Isoflavone profiles of soymilk as affected by high pressure treatments of soymilk and soybeans

ARTICLE IN PRESS
Food Chemistry xxx (2008) xxx–xxx
Contents lists available at ScienceDirect
Food Chemistry
j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / f o o d c h e m
Iso?avone pro?les of soymilk as affected by high-pressure treatments
of soymilk and soybeans
Stephanie Jung *, Patricia A. Murphy, Ileana Sala
Department of Food Science and Human Nutrition, Iowa State University, 2312 Food Science Building, Ames, IA 50011-1061, USA
a r t i c l e
i n f o
a b s t r a c t
Article history:
High hydrostatic pressure was applied to hydrated soybeans (100–700 MPa, 25 °C) and soymilk (400–
Received 15 November 2007
750 MPa; 25 and 75 °C) to assess its effect on iso?avone content, pro?le and water-extractability. Neither
Received in revised form 31 March 2008
pressure level nor initial treatment temperature affected soymilk iso?avone content. However, combined
Accepted 9 April 2008
pressure and mild thermal treatment modi?ed the iso?avone distribution. At 75 °C, the iso?avone pro?le
Available online xxxx
shifted from malonylglucosides toward b-glucosides, which was correlated to the effect of adiabatic heat-
ing. When pressure was applied to the hydrated soybeans, the soymilk iso?avone concentration varied
Keywords:
between 4.32 and 6.06 lmol/g. The content of protein decreased and fat increased in soymilks prepared
Iso?avone
from pressurized soybeans with increasing pressure level.
Soybean
Soymilk
Published by Elsevier Ltd.
High-pressure processing
1. Introduction
by adjusting temperature and pH (Prabhakaran & Perera, 2006;
Rickert, Johnson, & Murphy, 2004; Rickert, Meyer, Hu, & Murphy,
Iso?avones, which have received considerable attention due to
2004; Speroni, Milesi, & Anon, 2007). These adjustments should
their biological activity over the past 20 years, are phytoestrogens
be done carefully as an increase of extracted iso?avone can nega-
that are present in a concentration of 0.3–0.8% (db) in soybean
tively affect the extractability of protein (Barbosa, Lajolo, &
seeds (USDA-Iowa State University Database on the Iso?avone
Genovese, 2006). Besides their extractability, de?ned in this paper
Content of Foods, 2002). Interest in soy iso?avones is based on
as the amount of iso?avones recovered from the starting material
data suggesting potential of iso?avones in lowering cholesterol
into water, other determining parameters accounting for iso?av-
levels, preventing both prostate and breast cancers and attenuat-
ones’ ?nal concentration and prevalence of the different forms
ing bone loss in postmenopausal women, and alleviating meno-
are their rate of conversion. The conversions of iso?avones during
pausal symptoms (Hendrich & Murphy, 2007; Liu, 2004).
processing are dictated by both their chemical structure, and
Daidzin, genistin and glycitin are the three iso?avone glucosides
other parameters such as pH, temperature, moisture and activity
existing in soybeans and soy-based foods. They can be found as
of endogenous b-glucosidases (Ismail & Hayes, 2005).
non-conjugated b-glucosides and conjugated malonyl- or acetyl-
Soy proteins constitute another critical constituent in soy-
b-glucosides. In raw soybeans, although the iso?avone forms de-
beans. For decades they have been recognized for their excellent
pend on many seed characteristics such as variety, crop year
nutritional quality and functionalities as food ingredients. In
and growth location, the most predominant ones are malonyl-
1999, the soy protein-cardiovascular health claim was approved
daidzin and -genistin which together constitute 71–81% of total
by the US Food and Drug Administration and apparently spurred
iso?avones (Charron, Allen, Johnson, Pantalone, & Sams, 2005).
increased interest in soy-based products. This claim has certainly
The processing operations and conditions applied for production
contributed to the steady increase in soymilk consumption in
of soy-based products and ingredients determine the ?nal content
Western countries, which added to their long-standing popularity
and pro?le of iso?avones (Liu, 2004; Murphy et al., 1999; Wang &
in East Asian countries, making soymilk a popular beverage
Murphy, 1994). Extraction in water is the ?rst important process-
around the world. Soymilk production has several thermal treat-
ing step in the recovery of iso?avones from soy matrices to pro-
ments, which can negatively affect its nutritional quality, color
duce soymilk, tofu and soy protein isolate (SPI). During SPI and
and sensory attributes (Kwok & Niranjan, 1995) and modify iso-
soymilk production iso?avones can be partially retained in the ?-
?avone distribution. High-pressure processing (HPP), a newly
bre fractions. These losses in the ?bre fractions can be minimized
developed food technology, could be an alternative to thermal
processing for soymilk production (Kajiyama, Isobe, Uemura, &
Nohuchi, 1995; Lakshmanan, de Lamballerie, & Jung, 2006; Zhang,
* Corresponding author. Tel.: +1 515 294 2544; fax: +1 515 294 8181.
E-mail address: jung@iastate.edu (S. Jung).
