ANNALS OF

PHYTOMEDICINE
62
Annals of Phytomedicine 1(1): 62-68, 2012 An International Journal

Effect of processing conditions on saponin content and

antioxidant activity of Indian varieties of soybean
(Glycine max Linn.)
Shivani Chaturvedi, R. Hemamalini and Sunil K. Khare
Enzyme and Microbial Biochemistry Laboratory, Department of Chemistry,
Indian Institute of Technology, Delhi, Hauz Khas, New Delhi-110016, India.
Received for publication May 1, 2012; accepted June 1, 2012

Abstract
Saponin content and antioxidant activity of three major Indian varieties, i.e., Punjab
long, Hardee and Kalitur were analyzed. These contained 80.4 g/kg, 49.8 g/kg and
119.7 g/kg saponin in soyflour, respectively. The black variety ‘Kalitur’ showed a
higher saponin content and antioxidant activity as compared to the other two
yellow varieties. Saponins A and B were present as major saponins. Effect of
different processing conditions on soysaponin content of soybean cake were
evaluated. A variety of commonly prevalent processing practices including boiling,
soaking, sprouting and drying were included. The saponin content reduced on
boiling and defattening. Soaking and sprouting under 50oC temperature increased
the saponin content. Saponins exhibited significant antioxidant activity which
can be attributed to saponins A and B.
Key words: Soybeans; Soysaponins; HPLC; Processing conditions; Antioxidant
activity.
Introduction class-B is monodesmosidic, having only one glycosylation
site and two different aglycon moieties (soyasapogenol-B
Saponins are a class of oleanane triterpenoids which and E) (Shiraiwa et al., 1991). Soysaponin class-E is only a
commonly occur in different plants (Hostettman and Marston, transitory form which quickly converts into either of the other
1995). These are secondary metabolites, having important two classes. The type of saponin and its content largely
biological activity like anticarcinogenic, hypocholesterolemic, depends on the soya variety, agroclimatic factors and the
haemolytic (Francis et al., 2002), immunostimulatory, antiviral conditions of growth. Zou and Chang (2011) have reported
and hepatoprotective agent (Wang et al., 2007). The saponin the difference in saponin content of the commonly grown
content in soybean is 6.500 g kg-1, which is the highest among soybean varieties.
legumes (Sagratini et al., 2009).
Soybean having a black seed coat, commonly called ‘black
The saponins present in soybean are collectively termed soybean’, has widespread consumption in Asia. It is rich in
soysaponins. The saponins are divided into two major protein and also contains important bioactive components
categories-A and B, based on their aglycone structures. like isoflavones, anthocyanins, saponins and vitamin E
Soysaponin class-A includes bidesmosidic saponins and (Messina, 1999). Recent studies have shown enormous
consists of two glycosylation sites with one aglycone moiety health benefits of black soybean. It helps in preventing
(soysapogenol-A ) (Heng et al., 2006). The soysaponin cardiovascular diseases, breast cancer and it also acts as an
immunostimulatory agent (Xu and Chang, 2008).
Since, the soybean is consumed in the form of various food
Author for correspondence: Dr. S.K. Khare
Professor, Enzyme and Microbial Biochemistry Laboratory, products according to regional preferences, it is important to
Department of Chemistry, Indian Institute of Technology, Delhi, understand the effects of various processing methods on
Hauz Khas, New Delhi-110016, India. the availability of the bioactive component in it. The present
E-mail: skkhare@chemistry.iitd.ac.in; skhare@rocketmail.com work was done to mainly address the effect of processing on
Phone: +91-11-26596533, Fax: +91-11-26581102
Cell: +91-09350408194 the saponin contents of soybean. Soybean is generally
 63

