11174 J. Agric. Food Chem.

2009, 57, 11174–11185

DOI:10.1021/jf902609n

Antioxidant Properties, Phytochemical Composition, and

Antiproliferative Activity of Maryland-Grown Soybeans with
Colored Seed Coats
MARGARET SLAVIN,† WILLIAM KENWORTHY,‡ AND LIANGLI (LUCY) YU*,†
†
Department of Nutrition and Food Science, ‡Department of Plant Science and Landscape Architecture,
and University of Maryland, College Park, Maryland 20742

This study characterized and compared 18 colored seed coat soybeans for the isoflavone, total
phenolic, and cyanidin-3-glucoside (Cy-3-glc) contents of their flour extracts and the fatty acid
composition and carotenoid and R-tocopherol contents of their oils. Antioxidant assays also
assessed activity of the flour extracts against peroxyl, hydroxyl, and ABTSþ radicals. Black seed
coat soybeans had the highest TPC, ORAC, HOSC, and ABTSþ radical scavenging values, in
addition to the highest isoflavone content, and were the only color to contain Cy-3-glc. Five
soybeans (two black and one each brown, yellow, and green) were selected to test their effects
on HT-29 human colorectal cancer cell growth. The effects of the hydrolyzed and unhydrolyzed
extracts were compared to an aglycone isoflavone standard mixture of the same total molar
concentration as the highest soybean concentration of 15 mg of seed flour equiv/mL of treatment
media. All high doses of hydrolyzed soybean treatments except the green genotype significantly
reduced cell number compared to control at 3 h of treatment time, whereas the high dose of
isoflavone standard treatment took 72 h to show a significant reduction (P < 0.05).

KEYWORDS: Antioxidant; isoflavone; soybean; HT-29; tocopherol; carotenoid; free radical; antiproli-
feration

INTRODUCTION Interestingly, a few studies have indicated that soybeans with

Consumption of soy foods has been recognized to lower the black, brown, green, and yellow seed coats might differ in their
risk of aging-associated diseases, including cardiovascular disease antioxidant properties, flavonoid levels, total phenolic contents,
and cancer, among others (1-3). These health benefits have often and proanthocyanidins (4-6), indicating that this may alter their
been studied in relation to a particular soy component. Isofla- ability to affect health. Soybean extracts (4, 7) and peptides (8)
vones, a class of flavonoids found almost exclusively in legumes have both been shown to reduce the oxidation of LDL cholesterol
and most prominently in soy, have been studied heavily in this in vitro (4,7) or in rats (8), but black soybeans have been shown to
regard. There are three aglycone isoflavones;genistein, daidzein, have greater inhibitory effect against lipid peroxidation in human
and glycitein. Each aglycone has three derivatives based on the LDL than yellow ones (4). Additionally, in vitro studies have
placement of sugar constituents;β-, malonyl-, and acetyl-gluco- shown that the isoflavone genistein and anthocyanins are inde-
side. The aglycone forms are absorbed more rapidly by intestinal pendently capable of inhibiting the growth of cancer cells through
cells due to their less polar structure. Additionally, enzymes various mechanisms (2, 9). Because black soybeans are the only
present in the intestine are capable of cleaving the glucosides to color reported to contain anthocyanins (10, 11) and only brown
their aglycones, thus providing for better absorption. Isoflavones’ and black soybeans contain proanthocyanins, this may result in a
antioxidant abilities are well-known, and their biological activity differing level or type of bioactivity among soybeans with
has been demonstrated in a great number of studies;they inhibit different seed coat colors. Also, these data suggest the possibility
cancer cell growth in vitro, prevent tumor development in animal of developing novel soybean lines with a selected seed coat color
models, inhibit LDL oxidation, are capable of binding to estrogen to be used as bioactive ingredients in functional foods targeting
receptors, and inhibit bone resorption by osteoclasts, among a list different health problems.
of other activities (2, 3). These results suggest that isoflavones The farm value for soybeans has varied greatly in recent years
may be at least in part responsible for the health benefits and is highly dependent on factors well beyond the control of
associated with soy food consumption mentioned above. farmers, including international demand, weather conditions,
and market supply size (12). If a high-demand, consistent retail
*Address correspondence to this author at the Department of
outlet can be established for a novel soybean line because of its
Nutrition and Food Science, 0112 Skinner Building, University of special health properties, small farms growing it may be better
Maryland, College Park, MD 20742 [telephone (301)-405-0761; fax able to withstand the uncertain market and retain more consistent
(301)-314-3313; e-mail [email protected]]. profitability. As part of our continuous effort to enhance the