Li, Tatsumi, & Isobe, 2005). The unique effects of HPP are due to
0308-8146/$ - see front matter Published by Elsevier Ltd.
doi:10.1016/j.foodchem.2008.04.025
Please cite this article in press as: Jung, S. et al., Iso?avone pro?les of soymilk as affected by high-pressure treatments ..., Food Chemistry
(2008), doi:10.1016/j.foodchem.2008.04.025

---

ARTICLE IN PRESS
2
S. Jung et al. / Food Chemistry xxx (2008) xxx–xxx
the effect of pressure on non-covalent bonds while leaving the
For pressure treatment of the hydrated soybeans, 28 g of soy-
covalent bonds of the food intact, in contrast to the changes
beans were soaked as described above and after a water uptake
occurring during thermal processing. Because of the pressure-in-
determination, the appropriate amount of water needed to reach
duced modi?cation in protein structure and interactions, and
a dry bean-to-water ratio of 1:8 (w:w) was added to the soybeans
the potential for creation of food products with unique properties,
directly in a polyester bag. After HPP treatment, the content of the
the effects of high-pressure processing have been investigated on
bag, i.e. pressurized soybeans and water was transferred in a 1-L
functionality of soy protein isolates, soy protein gels and tofu
Waring heavy-duty laboratory blender (Torrington, CT) and the
(Lakshmanan et al., 2006). There is, however, a lack of data on
soymilk was prepared as described above.
the effect of high-pressure processing on the stability and inter-
conversion of the health-promoting iso?avones. In addition, little
2.3. Thermal treatment
information is available on the interaction between soy iso?av-
ones and proteins and how this interaction modi?es iso?avones
The soymilk was heated at 95 ± 2 °C for 15 min in a 2-L reaction
extractability in water. During soymilk production, about 30% of
vessel (model CG-1929-16, ChemGlass, Vineland, NJ) with contin-
the total mole mass of iso?avones is lost in the insoluble fraction,
uous stirring at 695 rpm (stirrer model BDC 3030, Caframo, Wiar-
the okara (Wang & Murphy, 1996). Protein–polyphenol interac-
ton, Ontario) prior to HPP. It took $15 min to reach the 95 °C
tions involve hydrogen bonding, ionic, and hydrophobic interac-
temperature. After the 15 min at 95 °C, the soymilk was immedi-
tions (Boye, 1999), which are known to be pressure-sensitive.
ately cooled down in an ice-water bath.
There is body of evidence that denaturation and modi?cation of
physico-chemical characteristics of the two main soy proteins,
2.4. High-pressure processing
glycinin and b-conglycinin, occurs during high-pressure process-
ing (Torrezan, Tham, Bell, Frazier, & Cristianini, 2007; Puppo
Soymilk and hydrated soybeans in water were vacuum-pack-
et al., 2004). Iso?avones seem to have an af?nity for denaturated
aged in polyester bags (SealPaks, KAPAK, Minneapolis, MN) in a
soy proteins (Rickert et al., 2004). Therefore under pressure, the
tabletop Multivac machine (Model C 100, Multivac, Kansas City,
unfolding and denaturation of soy protein and/or changes in the
MO). The samples were pressurized with a Food-Lab 900 high-
polyphenol interaction, could modify iso?avone extractability in
pressure food processor (Stansted Fluid Power, Stansted, UK). The
water. In addition, considerable changes in the structure of coty-
sample holder was 6.5-cm i.d. and 23-cm height. The rates of pres-
ledon surface and epidermal cells of pressurized soybean seeds
surization and depressurization were 260 and 500 MPa/min,
were observed as well as some release of the soy protein in soak-
respectively. Distilled water containing 10% vegetable soybean
ing water surrounding the beans during treatment (Omi, Kato,
oil, 0.18% Tween 80, 0.02% Span 80, and 0.1% potassium sorbate
Ishida, Kato, & Matsuda, 1996). These changes in cotyledon sur-
was used as the pressure transmitting ?uid. In preliminary exper-
face and epidermal cells induced by HPP may also affect iso?av-
iments, the pressures and temperatures of soymilk, pressurization
one extractability. To the best of our knowledge, no studies
?uid, and vessel were recorded over the entire period using a Stan-
have investigated the possible changes in the iso?avone water-
sted ?uid power FPG55000 RAP system and a Scan 1000 supervi-
extractability as a result of high-pressure processing.
sory control and data acquisition system (Hexatec, Hexham, UK).
This study reports on the effect of pressure level and initial
Both the pressurization ?uid and soymilk had the same tempera-
treatment temperature on the content and pro?le of iso?avones
ture increase during treatment due to adiabatic heating. Therefore,
in soymilk. The potential of using high-pressure processing as a
for practical reasons the temperature was directly measured in the
means to modify iso?avone water-extractability was determined,
pressurization ?uid. The quasi-adiabatic temperature increase
in addition to the composition, viscosity and protein characteristics
upon compression (ds) was de?ned as the ratio of the change of
of soymilk obtained from pressurized soybeans.
temperature (DT, °C) under pressure divided by the pressure
(MPa) Â 100 (Patazca, Koutchma, & Balasubramaniam, 2007).