cooked before consumption and processed by different minutes at room temperature. The supernatant was collected
methods. The commonly used processing methods are and dried in a flash evaporator. The residue was dissolved in
boiling, germination, soaking, sprouting and drying (Della et 1 ml butanol-water mixture (1:1 ratio). This was kept overnight
al., 1994). These are known to improve the nutritional status and was dried again. The dried mixture was dissolved in 2 ml
and taste. But the individual effect of temperature, crushing, methanol and was analysed by HPLC.
incubation etc., may cause an alteration in the saponin level
HPLC analysis of extracted saponins
and their bioavailability (Chau et al., 1997). The saponin
content and its antioxidant activity are likely to diminish The extracted saponins were analyzed by HPLC (Agilent
during harsh processing conditions. No systematic study Technologies, 1200 series, USA) equipped with UV detector.
has been done in this aspect, especially with reference to A C-18 reverse phase column was used. Conditions were
black soybean. The studies on the impact of soaking and followed as per the procedure of Zhang et al. (2009).
cooking on the saponin contents are limited and often Trifluoroacetic acid (0.05%) in water (solvent A) and
contradictory. This study aimed to compare the effects of the acetonitrile (solvent B) were used as solvents. The flow rate
different processing conditions and varietal differences on was kept at 1ml/min with elution under gradient condition.
the saponin content and the antioxidant activity. The solvent A started with 60% and decreased up to 52%
over 60 min. The peaks were analyzed by UV-VIS detector.
The level of saponin in Kalitur variety is being reported first
time. The effect of processing condition, viz. soaking, ESI- MS of saponins
sprouting, drying, roasting and boiling etc., on saponin and
Individual peaks of soysaponins were collected by HPLC and
antioxidant activity of soybean meal of Punjab long
analyzed by ESI-MS on a Bruker Compass MicroTOF-QII.
(commonly used in India) is also presented.
Antioxidant activity of saponins
Materials and Methods
The cupric ion reducing antioxidant capacity of soybean was
Materials determined according to the method of Apak et al. (2008).
Three cultivars of soybean (Glycine max Linn.) Punjab long, Briefly, 0.1 ml of sample was mixed with 1ml each of copper
Hardee and Kalitur were obtained from ‘National Research chloride solution (1.0 × 10"2 mol/l), neocuproine alcoholic
Centre for Soybean, Indore, India’. solution (7.5 × 10"3 mol/l), and ammonium acetate buffer (1
mol/l, pH 7.0) solution and water. The mixture was incubated
Preparation of defatted soybean meal and other processing for 30 min. and absorbance was recorded at 450 nm against
treatments the reagent blank. Trolox was used as the standard. The
The soybean flour was prepared by milling the beans at results were expressed as µmol Trolox /g, using molar
‘Centre for Rural Development and Technology’ facility. The absorptivity of Trolox as 1.67x104 l mol-1 cm-1.
flour was subjected to different processing treatments. (i) It Results and Discussion
was defatted for 48 h with hexane, (ii) boiling of flour in an
open vessel without pressure for 5 min. at 100 ÚC, (iii) boiling Soysaponins in different variety of soybeans
of the flour in a closed cooker with pressure (1.4 psi) for 5
Saponin contents from full fat soybean cake were extracted
min, (iv) soaking of seeds for 6 h at 30oC followed by sprouting by methanol and analysed by HPLC . The typical HPLC
for 72 h at 30oC, (v) soaking of seeds for 6 h at 30oC, (vi) chromatogram of saponin extracted from the full fat soy flour
ultraviolet irradiation of the flour for 20 min., (vii) microwave (FFSF) of all the three varieties are shown in Figure1. The
based processing of the flour for 5 min., and (viii) roasting of results show presence of four major saponins and five minor
the flour for 5 min. derivatives in each of the Indian soybean varieties, Punjab
long, Hardee and Kalitur.
Distilled water, pH 6.9 was used for all processing. The wet
samples were air-dried at room temperature inside a laminar Soybean is reported to have nine types of saponin or its
hood for 48 h so that any type of impurities could be avoided. derivatives (Zhang and Popovich, 2008). In the present study,
These dried samples were grounded properly into powder four major and five minor peaks by FFSF are shown. The first
and sieved through a sieve 8-mesh size and then powdered major peak from HPLC chromatogram appeared between 13.3-
samples were stored in dark brown, air-tight bottles at -80 ÚC 13.9 min. The second and the third major peaks appeared
for further extraction process. between 16.9-17.7 and 17.6-19.0 min. The fourth major peak
appeared between 25.4-26.8 min. Zhang and Popovich (2008)
Extraction of saponins
have previously reported four major peaks and identified them
Ten grams of soybean flour was mixed with 50 ml methanol in as Group I-B non DDMP soysaponin, Group III-B non DDMP
a capped and sealed flask and stirred for 48 h at 120 rpm in a soysaponin, soysapogenol-A and B Group-B DDMP
shaker. The samples were then centrifuged at 8000 rpm for 10
64