pubs.acs.org/JAFC Published on Web 11/09/2009 © 2009 American Chemical Society

Article J. Agric. Food Chem., Vol. 57, No. 23, 2009 11175
quality and value of Maryland-grown soybeans, this study Quantification was based on the area under each fatty acid peak as
aimed to characterize the antioxidant activity and phyto- compared to the total area of all fatty acid peaks.
chemical composition of Maryland-grown soybeans with Carotenoid and r-Tocopherol Contents. Oil samples and standards
various seed coat colors. Furthermore, this study sought to were dissolved in hexane and analyzed via normal phase liquid chromato-
compare the phytochemical composition of extracts of these graphy-atmospheric pressure chemical ionization-tandem mass spectro-
soybeans with their bioactivity in HT-29 human colorectal metry (NP-LC-APCI-MS/MS) according to the method of Hao et al. (16).
Briefly, a Zorbax RX-SIL column, 2.1 mm i.d.150 mm, 5 μm particle size
cancer cells, as compared to the activity of treatments of pure
(Agilent Technologies, Palo Alto, CA), was used at ambient temperature.
isoflavones.
Separation was achieved through gradient elution with a flow of 0.5 mL/
min, with hexane for solvent A and 1% isopropanol in EtOAc for solvent
MATERIALS AND METHODS B. A 5 min linear gradient from 1 to 10% solvent B was followed by a
15 min linear gradient from 10 to 50% B. The column was allowed to
Materials and Chemicals. Whole soybeans of brown, green, yellow, re-equilibrate at initial conditions for 10 min prior to injection of the next
and black seed coat colors from the 2007 growing season grown at a single sample. The injection volume was 5 μL. Quantification was done using
location in Maryland were obtained from Dr. William Kenworthy of the total ion counts compared to that of external standards.
Department of Plant Science and Landscape Architecture, University of
Total Phenolic Contents (TPC). The TPC of each soybean extract
Maryland (College Park, MD). In Maryland, 2007 was a drought season,
was measured according to a laboratory procedure described pre-
which should be considered in interpreting results due to the well-accepted
viously (17). The reaction mixture comprised 100 μL of 50% acetone
link between growing conditions and levels of various phytochemicals,
soybean extract, 500 μL of Folin-Ciocalteu reagent, 1.5 mL of 20%
including isoflavones and tocopherols.
sodium carbonate, and 1.5 mL of ultrapure water. Absorbance was read at
ABTS chromophore diammonium salt was manufactured by Calbio-
765 nm on a Thermo Spectronic (Waltham, MA) Genesys spectrophot-
chem (San Diego, CA). Iron (III) chloride, fluorescein (FL), biotech grade
ometer after 2 h of reaction at ambient temperature. Different concentra-
DMSO, 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid (Trolox),
tions of gallic acid were used to create the standard curve. Reactions were
2,2-diphenyl-1-picrylhydrazyl radical (DPPH•), genistein, daidzein, cyani-
conducted in triplicate, and results were reported as milligrams of gallic
din, lutein, β-carotene, β-cryptoxanthin, zeaxanthin, and R-, δ-, and
acid equivalents (GAE) per gram of soybean flour.
γ-tocopherols were purchased from Sigma-Aldrich (St. Louis, MO).
2,20 -Azinobis(2-amidinopropane) dihydrochloride (AAPH) was pur-
Isoflavone Composition. Fifty percent acetone soybean extracts were
hydrolyzed with acid, redissolved in methanol, and filtered through a 0.45 μm
chased from Wako Chemicals USA (Richmond, VA). Thirty percent
ACS grade H2O2 was purchased from Fisher Scientific (Fair Lawn, NJ). syringe filter prior to being subjected to HPLC analysis. The protocol by
Glycitein was obtained from Indofine Chemical Co. (Hillsborough, NJ). Lee et al. (18) was followed with modifications. The column used was
Cyanidin-3-glucoside was purchased from Polyphenols Laboratory AS a Phenomenex (Torrance, CA) Gemini C18 column (150 mm  4.6 mm 
(Norway). Ultrapure water was manufactured by Cayman Chemical Co. 5.0 μm) and was housed in an oven set to 40 °C. A binary solvent system
(Ann Arbor, MI) and was used for all experiments. An ATP-Lite 1step was employed with solvent A consisting of water/acetic acid at a ratio of
Luminescence Assay System was obtained from Perkin-Elmer (Waltham, 99.9:0.1 (v/v) and solvent B consisting of acetonitrile/acetic acid (99.9:0.1,
MA). All cell culture media components were purchased from Invitrogen v/v). The gradient changed linearly from 75:25 (A:B, v/v) to 67:33 (A:B, v/
(Carlsbad, CA). All other chemicals and solvents were of the highest v) from 0 to 20 min and then was returned to initial conditions for 5 min to
commercial grade and used without further purification. re-equilibrate the column prior to the next run. The flow rate was 1 mL/
Seed Coat Color. A monolayer of whole, undamaged soybeans was min, the injection volume was 10 μL, and the detection wavelength was set
placed in a clean glass sample container and analyzed for Hunter color to 254 nm. The isoflavones were identified and quantified via comparison
values L*, a*, and b* using a HunterLab ColorFlex (Reston, VA) to external standards.
according to the manufacturer’s directions. Color value was obtained Cyanidin-3-glucoside (Cy-3-glc) Contents. Cy-3-glc was deter-
using a D65/10° (daylight 65 illuminant/10° observer) setting. Triplicate mined because it is the predominant anthocyanin present in black soybean
measurements were taken, with seeds gently shaken between readings. seed coats (10, 19, 20). An HPLC protocol by Lee and colleagues (19) was
Oil Extraction. Whole soybeans were ground in a standard household adapted for use on a Shimadzu Prominence UFLC system (Columbia,
coffee grinder to a 20 mesh particle size. Five grams of ground soybeans MD) with an autosampler, an in-line degasser, a CTO-20AC oven, and an
was extracted at room temperature in 10 mL of petroleum ether (BP 35-60 °C) SPD-20A UV-vis detector. A gradient flow of 1.0 mL/min was employed
a total of four times. The petroleum ether was evaporated in a nitrogen on a Phenomenex Gemini C18 column (150 mm 4.6 mm5.0 μm) in an
evaporator to a constant weight. The oils were stored at ambient oven set to 25 °C. Solvent A consisted of 0.1% acetic acid in water, whereas
temperature under nitrogen in the dark until testing. solvent B consisted of 0.1% acetic acid in acetonitrile. The gradient
Antioxidant Extraction. The soy flour remaining after oil extraction changed linearly from 100:0 (A:B, v/v) to 85:15 (A:B, v/v) over 10 min
was then extracted with 50% acetone at a ratio of 1 g/10 mL. The solvent and then to 75:25 (A:B, v/v) over another 10 min. Then, a 5 min wash at
was chosen according to the results of Xu and Chang (13) and to coincide 10:90 (A:B, v/v) was performed, after which the column was re-equili-
with our previous work on soybeans (14). The tubes were stirred to disturb brated at 100:0 (A:B, v/v) for 5 min. Absorbance was read at 530 nm, and
the unwetted portion at the bottom, blown off with nitrogen, vortexed the injection volume was 20 μL of unhydrolyzed extract. Cy-3-glc was
three times for 15 s each, and allowed to extract overnight in the dark. identified and quantified via comparison to peak area of an external
The soy flour-solvent mixtures were then sonicated for 15 min, and the standard.
extracts were filtered through 0.45 μm syringe filters. The extracts were Antioxidant Activity Assays. ABTS•þ Scavenging Ability. The
held under nitrogen in the dark until testing. scavenging capacity against ABTS•þ was measured according to a
Fatty Acid Composition. Fatty acid methyl esters (FAME) were previously reported method (21). Briefly, ABTS•þ working solution was
prepared from the oils by saponification followed by methylation accord- prepared by reacting 2,2-azobis(3-ethylbenzothizide-6-sulfonic acid) dia-
ing to a previously described laboratory protocol and subjected to gas mmonium salt with manganese oxide (MnO2) in solution, which was then
chromatography (GC) analysis (15). A Shimadzu GC-2010 with an FID, filtered and diluted to an absorbance of 0.700 ( 0.005 at 734 nm on a
an AOC-20i injector, and an AOC-20S autosampler (Shimadzu, Columbia, Thermo Spectronic Genesys spectrophotometer. Eighty microliters of
MD) was used for fatty acid analysis. The carrier gas was helium at a total sample or standard was added to 1 mL of the working ABTS•þ solution