The soymilk previously heated at 95 °C was pressurized from
2. Materials and methods
400 to 750 MPa for 10 min at 25 and 75 °C. Soymilk was pre-
warmed to the initial temperature of treatment for 10 min in a
2.1. Materials
water bath. The control was pre-warmed in the same conditions
but not submitted to any pressure treatment. Immediately after
Vinton 81 cultivar soybeans (Glycine max L.) were purchased lo-
pressure treatment, the samples were cooled to room temperature
cally (Pattison Bros., Fayette, IA) and stored in the dark at 4 °C. All
in an ice-water bath. Aliquots were taken from all soymilk samples
reagents were of analytical grade and purchased from Fisher Scien-
and kept at 4 °C for further analyses, and the remaining portions
ti?c (Pittsburgh, PA) and Sigma–Aldrich (St. Louis, MO). Acetoni-
were immediately frozen. The hydrated soybeans were treated at
trile and methanol from Fisher were of HPLC-grade.
100, 200, 300, 400, 500, 600, and 700 MPa for 10 min at 25 °C.
The control was raw soymilk, i.e. soymilk prepared without ther-
2.2. Soymilk and soaked soybean preparation
mal or pressure treatment. After soymilk preparation, aliquot of
soymilk was immediately frozen. The experiments were conducted
One hundred grams of soybeans were washed to remove dirt
in duplicate.
and soaked for 12 h at room temperature in tap water. The drained
soybeans were weighed to determine the water uptake and then
2.5. Moisture, fat and crude protein determination
ground at low speed for 1 min with water to yield a dry bean-to-
water ratio of 1:8 (w:w) in a 4-L Waring heavy-duty laboratory
The moisture content of the samples was determined according
blender (Torrington, CT). The slurry was ?ltered through a 100-
to the AOCS (1995) method Ba 2a-38 with slight modi?cations.
mesh nylon ?lter-sack and water was added to reach a dry soy-
One gram of freeze-dried soymilk or 3 g of liquid soymilk were
bean-to-water ratio of 1:10 (w:w). Finally, the slurry was squeezed
weighed into tared aluminum dishes and dried in a Precision Econ-
manually to separate the insoluble residue, okara, from the ?ltrate.
omy forced-air oven (Thermo Electron, Waltham, MA) for 3 h at
The soymilk had a pH of 6.6 and a Brix of 6.5° ± 0.2°. Soymilk yield
130 °C. Lipid content of freeze-dried soymilk was determined with
was calculated as the weight (g) of soymilk obtained per 100 g of
AACC method 30-25. The freeze-dried soymilk was extracted with
soybeans.
petroleum ether with a Gold?sh extractor for 4 h. Crude protein
Please cite this article in press as: Jung, S. et al., Iso?avone pro?les of soymilk as affected by high-pressure treatments ..., Food Chemistry
(2008), doi:10.1016/j.foodchem.2008.04.025

---

ARTICLE IN PRESS
S. Jung et al. / Food Chemistry xxx (2008) xxx–xxx
3
content was measured by the Dumas method using a Rapid NIII
2.8. SDS PAGE
nitrogen analyzer (Elementar Americas, Mt. Laurel, NJ) as described
by Jung et al. (2003). A factor of 6.25 was used to convert nitrogen
Freeze-dried soymilk samples were defatted as described in &
to crude protein content.
2.7 and prepared at 6 g/L in 0.5 M Tris–HCl, 30% urea, 20% glycerol,
2.5% bromophenol-blue, 2% SDS solution and 2% 2-mercap-
2.6. Iso?avone content
toethanol, pH 6.8. Electrophoresis was performed as previously de-
scribed (Lakshmanan et al., 2006) with a mini Protean III
Iso?avone content of control soymilks, pressurized soymilks
electrophoresis system using 4–20% linear gradient gel (Bio-Rad
and soymilks prepared from pressurized beans was determined
Laboratories, Hercules, CA).
from freeze-dried soymilk. Soybeans were ground in a coffee grin-
der and extracted and analyzed as described for freeze-dried soy-
2.9. Apparent viscosity
milk. Freeze-dried soymilk was ground with pestle and mortar,
and approximately 2 g were accurately weighed into a 125-mL
Rheological measurements of 6.3 mL of soymilk sample were
screw-capped Erlenmeyer ?ask. After 10 mL of acetonitrile and
performed using a Haake RS150 Rheometer (ThermoOrion, Kar-
7 mL of Milli-Q system HPLC-grade water (Millipore, Bedford,
lsruhe, Germany) equipped with a DG41 sensor system. The sys-
MA) were added, the ?ask was capped and stirred for 2 h at room
tem consisted of a cup-and-bell-shaped rotor. The inner and
temperature in a rotary shaker (Innova Model 2050, New Bruns-
outer cylinders had an outer diameter of 36.0 and 43.4 mm, respec-
wick Scienti?c, Edison, NJ) at 300 rpm. The mixture was vacuum-
tively. The shear rate was increased from 10 to 1500 sÀ1 within
?ltered (No. 42 ?lter paper, Whatman, Hillsboro, OR), and the
7.5 min. Apparent viscosity was determined at 1500 sÀ1. Samples
?ltrate was evaporated to dryness under vacuum at 630 °C. Dry
were tested a minimum of three times.
matter was dissolved to a ?nal volume of 10 mL with 80% metha-
nol in water. The sample was ?ltered through a 0.45-lm polytetra-
2.10. Statistical analysis
?uoroethylene ?lter unit (Alltech, Deer?eld, IL) and iso?avones
were quanti?ed by HPLC according to the method of Murphy
Least signi?cant differences (LSD) were calculated at a 5% level
et al. (1999). Total iso?avone contents were expressed as lmol/g
using the Statistical Analysis System Version 9.1 (SAS, Cary, NC)
of dry sample. The total iso?avone recovery (%) was the ratio of
software package.
the total iso?avone amount (lmol, db) in soymilk divided by the
total iso?avone amount in soybeans (lmol, db) multiplied by 100.
