatiles in soymilk as affected by soybean variety, grinding, and heat processing methods. J Agric Food Chem 60(30):7457– 62.
The content of each isoflavone form was greatly influenced by grinding, heating, and variety. Interconversion, degradation, leaching, and heat-induced release were all involved in the whole process. This study provides a foundation for the soymilk industry Supporting Information to optimize its processing conditions. Additional Supporting Information may be found in the online version of this article at the publisher’s website:
Acknowledgments
ND Soybean Council, NDSU-AES, MSU-MAFES, USDA- Table S1. Effect of grinding methods, heating methods, and vaARS SCA 58-6402-2729 contributed funding. Approved for riety on isoflavone profile in soymilk.
C988 Journal of Food Science r Vol. 80, Nr. 5, 2015