flow rate of 12.2 mL/min through a fused silica capillary column SP-2380 and vortexed for 30 s, and the absorbance was read at 734 nm after 90 s of
(30 m  0.25 mm with a 0.25 μm film thickness) from Supelco (Bellefonte, reaction. Trolox standards in 50% acetone were used to create a standard
PA). Injection volume was 1 μL at a split ratio of 10/1. Oven temperature curve, and samples were diluted as necessary to fit on the curve.
was initially 136 °C, increased by 6 °C/min to 184 °C, at which it was held Oxygen Radical Absorbance Capacity (ORAC) Assay. The
for 3 min, and then increased again by 6 °C/min to a final temperature of ORAC assay was conducted according to a previously reported laboratory
226 °C. Individual fatty acids were identified through comparison of GC protocol (22) with minor modifications. Trolox standards were prepared
retention times with those of fatty acid methyl ester external standards. in 50% acetone, whereas all other reagents were prepared in 75 mM
11176 J. Agric. Food Chem., Vol. 57, No. 23, 2009 Slavin et al.
Table 1. Isoflavone Composition of Soybean Extract Treatments at Three Table 2. Isoflavone Composition of Isoflavone Treatment Medium Concen-
Concentrations (Micromolar)a trations (Micromolar)a
low dose medium dose high dose Daid Gen Gly Cy
Daid Gen Gly Daid Gen Gly Daid Gen Gly control 0 0 0 0
Daid only 12.0 0 0 0
Emerald 0.5 0.7 0.9 1.6 2.3 3.0 4.8 6.9 8.9
Gen only 0 13.6 0 0
MD 0304 WN-46 0.4 0.4 1.2 1.2 1.5 4.0 3.6 4.5 12.0
Gly only 0 0 25.4 0
MD 0304 WN-46-1 0.2 0.3 0.7 0.6 1.0 2.2 1.7 2.9 6.7
Daid þ Gen 12.0 13.6 0 0
Peking 1.8 1.2 2.1 5.9 4.0 7.0 17.7 12.1 20.9
Daid þ Gly 12.0 0 25.4 0
Pi 88788 1.2 1.4 2.5 4.1 4.5 8.5 12.2 13.6 25.4
Gen þ Gly 0 13.6 25.4 0
Isoflavone mixture 1.2 1.4 2.5 4.0 4.5 8.5 12.0 13.6 25.4
Daid þ Gen þ Gly 12.0 13.6 25.4 0
a
Daid, daidzein; Gen, genistein; Gly, glycitein. Pi 88788 mimic 12.0 13.6 25.4 7.0
Peking mimic 17.7 12.1 20.9 17.7
sodium phosphate buffer (pH 7.4). The initial reaction mixture contained a
225 μL of freshly made 8.16  10-8 M fluorescein and 30 μL of sample, Daid, daidzein; Gen, genistein; Gly, glycitein; Cy, cyanidin.
standard, or blank solution. Initial reaction mixtures were pipetted into a
96-well plate and preheated at 37 °C for 20 min. Then, 25 μL of freshly made
0.36 M AAPH was added to each well. The fluorescence of the assay Table 3. HunterLab Color Values of Soybeansa
mixture was recorded on a Victor3 multilabel plate reader (Perkin-Elmer, L* a* b*
Turku, Finland) once every 2 min for 3 h at 37 °C, with λEx = 485 nm and
λ Em =535 nm. Trolox equivalents (TE) were calculated for samples on Black
the basis of area under the curve (AUC) calculations used by Ou and
others (23). Results are expressed as micromoles of TE per gram of MD 06-5437-1 19.6 ( 0.3 0.2 ( 0.2 1.0 ( 0.2
defatted soybean flour. MD 06-5440-2 16.5 ( 0.44 0.1 ( 0.2 -0.5 ( 0.1
Hydroxyl Radical Scavenging Capacity (HOSC) Estima- Peking 22.7 ( 0.3 0.0 ( 0.0 0.3 ( 0.0
tion. The HOSC assay was conducted also using fluorescein as the Pi 88788 22.3 ( 0.2 0.0 ( 0.1 0.0 ( 0.1
fluorescent probe and a Victor3 multilabel plate reader (Perkin-Elmer) Pi 90763 19.3 ( 0.6 0.0 ( 0.1 -0.6 ( 0.1
according to a previously reported laboratory protocol (24). The reaction
Brown
mixture contained 170 μL of 9.28  10-8 M fluorescein, 30 μL of sample,
40 μL of 0.1990 M H2O2, and 60 μL of 3.43 M FeCl3. The fluorescence of MD 0304 WN46-1 33.9 ( 0.2 11.6 ( 0.2 14.5 ( 0.1
the reaction mixture was recorded approximately once every 4 min for 7 h MD 0304 WN46-2 33.3 ( 0.4 11.4 ( 0.3 14.4 ( 0.6
at ambient temperature, with λEx=485 nm and λ Em=535 nm. Standards MD 0304 WN46-3 33.6 ( 0.5 11.4 ( 0.3 13.9 ( 0.6
were prepared in 50% acetone. The 9.28  10-8 M fluorescein was pre- MD 0304 WN46-4 32.7 ( 1.0 10.9 ( 0.6 13.5 ( 0.7
pared fresh for each assay from stock solution and 75 mM sodium
phosphate buffer (pH 7.4). Trolox equivalents were calculated for samples Green
using the same AUC calculations as in ORAC (24). Results are expressed Emerald 49.6 ( 1.5 -3.4 ( 0.2 24.6 ( 0.6
as micromoles of TE per gram of defatted soybean flour. Verde 51.4 ( 1.1 -3.4 ( 0.5 24.9 ( 0.3
Antiproliferative Activity against HT-29 Cells. The antiprolifera-
tion test was adopted from Wang et al. (25). HT-29 human colorectal Yellow
adenocarcinoma cells were cultured in a humidified incubator at 37 °C and
MD 0304 WN-46 61.1 ( 0.7 7.9 ( 0.4 30.8 ( 0.5
5% CO2. Cell culture media consisted of McCoy’s 5A media supplemented
MD 05-6073 63.7 ( 0.5 6.7 ( 0.1 29.3 ( 0.2
with 10% fetal bovine serum and 1% antibiotic/antimycotic solution.
MD 05-6077 62.4 ( 0.4 6.9 ( 0.3 30.6 ( 1.1
Cells were grown to 95% confluence and then plated at 2500 cells/well
MD 05-6079 62.4 ( 0.5 6.5 ( 0.2 29.7 ( 0.4
in a black 96-well viewplate treated for tissue culture. After 24 h, the
MD 06-5433-1 60.7 ( 1.5 7.2 ( 0.2 34.6 ( 1.8
medium was replaced with 100 μL of control or treatment medium.
MD 06-5433-3 60.7 ( 1.5 6.6 ( 0.1 30.7 ( 0.5
Proliferation of cells was assessed using the ATP-Lite 1step Luminescence
MD 06-5445-5 58.3 ( 1.3 6.0 ( 0.1 30.9 ( 1.0
Assay System from Perkin-Elmer prior to treatment and at 3, 24, 48, 72,
a
and 96 h after initial treatment. To take a reading, plates were allowed to Values based on triplicate readings. Mean ( standard deviation shown. (n = 3).
equilibrate at room temperature for 30 min, and then 100 μL of
reconstituted ATPlite solution was added to each well immediately prior daidzein, and glycitein (each g99% pure and purchased from Sigma-
to taking luminescence readings on a Victor3 multiwell plate reader Aldrich or Indofine), and the concentration of isoflavones in the mixture
(Perkin-Elmer). A separate plate was used for each reading. Treatment was equivalent to that of Pi 88788, which had the highest isoflavone
and control media were replaced every 24 h until a reading was taken on contents of soybean extracts tested on cells.
that plate. For the study of the antiproliferative activity of isoflavones, cells were
Treatment Media Preparation. All treatment media contained treated with the aglycone isoflavones (genistein, daidzein, and glycitein)
1.2% DMSO, including the control, and were filtered through a 0.2 μm individually and in all possible combinations at three doses each. Again,
retrograde cellulose syringe filter prior to treatment of cells. the final concentration of DMSO in the treatment media was 1.2%. Doses
For the study of the antiproliferative activity of soybean extracts, cells were chosen to be concentrations equivalent to the high dose of soybean Pi
were treated with hydrolyzed or unhydrolyzed soybean extracts in three 88788. Actual doses and combinations can be seen in Table 2.
doses: 1.5, 5, and 15 mg of seed flour equiv/mL of media. All treatment Another study was conducted to test the antiproliferative effect of an
media, including the control, contained a final concentration of 1.2% isoflavone and cyanidin mixture that mimic the concentration of both
DMSO (v/v). Isoflavone compositions of the hydrolyzed extracts were isoflavones and cyanidin in Peking and Pi 88788 extracts. The isoflavone
calculated for genistein, daidzein, and glycitein on the basis of HPLC data and cyanidin concentrations of these two soybeans were recreated with
and are shown in Table 1. Because the unhydrolyzed extracts had not had standard compounds, with final treatment media again containing 1.2%
their glucoside isoflavones hydrolyzed to their aglycone forms, it is likely DMSO, and antiproliferative effects were again compared to a control
their isoflavones were mainly in the glucoside form. However, all 12 with 1.2% DMSO. Actual concentrations are shown in Table 2.
isoflavone forms were not quantified in unhydrolyzed extracts. For the Statistical Analysis. Tests were conducted in triplicate with data
unhydrolyzed extracts, the values in Table 1 therefore represent isoflavone reported as mean ( standard deviation. Differences in means were
“aglycone equivalents”. For comparison purposes, a standard isoflavone detected using one-way ANOVA and Tukey’s test. For those instances
treatment was tested on the cells. It included a mixture of genistein, when average values were calculated for a seed coat color group,
Article J. Agric. Food Chem., Vol. 57, No. 23, 2009 11177
a
Table 4. Fatty Acid Composition of Soybean Oil by Seed Coat Color (Grams per 100 g of Oil)
16:0 18:0 total SFA 18:1 18:2 18:3 total PUFA