soysaponin, respectively. The five minor peaks were also in saponins are heat labile. The level of saponin loss is
agreement with their reports. Individual peaks and their consistent with previous reports by Xu and Chang (2009)
identification are summarized in Table 1. and Wang et al. (2009), where 21.9 and 15.9% reductions are
observed after cooking (boiling).
Punjab long variety contain highest amount of I-Group-B
Non-DDMP Soy Saponin, Soy Sapogenol B, Ba and Bg – Boiling under pressure (1.4 psi) caused more reduction
GroupB-DDMP Soy Saponin. Kalitur variety contain highest (53.11%) as compared to open vessel cooking. Elhardallow
amount of III GroupB-non- DDMP Soy Saponin, Soy and Walker (1994) also found significant reduction (60%) in
Sapogenol A, Ba group B-DDMP Soy Saponins and Ya Group saponin content after closed pressure cooking of the lentils.
B-DDMP Soy saponins. Pressure and ordinary cooking of soaked legumes also reduce
the saponin content by 28-38% (Duhan et al., 2001). Simple
Punjab long and Kalitur varieties have about equal amount
soaking of seeds led to a 32.2% increase in saponin content.
of Soy sapogenol (A and B), which are the major and well
This is in contrast with earlier reports, wherein saponin content
characterized saponins in soybean.
was found to decrease due to leaching out of saponin in
The total saponin content in FFSF of each soybean variety is water (Kataria et al.,1988). However, short duration (6 h)
summarized in Figure 2. It is evident that Kalitur, the black soaking of navy beans did not complete the hydration of
variety of soybean showed significantly higher values of beans so the water soluble saponin content remains as such
saponin content in comparison to Punjab long and Hardee. in the sample so more or less similar amount of saponin was
There are no previous reports of the quantification of saponin found in soaked samples as compared to unprocessed sample
content in these varieties. Due to the lack of purification (Shi et al., 2009).
standards and an effective quantification method of
The exposure of the seeds to microwave radiation, prior to
saponins, the research on the effect of processing conditions
grinding to prepare flour results in increase in the content of
on saponin content has received less attention than
saponin by 39.2%. An increase in the isoflavone content
isoflavones (Rickert et al., 2004).
under similar processing conditions was reported by Aguiar
The antioxidant activity of the saponins et al. (2010). They found that their processing methods
Antioxidant activity of saponins extracted from each variety decreased the malonyl form while aglycon and glycoside
is shown in Figure 3. The Punjab long and Kalitur have almost forms increased. The saponin content has aglycone and
same total antioxidant activity, though Kalitur has much higher glycoside moieties in it, so perhaps because of this reason
saponin content (Figure 2). However, they have same amount our experiment also showed an increase in saponin values.
of soy sapogenol content which might be contributing to the Microwave treatment has been reported to reduce the
antioxidant property. Lee et al. (2002). also found varietal antinutritional factors in some legumes (El-Beltagy, 1996).
difference and processing conditions, affect the antioxidant However, the present study has displayed the opposite result.
activity. Roasting of soy flour also caused 23.38% increase in saponin
content. The effect of roasting has been previously reported
Effect of processing method on the saponin contents of
in green gram where a reduction in saponin content was
soybean flour
observed, (Kataria et al., 1988). However, some researchers
Processing methods are known to affect the availability of reported that saponins are relatively heat stable components
nutrients and bioactive constituents of the seed cakes (Chau (Oenning et al., 1994).
et al., 1997). The impact of processing on saponins are rarely
From the results, it can be concluded that there was a
available in literature (Rickert et al., 2004). In the present
substantial reduction in the saponin content of soyabean
study, the soybean flour of Punjab long was subjected to
flour when it was processed through defatting and boiling
commonly employed processing conditions. The residual
(in a closed vessel) but increased in soaking, sprouting,
content was monitored by HPLC. The changes in total
roasting, exposure to ultra violet radiation and exposure to
saponins and individual saponins content are also shown in
microwave irradiation.
Tables 2 and 3, respectively. It is evident that soaking, boiling
and defatting caused a substantial reduction in the saponin ESI-MS of soysaponin samples
content of soybean flour. The saponin content was reduced
The molecular mass of the different peaks of soysaponin
by about 18% on defatting. The removal of saponins may be
separated by HPLC were analyzed by ESI-MS. The peak
due to their dissolution in n-hexane that had been used for
numbers 6, 7 and 2 (Table.1) gave mass values corresponding
defatting the soybean flour. The free aglycons may be
to 1035.86, 1225.07 and 568.29. These were identified as for
extracted by hexane during defatting (Booth et al., 1960).
the peaks-soysaponin βa, soysaponin βg, and soysaponin
Boiling in an open vessel caused a 28.6 % reduction in the
III. The values for soysapogenol A and soysapogenol B
saponin content. This may be due to the fact that some
correspond to the values reported in literature viz., 473.3 and
 65