3. Results and discussion
2.7. Differential scanning calorimetry
3.1. Iso?avone content and pro?le in soybeans, raw soymilk and
Differential scanning calorimetry (DSC) was performed on
thermal processed soymilk
defatted freeze-dried soymilks. Defatting procedure involved
extracting 5 g of freeze-dried sample on a stir plate with 25 mL
The total iso?avone concentration in soybean seeds was
of hexane in the fume hood for 1 h. The mixture was ?ltered
4.81 lmol/g or 2130 lg aglucons/g, which was in the same range re-
using 41-mesh paper ?lters and the procedure was repeated until
ported in previous studies (Hou & Chang, 2002; Murphy et al., 1999;
the hexane fraction was clear. The residual hexane was allowed
Prabhakaran & Perera, 2006; Table 1). Acetyl-b-glucosides were
to evaporate overnight and the defatted samples were stored in
only detected in trace amounts as expected since the acetylated iso-
a desiccator until analyzed. Calibration of the Exstar 6000 Seiko
?avones production requires dry thermal treatment. The aglucons,
II calorimeter (Seiko Instruments, Torrance, CA) was performed
daidzein, glycitein and genistein, represented less than 2% of the
with indium. The freeze-dried soymilk sample (0.1 g) was dis-
mole mass, and were considered as negligible. The malonyl-b-glu-
solved in 0.9 mg of 0.05 M Tris–HCl buffer, pH 7.0, and 13 mg
cosides and b-glucosides were the predominant forms, both repre-
of this solution was hermetically sealed in pre-weighed alumi-
senting 98.3% of the total concentration. Total daidzein mole
num pans. A pan containing buffer was used as reference. Calori-
content was the highest at 51% of the total iso?avones followed
metric measurements were carried out at a 10 °C/min scan rate
by genistein and glycitein with 41 and 7.5%, respectively. This was
from 25 to 110 °C. Enthalpy of thermal denaturation was esti-
very similar to daidzein and genistein, expressed in lmol/g, of Pra-
mated from the DSC curve. Triplicates were run for each sample.
bhakaran and Perera (2006) and Murphy, Barua, and Hauck (2002).
After DSC analysis, pans were punctured on top and samples were
In this study, several soymilks submitted or not to a thermal
dried in a forced-air oven at 130 °C for 3 h to determine their
treatment were prepared. To study the effect of pressure on iso-
moisture contents. The results were expressed as J/g of dry
?avone content and pro?le in soymilk, the soymilk was submitted
protein.
to a thermal treatment at 95 °C for 15 min prior to pressure treat-
Table 1
Iso?avone contents of soybeans, control soymilk and heat-treated soymilks (lmol/g, dry basis)
Glucoside
Malonylglucoside
Aglucon
Total
Din
Gin
Glin
Total
MD
MG
MGl
Total
Dein
Gein
Glein
Total
TD
TG
TGl
TI
Soybeans
0.27a
0.26a
0.12
0.65a
2.12a
1.71a
0.25a
4.08a
0.05a
0.03a
ND
0.08a
2.44
2.01
0.36
4.81
Raw soymilk
0.29a
0.30a
0.12
0.71a
2.02a
1.70a
0.25a
3.97a
0.16a
0.12a
0.05
0.33a
2.47
2.12
0.42
5.01
95 °C Soymilk
0.50b
0.59b
0.14
1.23b
1.55b
1.34b
0.18b
3.07b
0.31b
0.26b
0.07
0.64b
2.27
2.20
0.39
4.86
95/75 °C Soymilk
0.51b
0.60b
0.14
1.25b
1.45b
1.29b
0.18b
2.92b
0.30b
0.26b
0.07
0.63b
2.36
2.19
0.39
4.94
LSD
0.08
0.06
NDif
0.16
0.39
0.34
0.05
0.80
0.14
0.10
NDif
0.24
NDif
NDif
NDif
NDif
Values in the same column with different letters were signi?cantly different (p 6 0.05). Abbreviations: Din: daidzin; Gin: genistin; Glin: glycitin; Dein: daidzein; Gein:
genistein; Glein: glycitein; MD: malonyldaidzin; MG: malonylgenistin; MGl: malonylglycitin. TD: total daidzein; TG: total genistein; TGl: total glycitein; TI: total iso?avones.
95 °C Soymilk was heated at 95 °C for 15 min. 95/75 °C Soymilk was heated at 95 °C for 15 min and 75 °C for 10 min. NDif: Not different.