Black

MD 06-5437-1 11.27i ( 0.05 3.91d ( 0.03 15.18k ( 0.03 21.26c ( 0.07 52.35ij ( 0.12 8.48k ( 0.06 60.94j ( 0.17
MD 06-5440-2 11.44j ( 0.05 4.05e ( 0.01 15.49m ( 0.05 22.99d ( 0.04 50.63 g ( 0.05 7.99i ( 0.01 58.66 h ( 0.04
Peking 11.21i ( 0.08 3.46b ( 0.05 14.67i ( 0.04 17.13a ( 0.04 55.14p ( 0.12 10.78p ( 0.07 66.03o ( 0.18
Pi 88788 11.41j ( 0.02 3.48b ( 0.03 14.89j ( 0.02 18.72b ( 0.05 53.57 m ( 0.06 10.27n ( 0.03 63.89m ( 0.05
Pi 90763 11.64k ( 0.02 3.03a ( 0.03 14.67i ( 0.01 18.16b ( 0.03 53.93no ( 0.11 10.45o ( 0.02 64.49n ( 0.11
black av 11.39z ( 0.16 3.59yz ( 0.38 14.98z ( 0.33 19.65w ( 2.23 53.12z ( 1.59 9.59z ( 1.17 62.71z ( 2.73
Brown