457.4 (Zhang et al., 2008). The results confirm the presence 2. While producing soy foods it should be kept in mind
of major types of saponins in Indian soybean varieties Punjab that excessive boiling might leads to a loss of saponins.
long, Hardee and Kalitur.
3. The processing methods such as microwave treatment,
Conclusions soaking or sprouting are better techniques to improve
the saponin content.
The black variety ‘Kalitur was found superior in the saponin
content. The processing conditions especially boiling 4. Among individual saponins, soy sapogenols A and B are
reduced the saponin significantly whereas soaking and important for antioxidant activity.
sprouting increased the saponin content. The study has the Acknowledgements
following important implications:
The authors are highly grateful to Indian Institute of
1. ‘Kalitur’ which is a non-edible variety of soybean can be
Technology, Delhi, New Delhi for providing research facilities.
very useful for obtaining saponin at high concentration,
The financial support for the study provided by Ministry of
which can be marketed as cholesterol-lowering and
Food Processing Industry, New Delhi is gratefully
anticarcinogenic agents.
acknowledged.

(a)

(b)

(c)
Figure 1. HPLC chromatogram of saponin extracted from the full fat soyflour of three soybean varieties (a) Punjab long (b)
Hardee (c) Kalitur
66

Figure 2. Total saponin content in full fat soyflour of Indian soybean varieties (a) Punjab long (b) Hardee (c) Kalitur

Figure 3. Antioxidant activity of saponins extracted from Indian soybean varieties (a) Punjab long (b) Hardee (c) Kalitur
Table 1. Saponin types in full fat soy flour (FFSF) of different soybean varieties
Peak No. Retention Saponin Relative amount of individual saponins
time (min) “Punjab long”a “Hardee” “Kalitur”

1. 13.3 I-Group-B Non-DDMP Soy Saponin 100 (-)24.6 (-)21.4

2. 16.9 III-Group-B Non-DDMP Soy Saponin 100 (-)72.2 (+)37.3
3. 17.5 Soy Sapogenol-A 100 (-)61.6 (+)51.6
4. 18.9 Soy Sapogenol-B 100 (-)55.7 (-)40.3
5. 21.0 Be-Group-B -DDMP Soy Saponin 100 nd nd
6. 24.6 Bg-Group-B -DDMP Soy Saponin 100 nd nd
7. 25.7 Ba-Group-B -DDMP Soy Saponin 100 (-)78.8 (+)16.2
8. 31.9 Not identified nd (-)39.2 (+)22.2
9. 37.5 Yg-Group-B -DDMP Soy Saponin 100 (-)67.6 (-)54.8
10. 46.3 Ya-Group-B -DDMP Soy Saponin 100 (+)8.8 (+)15.2
a
The content of Punjab long variety are taken as 100, nd- Not detected
 67

Table 2. Effect of various processing conditions on saponin content of Punjab long variety of soybean
Processing conditions Soy saponin Percent residual
content (g/kg) saponin respect to
unprocessed (control)
Full fat soy flour without any process (control) 80.4±1.32 100
Defatting 73.67±1.15 92
Boiling in an open vessel 57.4±1.36 71
Pressure cooking 37.7±0.93 47
Soaking of seed (6 h) and than sprouting (72 h) at 30ºC 87±1.52 108
Soaking of seeds for 6 h 106.3±2.77 132
Flour exposed to Ultra violet rays for 20 min 92.4±1.91 115
Microwave cooking for 5 min 111.9±2.06 139
Roasting 99.2±1.75 123

Table 3. Effect of processing conditions on individual saponin content of Full Fat Soya Flour (Punjab long variety of soybean)