Please cite this article in press as: Jung, S. et al., Iso?avone pro?les of soymilk as affected by high-pressure treatments ..., Food Chemistry
(2008), doi:10.1016/j.foodchem.2008.04.025

---

ARTICLE IN PRESS
4
S. Jung et al. / Food Chemistry xxx (2008) xxx–xxx
ment. This thermal treatment was chosen to totally inactivate b-
from 400 to 750 MPa at 25 and 75 °C (p > 0.05; Table 2). The total
glucosidases, known to convert glucosides to free aglucons (Matsu-
iso?avone concentration of pressurized soymilks was not statisti-
ura, Obata, & Fukushima, 1989). No residual b-glucosidase activity,
cally different from their 95 °C control soymilks. While the content
determined by the method of Matsuura and Obata (1993), was
of iso?avone was unchanged by the HPP treatment applied alone
found on this thermal-treated soymilk. Effect of pressure on iso-
or combined with a mild thermal treatment, the pro?le of iso?av-
?avone of soymilk was investigated at an initial temperature of
one was modi?ed depending on the initial treatment temperature.
25 and 75 °C. The samples were pre-heated at 25 or 75 °C for
At initial temperature of 25 °C, the distribution of the iso?avones in
10 min, and the controls were prepared in the same conditions
soymilk pressurized at 400 MPa was similar to the corresponding
but not submitted to HPP. When soymilk was prepared from pres-
control soymilks (Table 1). A pressure increase up to 750 MPa did
surized soybeans, the control was prepared from soybeans not sub-
not affect the iso?avone pro?le. Samples pressurized at 500 and
mitted to any thermal treatment and, therefore, corresponded to a
600 MPa and 25 °C were not statistically different from the ones
raw soymilk.
treated at 400 and 750 MPa (data not shown). At 75 °C, the three
The total and individual iso?avone content of raw soymilk and
malonyl-b-glucosides and the b-glucosides decreased and in-
soymilks prepared with thermal treatments is presented in Table 1.
creased, respectively, with increasing pressure while the aglucons
For all the samples, the acetyl-b-glucosides concentration repre-
and acetyl-b-glucosides remained constant (p > 0.05). These results
sented less than 2% mole mass, and considered negligible, there-
suggest that pressure combined with mild temperature promoted
fore, these data are not shown. There was no statistical
the interconversion of the malonyl forms to b-glucosides. Daidzin
difference in the ?nal iso?avone concentration, expressed in
and genistin forms showed greater interconversion than glycitin.
lmol/g (dry basis), between the soybeans and different soymilks.
When compared to the 400 MPa, 25 °C, the same pressure treat-
The changes in the iso?avone distribution depended on thermal
ment at 75 °C converted $0.30 lmol/g of malonyldaidzin and
treatment applied. The raw soymilk had the same iso?avones dis-
malonylgenistin while only 0.04 lmol/g of malonylglycitin was
tribution as the soybeans. In soymilk treated at 95 °C for 15 min as
converted. At the highest pressure and initial temperature of
compared to soybeans and raw soymilk, the distribution of the
75 °C, b-glucosides daidzin and genistin increased by 0.63–
individual iso?avones was shifted towards the b-glucoside and
0.9 lmol/g from the malonyl esters.
aglucon forms at the expense of a decrease of 23% in the malo-
During high-pressure processing, soymilk samples were sub-
nyl-b-glucoside content. There was an equi-mole conversion from
jected to a temperature higher than the initial temperature due
malonylglucosides to b-glucosides and aglucons forms. Small but
to adiabatic heating. This increase of temperature was the result
signi?cant conversion of the b-glucosides to aglucons in 95 °C soy-
of compression heating and depends on the nature of the product,
milk compared to raw soymilk can be attributed to the action of
the initial product temperature, and the pressure level (de Heij
endogenous b-glucosidase. For the raw soymilk, aliquot was imme-
et al., 2003). When samples were pressurized at 25 °C, the average
diately frozen after grinding/?ltration while for the 95 °C soymilk,
quasi-adiabatic temperature increase upon compression ds was
the enzyme might have been active longer before its thermal inac-
1.0 °C/100 MPa. The ds calculated at each pressure was not affected
tivation, leading to the conversion of b-glucosides to corresponding
by pressure level. When initial temperature was 75 °C, the average
aglucons. The additional heating step at 75 °C for 10 min did not
ds was increased to 2.0 °C/100 MPa. The ds value calculated at each
further alter the distribution of the different iso?avone forms of
pressure increased as pressure increased. Similar trends were re-
this sample compared to the control at 95 °C (p > 0.05).
ported for foods with high water content by Patazca et al. (2007).