MD 0304 WN46-1 5.86b ( 0.02 4.24gh ( 0.05 10.10c ( 0.05 41.48j ( 0.04 42.84bc ( 0.08 3.10a ( 0.05 46.02b ( 0.11
MD 0304 WN46-2 6.19d ( 0.02 4.11ef ( 0.03 10.30d ( 0.03 41.02j ( 0.04 43.13d ( 0.03 3.31bc ( 0.07 46.47c ( 0.05
MD 0304 WN46-3 6.05c ( 0.01 4.19fg ( 0.03 10.24cd ( 0.04 41.01j ( 0.08 42.92cd ( 0.09 3.27bc ( 0.05 46.28bc ( 0.13
MD 0304 WN46-4 6.14 cd ( 0.06 4.36h ( 0.03 10.49e ( 0.08 41.07j ( 0.10 42.63b ( 0.07 3.23b ( 0.01 45.93b ( 0.07
brown av 6.06x ( 0.13 4.22z ( 0.10 10.28y ( 0.15 41.15z ( 0.21 42.88y ( 0.20 3.23w ( 0.09 46.11x ( 0.25
Green

Emerald 9.58e ( 0.07 3.45b ( 0.03 13.03f ( 0.05 23.51de ( 0.11 52.53jk ( 0.09 8.33j ( 0.03 60.96j ( 0.12
Verde 10.97h ( 0.08 3.51bc ( 0.03 14.48h ( 0.08 26.08f ( 0.13 48.88f ( 0.19 7.76h ( 0.04 56.83f ( 0.16
green av 10.28y ( 0.76 3.48y ( 0.04 13.76yz ( 0.73 24.80x ( 1.41 50.71z ( 2.01 8.05y ( 0.31 58.75y ( 2.31
Yellow

MD 0304 WN-46 6.04c ( 0.04 3.86d ( 0.01 9.90b ( 0.04 36.03i ( 0.06 48.31e ( 0.18 3.37c ( 0.01 51.87d ( 0.18
MD 05-6073 3.54a ( 0.04 3.61c ( 0.02 7.15a ( 0.05 33.26g ( 0.02 53.68mn ( 0.05 3.76e ( 0.92 57.50g ( 0.07
MD 05-6077 3.55a ( 0.2 3.61c ( 0.02 7.16a ( 0.02 33.11g ( 0.07 53.99o ( 0.05 3.52d ( 0.01 57.56g ( 0.05
MD 05-6079 3.59a ( 0.03 3.49bc ( 0.01 7.08a ( 0.02 34.79h ( 1.11 51.36h ( 0.03 3.88f ( 0.01 55.27e ( 0.02
MD 06-5433-1 16.60m ( 0.04 6.08i ( 0.02 22.68n ( 0.04 23.86e ( 0.04 38.98a ( 0.05 5.92g ( 0.01 44.95a ( 0.04
MD 06-5433-3 10.61g ( 0.03 3.88d ( 0.11 14.50h ( 0.10 22.71d ( 0.04 52.23i ( 0.07 8.04i ( 0.03 60.34i ( 0.07
MD 06-5445-5 10.36f ( 0.02 3.48b ( 0.07 13.84g ( 0.05 21.71c ( 0.02 52.73k ( 0.04 9.07m ( 0.05 61.84k ( 0.08
yellow av 7.76xy ( 4.72 4.00yz ( 0.88 11.76yz ( 5.46 29.35y ( 5.96 50.18z ( 5.02 5.37x ( 2.24 55.55y ( 5.44
a
Values are based on triplicate readings. Mean ( SD shown. Individual sample values in the same column marked by the same letter are not significantly different (P e 0.05).
Average values in the same column marked by the same letter are not significantly different (P e 0.05). SFA, saturated fatty acids; PUFA, polyunsaturated fatty acids.

differences between color groups were determined using one-way ANOVA in their fatty acid compositions (P e 0.05). This is relevant due to
with the Tukey-b post hoc option, allowing for the comparison of unequal the considerable efforts dedicated to altering the fatty acid
sample sizes. Correlations among means were determined using a two- composition of soybeans oils for specific uses: low-linolenic
tailed Pearson correlation test. Statistics were analyzed using SPSS for varieties have been sought due to their improved oxidative
Windows (version rel. 10.0.5, 1999, SPSS Inc., Chicago, IL). Statistical stability, whereas other varieties (including low-saturated or
significance was defined at P e 0.05.
high-linolenic) have been sought for their health-promoting
attributes.
RESULTS AND DISCUSSION The black genotypes had the lowest 18:1 contents and the
Seed Coat Color. In the typical yellow soybean, chlorophyll highest average 16:0 level, along with relatively high levels of total
disappears as seeds mature and dry in the field, allowing the saturated fatty acids (SFA) around 15% of total fatty acids and
remaining flavonoid pigments’ color to show. Some soybeans do on average the highest 18:2 and 18:3 contents (Table 4). The
not lose their chlorophyll as they dry and remain green at brown genotypes had the highest 18:1 content, ranging from 41.0
maturation, whereas others have additional pigments, including to 41.5% of total fatty acids in their oils, along with around 10%
anthocyanins and proanthocyanins, that impart deep black and SFA and low 18:3 contents at 3%, with higher 18:1 and 18:2 to
brown seed coat colors. It is these pigments that are believed to compensate. The two green genotypes had SFA at 13 and 14.5%
contribute at least in part to the antioxidant capacity differences and an average 18:3 content of 8%. This 18:3 level is comparable
discussed later. to that of 8.0-10.5% detected in the soybeans with black coats
Seed coat color values are shown in Table 3. The L* reading is (Table 4). Finally, the most variation in fatty acid composition
an indicator of blackness of the sample, with closer to zero being was seen among the yellow genotypes, with SFA ranging from 7
more black. Appropriately, the black seed coat L* values are in to 23%, 18:1 from 22 to 35%, 18:2 from 39 to 54%, and 18:3 from
the vicinity of 20, whereas the others are higher: brown around 33, 3 to 9%. In summary, black-coated soybeans may serve as a
green around 50, and yellow around 60. Negative a* readings dietary source for n-3 fatty acid, whereas soybeans with brown
indicate a green color, whereas positive a* readings indicate a red coats may be candidates for future breeding efforts to develop
color. As expected, the green seed coats have the most negative a* high-oleic and low-linolenic and saturated fatty acids lines.
value, and the brown have the highest positive a* value. Finally, Carotenoid and r-Tocopherol Contents. Lutein, β-carotene,
b* represents the yellow to blue continuum, with positive values cryptoxanthin, and zeaxanthin contents of the soybean oils are
indicating yellow and negative values indicating blue. The yellow summarized in Table 5. Lutein was the primary carotenoid
seed coats have the highest positive b*, whereas the black seed compound present in the soybeans regardless of seed coat color,
coats have the lowest value. and zeaxanthin was also detected in most of the tested soybean
Fatty Acid Composition. As shown in Table 4, oils from oils. The greatest lutein content was 907 μg/g in the oil of MD
soybeans with different seed coat colors might differ significantly 05-6079 yellow soybean, and MD 05-6077 yellow soybean oil had
11178 J. Agric. Food Chem., Vol. 57, No. 23, 2009 Slavin et al.
a
Table 5. Carotenoid and R-Tocopherol Content of Soybeans by Seed Coat Color
μg/g of oil
R-tocopherol lutein β-carotene cryptoxanthin zeaxanthin total carotenoids (μmol/g of oil)