Type of saponin Unprocesse Defatted Boiling of Pressure Soaking of Soaking of Microwave Roasting
d FFSF a Soy flour open cooking the seed (6 h) seed (6 h) cooking
(control) Flour vessel flour followed by 5 min
(DSF) sprouting
(72 h) at 30
˚C

I-Group-B Non-DDMP Soy 100 (+)15.0 (+)0.4 (-)2.2 (-)17.9 (-)55.4 (+)67.2 (-)53.1
Saponin
III-Group-B Non-DDMP Soy 100 (+)93.5 (-)43.8 (+)33.7 (+)20.8 (+)40.8 (+)37.1 (+)42.5
Saponin
Soy Sapogenol-A 100 (+)53.5 (-)3.5 (+)36.7 (+)39.2 (+)57.4 (+)51.7 (+)52.2

Soy Sapogenol-B 100 (-)41.8 (+)45.9 (+)45.5 (-)59.9 (+)30.4 (-)59.5 (+)35.5

Be-Group-B -DDMP Soy Saponin 100 nd nd nd nd nd nd nd

Bg-Group-B -DDMP Soy Saponin 100 nd (+)66.5 (+)87.0 (+)69.1 (+)82.0 (+)86.8 (+)89.6

Ba-Group-B -DDMP Soy Saponin 100 (+)25.5 (-)78.3 (-)73.7 (-)69.7 (-)72.3 (-)14.6 (-)13.9

Un identified nd nd nd (-)69.6 (-)81.2 (-)77.1 (-)85.1 (-)77.1

Yg-Group-B -DDMP Soy Saponin 100 (+)17.11 nd nd nd nd (+)4.7 (+)25.0

Ya-Group-B -DDMP Soy Saponin 100 (+)31.5 (+)56.2 (-)32.5 nd nd nd (+)4.6

a
The value of FFSF (control) was taken as 100%, nd Not detected

References Della, V.; Balle, G.; Quillien, L., and Gueguen, J. (1994). Relationships
between processing conditions and starch and protein modifications
Aguiar, C.L.; Canniatti-Brazaca, S.G.; Park, Y. K. and Alencar, S. M. during extrusion-cooking of pea flour. J. Sci. Food Agric., 64(4):
(2010). Effects of the Processing Techniques on Isoflavone Profiles. 509–517.
In: Isoflavones: Biosynthesis, Occurrence and Health Effects,(ed. Duhan, A.; Khetarpual, N. and Bishnoi, S. (2001). Saponin content
Thompson, M. J), Nova Science Publishers Inc., New York, USA, and trypsin inhibitor activity in processed and cooked Pigeon pea
pp: 225-237, cultivars. Int. J. Food Sci. Nutr., 52(1): 53–59.
Apak, R.; Guclu, K.; Ozyurek, M. and Celik, S.E. (2008). Mechanism El-Beltagy, A. (1996). Eûect of home traditional methods on quality
of antioxidant capacity assaysand the CUPRAC (cupric ion reducing aspected of some legumes. MS Thesis, Faculty of Agriculture, Menoûya
antioxidant capacity) assay. Microchim. Acta, 160 (4): 413-419. University, Shibin El-Kom, Egypt.
Booth, A. N.; Bickoff, M. and Köhler, G. O. (1960). Estrogen-like Francis, G.; Kerem, Z.; Makkar, H. P. S. and Baker, K. (2002). The
activity in vegetable. oils and mill by-products. Science, 131(3416): biological action of saponins in animal system: A review. Brit.J. Nutr.,
1807-1808. 88(6):587-605.
Chau, C. F.; Cheung, P. C. and Wong, Y. S. (1997). Effect of cooking Heng, L.; Vincken, J. P.; Hoppe, K.; Koningsveld, G. A.; Decroos, K.;
on content of amino acids and antinutrients in three Chinese indigenous Gruppen, H.; Boekel, M. A. J. S. and Voragen, A. G. J. (2006). Stability
legume seeds. J. Sci. Food Agric., 75 (4): 447-452. of pea DDMP saponin and the mechanism of its decomposition.
Food Chem., 99(2):326–334.
68