Soymilk was pressurized from 400 MPa, which was identi?ed in
Therefore, when HPP was applied at an initial temperature of
our laboratory as the minimum pressure reducing the numbers of
25 °C, the temperature of soymilk was approximately 30–35 °C
spoilage microorganisms in soymilk. Soybeans were treated to
for 400–750 MPa treatment, which apparently did not affect iso-
modify component extractability and therefore pressurization
?avone distribution. When HPP was conducted at 75 °C, the tem-
was applied from 100 MPa.
perature of soymilk ranged from 85 to 90 °C for pressures from
400 MPa to 750 MPa. When pro?le obtained after 400 MPa treat-
3.2. Iso?avone content and distribution in pressurized soymilk
ment at 25 °C was compared to the same soymilk treated at
75 °C, an average of 0.6 lmol/g of total glucoside and total malo-
There was no signi?cant effect of pressure and initial tempera-
nylglucoside inter-converted. At the highest pressure, 1.7 lmol/g
ture on the total iso?avone concentration of soymilk pressurized
of total glucoside and malonylglucosides inter-converted. These re-
Table 2
Iso?avone contents of soymilk processed by high-pressure at 25 and 75 °C (lmol/g, dry basis)
Pressure (MPa)
Glucoside
Malonylglucoside
Aglucon
Total
Din
Gin
Glin
Total
MD
MG
MGl
Total
Dein
Gein
Glein
Total
TD
TG
TGl
TI
25 °CA
400
0.50
0.58
0.14
1.22
1.55
1.36
0.18
3.09
0.31
0.26
0.07
0.64
2.39
2.27
0.39
5.06
750
0.55
0.64
0.15
1.34
1.52
1.34
0.18
3.04
0.31
0.26
0.07
0.64
2.43
2.31
0.40
5.14
LSD
NDif
NDif
NDif
NDif
NDif
NDif
NDif
NDif
NDif
NDif
NDif
NDif
NDif
NDif
NDif
NDif
75 °CB
400
0.75a
0.88a
0.18a
1.81a
1.18a
1.06a
0.15a
2.39a
0.30
0.28
0.07
0.65
2.25
2.28
0.40
4.93
500
0.87b
1.03b
0.20ab
2.10b
1.03b
0.94b
0.13b
2.10b
0.30
0.25
0.07
0.62
2.23
2.29
0.40
4.92
600
1.04c
1.22c
0.22bc
2.48c
0.92c
0.85c
0.12c
1.89c
0.31
0.26
0.07
0.64
2.29
2.40
0.41
5.11
700
1.17d
1.38d
0.24c
2.79d
0.77d
0.71d
0.11d
1.59d
0.31
0.26
0.07
0.64
2.26
2.41
0.42
5.09
750
1.21d
1.45d
0.24c
2.90d
0.58e
0.55e
0.08e
1.21e
0.30
0.25
0.07
0.62
2.10
2.31
0.39
4.80
LSD
0.09
0.11
0.02
0.21
0.08
0.05
0.01
0.12
NDif
NDif
NDif
NDif
NDif
NDif
NDif
NDif
Values in the same column with different letters were signi?cantly different (p 6 0.05). Abbreviations: Din: daidzin; Gin: genistin; Glin: glycitin; Dein: daidzein; Gein:
genistein; Glein: glycitein; MD: malonyl-daidzin; MG: malonylgenistin; MGl: malonylglycitin; TD: total daidzein; TG: total genistein; TGl: total glycitein; TI: total iso?avones.
A Samples pressurized at 500 and 600 MPa were not statistically different from the ones treated at 400 and 750 MPa, so the data are not shown. NDif: Not different.
A,B Soymilks control corresponded to the 95 °C and 95/75 °C treated soymilk (Table 1).
Please cite this article in press as: Jung, S. et al., Iso?avone pro?les of soymilk as affected by high-pressure treatments ..., Food Chemistry
(2008), doi:10.1016/j.foodchem.2008.04.025

---

ARTICLE IN PRESS
S. Jung et al. / Food Chemistry xxx (2008) xxx–xxx
5
sults suggested that high-pressure promoted conversion of malo-
with Omi et al. (1996) which was carried out on soybeans pressur-
nyl- to b-glucoside iso?avones due to adiabatic heating. More re-
ized up to 700 MPa at 25 °C for 25 min. However, Omi et al. (1996)
search is needed to precisely calculate the conversion rate
reported 0.2–0.5% of soy proteins release in the soaked-pressurized
constants under high-pressure combined with mild thermal
water and related it to changes in cotyledon surface and epidermal
treatment.
cells of the soybeans as observed by scanning microscopy. In addi-
tion, small changes in the aglucon pro?le of the soymilk obtained
3.3. Iso?avone content and distribution in soymilk prepared from
from pressurized beans might be explained on the basis of residual
pressurized soybeans
endogenous b-glucosidase activity. We found that b-glucosidase
activity in soymilk is stable up to 400 MPa and gradually decrease
The total iso?avone mole recovery in control soymilk was 79%
when pressure is further increased to reach 60% inactivation after
of the 120 lmol originally present in the soybeans. This result sug-
700 MPa treatment (results not shown). Similar decrease of activ-
gested some iso?avone loss in the okara, the insoluble fraction, as
ity of b-glucosidase from strawberry was reported for treatment
only small amounts ($1%) are expected to be lost in the soaking
higher than 500 MPa (Zabetakis, Leclerc, & Kadja, 2000). The de-
water (Jackson et al., 2002; Wang & Murphy, 1996). The total iso-
crease in b-glucosidase activity starting at 500 MPa coincided with
?avone recovery was not modi?ed when the beans were pressur-
the decrease in the aglucons content. Observed iso?avone pro?le
ized from 100 to 700 MPa prior to soymilk production (data not
changes might also be the result of modi?cation of the iso?av-
shown). Pressure level applied to hydrated soybeans has a signi?-
one/protein interactions. Iso?avone pro?le of soymilk from pres-
cant effect on the total iso?avone concentration (lmol/g, db) in
surized
hydrated
soybeans
therefore
probably
resulted
in
soymilk (Table 3). The iso?avone content and pro?le of the soymilk
combined pressure-induced changes in enzymatic, cell structure,
prepared from the 100 MPa processed soybeans was not signi?-
protein of the beans and additional investigation will be needed
cantly different from the control raw soymilk (Tables 1 and 3).