Black

MD 06-5437-1 526.6hi ( 39.6 171.8b ( 3.7 30.4a ( 3.3 6.9f ( 0.6 95.3e ( 2.1 0.539fg ( 0.011
MD 06-5440-2 1083.8m ( 97.0 191.8b ( 4.4 ND ND 60.0d ( 3.2 0.443ef ( 0.013
Peking 144.1ab ( 5.3 270.7d ( 16.4 ND ND 62.5d ( 3.6 0.586g ( 0.034
Pi 88788 744.1jk ( 75.7 375.9e ( 8.2 ND 1.9cd ( 0.4 119.0f ( 1.4 0.874h ( 0.013
Pi 90763 360.8fg ( 17.3 510.3f ( 17.7 49.1b ( 1.8 6.4f ( 0.4 217.4g ( 10.4 1.382i ( 0.037
Brown

MD 0304 WN46-1 282.7def ( 6.5 201.4bc ( 8.6 ND ND 45.0c ( 1.7 0.433e ( 0.013
MD 0304 WN46-2 262.1cdef ( 6.5 391.9e ( 12.8 ND 3.1e ( 0.4 89.7e ( 2.5 0.852h ( 0.024
MD 0304 WN46-3 308.9ef ( 1.7 64.8a ( 2.7 ND ND 10.6a ( 0.9 0.133b ( 0.005
MD 0304 WN46-4 240.7bcde ( 10.5 178.4b ( 9.2 ND ND 37.9c ( 1.4 0.380de ( 0.016
Green

Emerald 250.0cde ( 6.7 ND ND ND ND 0.000a ( 0.000

Verde 431.3gh ( 26.7 82.4a ( 6.7 ND ND 14.4a ( 0.9 0.170b ( 0.013

Yellow

MD 0304 WN-46 332.7efg ( 15.8 226.9bcd ( 17.0 ND 0.5ab ( 0.3 27.0b ( 1.0 0.447ef ( 0.032
MD 05-6073 161.8abc ( 6.3 80.4a ( 4.7 ND ND ND 0.141b ( 0.008
MD 05-6077 109.9a ( 7.5 ND ND 2.9de ( 0.4 ND 0.005a ( 0.001
MD 05-6079 201.3abcd ( 14.4 906.5g ( 64.0 ND 1.3bc ( 0.2 14.4a ( 1.1 1.621j ( 0.111
MD 06-5433-1 618.7ij ( 23.2 66.9a ( 5.9 36.4a ( 3.4 ND 19.3ab ( 1.2 0.219bc ( 0.015
MD 06-5433-3 677.8jk ( 24.7 99.5a ( 3.7 ND 2.3de ( 0.3 60.6d ( 1.4 0.286cd ( 0.009
MD 06-5445-5 121.0a ( 8.4 256.9cd ( 6.9 ND 2.4de ( 0.2 86.8e ( 2.7 0.608g ( 0.016
a
Values based on triplicate readings. Mean ( SD shown. Values in the same column marked by the same letter are not significantly different (P e 0.05). ND, not detected.