Hostettman, K. A. and Marston, A. (1995). Saponins, Cambridge Zhang, W.; Teng, S. P. and Popovich, D. G. (2009). Generation of
University Press, Cambridge, United Kingdom. Group B soysaponins I and III by hydrolysis. J. Agric. Food. Chem.,
57(9):3620-3625.
Jyothi, T. C.; Sindhu, K. and Rao, A. G. A. (2007). Influence of
germination on saponin in soybean and recovery of soy sapogenol-I. Zhang, W.; Yeo, M. C.; Tang, F. Y. and Popovich, D. G. (2009).
J. Food Biochem., 31(1):1-13. Bioactive responses of Hep-G2 cells to soysaponin extracts differs
with respect to extraction conditions. Food Chem. Toxicol.,
Kataria, A.; Chauhan, B. M. and Gandhi, S. (1988). Effect of domestic 47(9):2202-2208.
processing and cooking on the antinutrient of black gram. Food Chem.,
30(2):149-156. Zou, Y. and Chang, S. K. C. (2011). Effect of black soybean extract
on the suppression of the proliferation of human AGS gastric cancer
Lee, S. J.; Chung, I. M.; Ahn, J. K.; Lee, S. K.; Kim, S. H., and Yoo, cells via the induction of apoptosis. J. Agric. Food. Chem.,
N.H. (2002). Variation in antioxidative activity of soybean (Glycine 59(9):4597–4605.
max L.) varieties with crop year and duration of storage time. Food
Sc.Biotechnol., 11:649–653.

Meredith, T. (2002). Solbar plant extracts, Ashdod, Israel, Personal

communication.

Messina, M. J. (1999). Legumes and soybeans: Overview of their

nutritional profiles and health effects. Am.J.Clin.Nutr. 70(3
Suppl.):439S-450S.

Oenning, G.; Jullerat, M.; Fay, L. and Asp, N. (1994). Degradation of

oat saponins during heat processing effect of pH, stainless steel and
iron at different temperatures. J. Agric. Food. Chem., 42(11):2578–
2582.

Rickert, D. A.; Meyer, M. A.; Hu, J., and Murphy, P. A. (2004).Effect

of extraction pH and temperature on isoflavone and saponin
partitioning and profile during soy protein isolate production. J.Food
Sc., 69(8):C623-C631.

Sagratini, G.; Zuo, Y.; Caprioli, G.; Cristalli, G.; Giardinà, D.; Maggi, F.;
Molin, L.; Ricciutelli, M.; Traldi, P. and Vittori, S. (2009).
Quantification of soyasaponins I and βg in Italian lentil seeds by
solid-phase extraction (SPE) and high-performance liquid
chromatography-mass spectrometry (HPLC-MS). J. Agric. Food.
Chem.., 57(23):11226-11233.

Shi, J.; Xue, S. J.; Mab, Y.; Li, D.; Kakuda, Y. and Lan, Y. (2009).
Kinetic study of saponins B stability in navy beans under different
processing conditions. J.Food Eng. 93: 59–65.

Shiraiwa, M.; Harada, K. and Okubo, K. (1991). Composition and

structure of “group B saponin” in soybean seed. Agr.Biol. Chem.
Tokyo, 55(4): 911-917.

Wang, N.; Hatcher, D. W.; Toews, R. and Gawalko, E.J. (2009).

Influence of cooking and dehulling on nutritional composition of
several varieties of lentils (Lens culinaris). LWT- Food Sc.Technol.
42(4):842-848.

Wang, Y.; Zhang, Y.; Zhu, Z.; Zhu, S.; Li, Y.; Li, M. and Yu, B. (2007).
Exploration of thecorrelation between the structure, hemolytic
activity, and cytotoxicity of steroid saponins. Bioorg.Med.Chem.,
15(7):2528–2532.

Xu, B. J. and Chang, S. K. C. (2008). Antioxidant capacity of seed

coat, dehulled bean, and whole black soybeans in relation to their
distributions of total phenolics, phenolic acid, anthocyanins and
isoflavone. J. Agric. Food. Chem., 56(18): 8365-8373.

Xu, B. and Chang, S. K. C. (2009). Phytochemical profiles and health

promoting effects of cool-season food legumes as influenced by thermal
processing. J. Agric. Food. Chem., 57(22):10718-10731.

Zhang, W. and Popovich, D. G. (2008). Effect of soyasapogenol A

and soyasapogenol B concentrated extracts on HEP-G2 cell
proliferation and apoptosis. J. Agric. Food. Chem., 56(8):2603–2608.