to
determine
the
involvement
of
the
soybean
structural
The lowest concentration of total iso?avones and total glucoside
modi?cation.
in soymilk were observed after pressurization of the soybeans at
300 MPa while the highest were obtained for soymilk prepared
3.4. Characterization of soymilk prepared from pressurized soybeans
with 700 MPa processed soybeans. Treatment at 300 MPa de-
creased the content of iso?avones in the total daidzein, genistein
The weight of soymilk and its solids, fat, crude protein content,
and glycitein by 0.55 lmol/g, 0.30 lmol/g and 0.10 lmol/g, respec-
and apparent viscosity varied depending on the pressure level sub-
tively. This decrease can mainly be explained because of signi?cant
mitted to the hydrated soybeans (Table 4). The yield of soymilk so-
changes in malonyl-b-glucoside contents. The total aglucon con-
lid decreased from 74.8% to 66.4% for the control and the soymilk
tent was stable for pressure up to 400 MPa but progressively de-
obtained from 700 MPa-treated beans, respectively. The protein
creased from 500 MPa to half at 700 MPa. The three aglucons,
content varied from 51.7% to 47.7% (db) for the same samples. This
daidzein, genistein and glycitein, contributed partially to this de-
decrease represented a total variation of 7.7% in the protein con-
crease. Soymilk prepared from 700 MPa processed soybeans con-
tent. Fat content increased with increasing pressure level. Soymilks
tained only 55% of daidzein and genistein compared to the
prepared from 400–700 MPa-treated beans had a 2-fold higher fat
control with no glycitein detected at this pressure treatment.
content than the control. The storage protein pro?le of the soy-
The soymilks treated up to 750 MPa at a temperature of 25 °C
milks was similar to the control regardless of the pressure level ap-
had similar iso?avone content and distribution than the control
plied to the beans (Fig. 1). Thermal properties of soymilks were
as reported above. Therefore, if the assumption that pressure did
determined. The control soymilk exhibited two thermal transitions
not modify either iso?avone content or pro?le in soybeans pres-
at 76 and 95 °C, which were attributed to b-conglycinin and gly-
surized at 25 °C is made, the iso?avone pro?le and content in soy-
cinin, respectively. The enthalpy of glycinin and b-conglycinin from
milk obtained from pressurized soybeans could be explained by
soymilk obtained from pressurized soybeans are summarized in
changes in their water-extractability during soymilk preparation.
Table 5. Enthalpy of denaturation of glycinin and b-conglycinin of
The increase of extraction yield could be due to an increase of mass
soymilk prepared from control soybeans were 8.24 and 1.51 mJ/
transfer by enhancement of solvent penetration into the solid
mg, respectively, which agrees with results of L’hocine, Boye, and
material and release of intracellular product due to disruption of
Arcand (2006). When HPP was applied to the soybeans, the native
cell walls (Houqin, Ruizhan, & Chanzeng, 2007). There was no
state of the extracted glycinin stayed similar for treatment up to
apparent alteration in shape, size, or color between pressurized
300 MPa and decreased signi?cantly at 400 MPa while having
and non-pressurized soybean seeds. This visual observation agrees
denaturation enthalpy lower than 0.4 mJ/mg for 600 and
Table 3
Iso?avone contents of soymilk prepared from pressurized beans (lmol/g, dry basis)
Pressure (MPa)A
Glucoside
Malonylglucoside
Aglucon
Total
Din
Gin
Glin
Total
MD
MG
MGl
Total
Dein
Gein
Glein
Total
TD
TG
TGl
TI
100
0.29bc
0.30ab
0.12a
0.72bc
2.12bc
1.80ab
0.26c
4.17bc
0.15b
0.11a
0.05a
0.31a
2.56a
2.21ab
0.43a
5.20bc
200
0.27c
0.28a
0.12a
0.67c
1.90d
1.65b
0.24d
3.78cd
0.15b
0.11a
0.05a
0.31a
2.32b
2.04a
0.40b
4.76cd
300
0.26c
0.30ab
0.10b
0.66c
1.61e
1.51b
0.19e
3.32d
0.14bc
0.11a
0.04ab
0.30a
2.01c
1.92a
0.34c
4.27d
400
0.36ab
0.36abc
0.14c
0.86ab
2.37a
1.90ab
0.28b
4.55ab
0.16ab
0.11a
0.04ab
0.31a
2.89d
2.37ab
0.46d
5.72ab
500
0.37ab
0.37bc
0.15c
0.89ab
2.26ab
1.82ab
0.31a
4.38ab
0.12cd
0.08b
0.03ab
0.23b
2.75ad
2.26ab
0.49e
5.51ab
600
0.33abc
0.33ab
0.15c
0.81bc
2.28ab
1.82ab
0.30a
4.41ab
0.10de
0.07bc
0.02bc
0.19bc
2.72ad
2.22ab
0.47de
5.41ab
700
0.41a
0.42c
0.17d
1.00a
2.23ab
2.28a
0.31a
4.82a
0.09e
0.06c
0.00c
0.16c
2.74ad
2.76b
0.48de
5.98a
LSD
0.08
0.08
0.02
0.17
0.16
0.54
0.01
0.50
0.03
0.02
0.02
0.06
0.20
0.59
0.03
0.60
Values in the same column with different letters were signi?cantly different (p 6 0.05). Abbreviations: Din: daidzin; Gin: genistin; Glin: glycitin; Dein: daidzein; Gein:
genistein; Glein: glycitein; MD: malonyl-daidzin; MG: malonylgenistin; MGl: malonylglycitin; TD: total daidzein; TG: total genistein; TGl: total glycitein; TI: total iso?avones.