no detectable lutein or zeaxanthin. The highest zeaxanthin con- study were low in these components. The predominant carote-
tent was 217 μg/g in the oil of Pi 90763 black soybean oil, and noid detected was lutein, and consumption of lutein has protec-
black soybean oils contained, on average, higher zeaxanthin than tive effects against a variety of ocular diseases. Tocopherols are
soybeans with other color coats (Table 5). The oils of Pi 90763 and members of the vitamin E family. Both carotenoids and toco-
MD 06-5437-1 black soybean lines had all four carotenoid pherols are lipophilic antioxidants and may have protective
compounds, but the oil of MD 05-6079 yellow soybean line had effects against oxidative damage for oils during storage and
the highest total carotenoid content of 1.62 μmol/g of oil, cellular components after ingestion/absorption.
followed by 1.38 μmol/g of oil detected in the oil of Pi 90763 Isoflavone Composition. Isoflavone compositions of these soy-
soybeans with black seed coats (Table 5). The average total beans with different seed coat colors were compared using the
carotenoid content was about 0.76, 0.48, 0.45, and 0.09 μmol/g hydrolyzed 50% acetone extracts of the defatted soy. Genistein,
oil for soybeans with black, yellow, brown, and green coats, daidzein, and glycitein were detected in all soybean flours, with
whereas the yellow genotypes contained the widest variation in glycitein as the predominating isoflavone in each sample
carotenoid content from 0.005 to 1.621 μmol/g of oil. It was also (Table 6). The black soybeans Peking, Pi 88788, and Pi 90763
noted that the oil of Emerald green soybean contained no had significantly higher total isoflavones among all of the
detectable carotenoids, and the other green-coated soybean soybeans tested (P e 0.05). Soybeans with a black seed coat
cultivar (Verde) contained <0.2 μmol of total carotenoids/g in had an average level of total isoflavone content of 2.1 μmol/g
its oil. However, two samples are not enough to determine of whole soybean, which was much higher than the 1.0, 0.8, and
whether a significant trend exists. 0.7 μmol/g of whole bean observed in the green, brown, and
The oils from soybeans with different color coats were also yellow soybeans, respectively (Table 6). This observation was
evaluated for their R-tocopherol contents (Table 5). All oils supported by a trend seen in two previous papers using 50%
contained R-tocopherol, with the low value from the yellow acetone as the extraction solvent, in which total flavonoid
MD 05-6077 at 109.9 μg/g of oil and the highest value from the contents were 3-4-fold higher in black soybeans than in yel-
black MD 06-5440-2 at 1083.8 μg/g of oil. No specific trend was low (5, 13). However, two other previous studies reported black
apparent between R-tocopherol content and seed coat color, soybeans to contain fewer isoflavones than yellow soybeans,
although the average levels of R-tocopherol were 572, 341, 318, although these studies used a different solvent system and
and 274 μg/g of oil for black, green, yellow, and brown soybean detection method (acetonitrile/water and HPLC), which limits
genotypes. Soybean oil with higher carotenoid content did not their relevance for direct comparison to this study (10, 26). It is
necessarily contain higher R-tocopherol. These data indicated recognized that the isoflavone values of this study may be lower
that soybeans with different coat colors generally contain com- than ranges of values presented in other analyses, a difference
parable levels of carotenoids and R-tocopherol, with higher believed to be mainly due to the extraction conditions.
possibility to find a black soybean line or cultivar rich in these It was also noted that soybeans with the same seed coat color
beneficial components. may differ significantly (P e 0.05) in their total isoflavone
Some carotenoids (including β-carotene and cryptoxanthin) contents and isoflavone compositions (Table 6). For instance,
are forms of provitamin A, although the soybeans in the present Pi 88788 and Peking black soybean samples had significantly
Article J. Agric. Food Chem., Vol. 57, No. 23, 2009 11179
a
Table 6. Isoflavones and Cyanidin-3-glucoside Detected in Soybeans by Seed Coat Color
μmol/g of flour
daidzein genistein glycitein total isoflavones Cy-3-Glc (mg/g of flour)

Black

MD 06-5437-1 0.18bc ( 0.00 0.16ab ( 0.00 0.47def ( 0.00 0.80 cd ( 0.00 0.03a ( 0.01
MD 06-5440-2 0.30f ( 0.00 0.32c ( 0.00 0.53gh ( 0.01 1.15g ( 0.01 0.25c ( 0.02
Peking 1.18i ( 0.03 0.81e ( 0.08 1.39k ( 0.03 3.38j ( 0.12 0.57e ( 0.02
Pi 88788 0.81g ( 0.01 0.91f ( 0.01 1.69m ( 0.02 3.41j ( 0.03 0.22b ( 0.01
Pi 90763 0.94 h ( 0.01 0.78e ( 0.00 1.41k ( 0.01 3.13i ( 0.01 0.30d ( 0.02
Brown

MD 0304 WN46-1 0.11a ( 0.00 0.20b ( 0.01 0.45de ( 0.01 0.76bc ( 0.02 ND
MD 0304 WN46-2 0.19bcd ( 0.00 0.28c ( 0.00 0.52hi ( 0.00 0.98f ( 0.01 ND
MD 0304 WN46-3 0.19 cd ( 0.00 0.28c ( 0.01 0.48efg ( 0.01 0.95ef ( 0.02 ND
MD 0304 WN46-4 0.19 cd ( 0.00 0.27c ( 0.00 0.44cde ( 0.01 0.90ef ( 0.01 ND
Green

Emerald 0.32f ( 0.00 0.46d ( 0.02 0.59i ( 0.01 1.37h ( 0.01 ND

Verde 0.26e ( 0.01 0.31c ( 0.01 0.38ab ( 0.01 0.95ef ( 0.03 ND

Yellow

MD 0304 WN-46 0.24e ( 0.00 0.30c ( 0.00 0.80j ( 0.01 1.34h ( 0.01 ND
MD 05-6073 0.12a ( 0.01 0.13a ( 0.00 0.34a ( 0.05 0.60a ( 0.05 ND
MD 05-6077 0.12a ( 0.00 0.17ab ( 0.00 0.36ab ( 0.00 0.65ab ( 0.00 ND
MD 05-6079 0.19cd ( 0.00 0.20b ( 0.00 0.50fgh ( 0.01 0.89ef ( 0.01 ND
MD 06-5433-1 0.22d ( 0.00 0.21b ( 0.00 0.50efg ( 0.01 0.93de ( 0.01 ND
MD 06-5433-3 0.17b ( 0.00 0.17ab ( 0.00 0.41abc ( 0.01 0.75bc ( 0.00 ND
MD 06-5445-5 0.17b ( 0.00 0.20ab ( 0.00 0.42bcd ( 0.00 0.79bcd ( 0.00 ND
a
Values based on triplicate readings. Mean ( SD shown. Values in the same column marked by the same letter are not significantly different (P e 0.05). ND, not detected.