Soymilk control corresponded to raw soymilk (Table 1).
A Pressure level applied to the hydrated beans before soymilk preparation.
Please cite this article in press as: Jung, S. et al., Iso?avone pro?les of soymilk as affected by high-pressure treatments ..., Food Chemistry
(2008), doi:10.1016/j.foodchem.2008.04.025

---

ARTICLE IN PRESS
6
S. Jung et al. / Food Chemistry xxx (2008) xxx–xxx
Table 4
Weight, proximate analysis and apparent viscosity of soymilk prepared from pressurized soybeans
Pressure (MPa)A
Soymilk weightB (g)
Solid content (%)
Crude protein (%, db)
Fat content (%, db)
Apparent viscosity
0.1
270a
6.96a
51.76a
11.10a
1.47a
100
263ab
6.93a
51.20a
13.75ab
1.47a
200
270a
6.95a
51.26a
14.36b
1.55ae
300
264ab
7.05a
51.47a
19.33c
1.85c
400
257bc
6.98a
49.53b
21.81cd
1.97d
500
252c
6.85a
48.91c
23.04d
1.96d
600
250c
6.57b
48.14d
22.93d
1.67b
700
254bc
6.58b
47.70d
23.98d
1.52e
LSD
10
0.24
0.59
3.21
0.06
Values in the same column with different letters were signi?cantly different (p < 0.05).
A Pressure level applied to the beans before soymilk preparation.
B Soymilk was prepared from 28 g of hydrated processed beans with a soybean-to-water ratio of 1:10.
Pressure level (MPa) A
0.1 100 200 300 400 500 600 700
,
7S/?, ?
7S/?
11S/?
11S/?
11S/?
7S: ?-conglycinin; 11S: glycinin; A: acidic subunits; B: basic subunits
A Pressure level applied to the hydrated beans before soymilk preparation.
Fig. 1. SDS PAGE pro?le of soymilk obtained from pressurized soybeans.
displayed a bell-shaped curve with highest value obtained after
Table 5
treatment of the beans at 400 and 500 MPa.
Enthalpy of denaturation of glycinin and b-conglycinin of soymilk prepared from
Our results have shown that there were no apparent differences
pressurized beans
with pressure level in extraction rate of individual glycinin and b-
Pressure (MPa)A
Denaturation enthalpy (mJ/mg of protein)
conglycinin from pressurized soybeans as peptide pro?le remained
unchanged. The solids and fat content, protein content, and native
Glycinin
b-Conglycinin
state of glycinin and b-conglycinin were affected by the HPP pre-
0.1
8.24a
1.51a
treatment. These parameters are known to impact viscosity of
100
8.45a
1.28b
200
8.25a
1.35b
soy products (Cheng, Shimizu, & Kimura, 2005; Forster & Ferrier,
300
7.68b
<0.4c
1979; Liu, Chang, Li, & Tatsumi, 2004; Toda, Chiba, & Ono, 2007;
400
4.70c
<0.4c
Wagner, Sorgentini, & Anon, 1992) and variations in these param-
500
0.85d
<0.4c
eters probably contributed to the small but signi?cant change ob-
600
<0.4e
<0.4c
served in the apparent viscosity of soymilk prepared from
700
<0.4e
<0.4c
pressurized soybeans, yet the mechanism responsible for the ob-
LSD
0.18
0.20
served changes in viscosity is not clear.
Values in the same column with different letters were signi?cantly different
(p < 0.05). <0.4: denaturation enthalpy was lower than 0.4 mJ/mg of proteins.
A
4. Conclusions
Pressure level applied to the beans before soymilk preparation.
This is the ?rst study reporting the effect of high-pressure pro-
cessing on the stability, conversion and water-extractability of soy
700 MPa treatment. b-Conglycinin enthalpy decreased by 15% after
iso?avones. The pressure alone did not affect the iso?avone con-
200–300 MPa treatment of the beans&nb