higher total isoflavone content than the other three black soy- attributions of anti-inflammation and antitoxic activities (9). In
beans (P e 0.05). Peking black soybeans had about the same various studies, anthocyanins of black soybean seed coats were
amounts of glycitein and daidzein, whereas Pi 88788 black able to suppress lipid peroxidation in human low-density lipo-
soybeans had twice as much glycitein as daidzein, although total proteins (LDL), inhibit the proliferation of HT-29 human colon
isoflavone contents did not differ between the two (Table 6). cancer cells, and reduce cyclooxygenase-2 (COX-2) mRNA level
Although overall trends may exist in isoflavones based on color, it and the number of aberrant crypt foci numbers in F344 rats (7,9).
cannot be assumed that any one color contains a specific level. Black soybean anthocyanins have also been shown to mediate the
Consequently, the genotype of soybean for use in a high-isoflavone damage incurred by ischemic myocardial infarction in rats (27)
food must still be evaluated individually and not solely by seed and by ultraviolet B radiation in both human and mice skin
coat color. cells (28). Various components of black seed coat soybeans have
These data indicated that black soybeans may serve as an been implicated as having antiobesity properties, including pep-
excellent dietary source of isoflavones, which have been shown to tides (8,29) and anthocyanins (30). Taken together, identification
slow or inhibit the development of a variety of chronic diseases and/or development of black soybean lines rich in anthocyanins
(cardiovascular disease, various cancers, among others), in addi- may well be an approach to enhance their farm gate value through
tion to their ability to lessen the symptoms of menopause. use in health beneficial functional foods.
Because of this, black soybeans may serve as a valuable ingredient Total Phenolic Content and Antioxidant Activities. It is prudent
in functional foods targeting these diseases and life stage. to perform a variety of antioxidant activity tests on samples to
Cyanidin-3-Glucoside Contents. Recently, nine anthocyanins gain a broad perspective of the antioxidant properties of the
were characterized in the seed coat of black soybeans, and Cy-3- samples. TPC, ORAC, HOSC, and ABTS radical cation scaveng-
glc was shown to be the predominant anthocyanin compound in ing capacity were chosen for this study because of their activities
black soybean seed coats and was detected in whole black as discussed here: The TPC assay is used to compare the reducing
soybeans (10, 19, 20). In the present study, all soybean samples power of a sample compared to a standard phenolic, in this case
were examined for their Cy-3-glc contents, which were detected in gallic acid. ORAC is perhaps the most commonly used antioxi-
all soybeans with black seed coats, but not in other colors. As dant activity assay and utilizes competitive kinetics between
shown in Table 6, Peking contained the highest amount at 0.57 mg/ a fluorescent probe and the sample’s antioxidants to provide
g of flour and was almost double the next nearest sample, Pi 90763. information on a sample’s ability to scavenge peroxyl radicals
Although Peking contained the most Cy-3-glc of the soybeans through a hydrogen atom transfer. The HOSC assay is a relatively
tested, Choung and others (20) reported its anthocyanin content to new assay compared to ORAC, but uses a similar method to
be in the middle of a range of 10 black soybeans tested with estimate a sample’s ability to scavenge hydroxyl radicals. ABTS
acidified MeOH as the extraction solvent. Data in Table 6 also radical cation scavenging capacity measures the ability of a
suggested the huge possible variation of anthocyanins among sample to reduce the nonphysiological radical.
black soybean cultivars and lines. All soybeans sampled contained significant levels of phenolics,
Soybeans with black seed coats have long been recognized by and the black seed coat soybeans exhibited much higher levels of
various Eastern cultures to have health beneficial properties, with TPC than soybeans of the other seed coat colors (Figure 1).
11180 J. Agric. Food Chem., Vol. 57, No. 23, 2009 Slavin et al.

Figure 1. Total phenolic contents of colored seed coat soybean extracts. Values are based on triplicate tests and expressed as milligrams of gallic acid
equivalents per gram of soybean flour. Means and SD are shown (n = 3). Genotypes marked by the same letter are not significantly different (P e 0.05).

Figure 2. ABTSþ radical scavenging capacity of colored seed coat soybean extracts. Values are based on triplicate tests and expressed as micromoles of
Trolox equivalents per gram of soybean flour. Means þ SD are shown (n = 3). Genotypes marked by the same letter are not significantly different (P e 0.05).

Peking had the highest TPC value of 12.1 mg of gallic acid other yellow, green, or brown soybeans. The values for TPC,
equivalents (GAE)/g of flour, and MD 06-5437-1 with black seed ORAC, HOSC, and ABTS•þ scavenging capacity were in very
coat had about 2.7 mg of GAE/g of flour, which was the lowest strong agreement;each of these four tests was significantly (P <
among all black soybeans. The soybeans with brown, green, and 0.01), strongly (r > 0.92), and positively correlated to each other
yellow seed coats had values between 0.8 and 2.2 mg of GAE/g of test.
flour, with most values clustered between 1.2 and 1.8 mg of GAE/g These data agreed with previously published studies in both
of flour. approximate values and the magnitude of differences seen between
All soybeans sampled also contained scavenging capacities black soybeans and other color soybeans for both TPC and
against peroxyl (ORAC), hydroxyl (HOSC), and ABTSþ radi- antioxidant activity, measured in other studies by ORAC, DPPH
cals, and black seed coat soybeans exhibited significantly higher radical scavenging capacity, and FRAP assays (5, 10, 13, 26, 31).
antioxidant activities (P e 0.05) than soybeans of the other seed Limited data were available on values of ABTS•þ scavenging
coat colors in all three assays (Figures 2-4). Peking black capacity and HOSC for soybeans. The HOSC assay was chosen in
soybeans, which had the highest TPC, also had the highest this study because it is the only radical scavenging capacity assay
ABTS•þ scavenging capacity of 102 μmol of TE/g of flour, validated for radical purity and concentration consistency, and it
whereas Pi 88788 had the greatest ORAC value of 255 μmol of is also performed under the physiological pH of 7.4 (24).
TE/g of flour and a HOSC value of 238 μmol of TE/g of flour. Soybeans with black seed coats have repeatedly been shown to
MD 0304 WN-46 yellow soybeans had significantly stronger have higher antioxidant contents than those with other seed coat
radical scavenging capacities against all three tested radicals than colors in almost every antioxidant measurement that is commonly
Article J. Agric. Food Chem., Vol. 57, No. 23, 2009 11181

Figure 3. Oxygen radical absorbance capacity of colored seed coat soybean extracts. Values are based on triplicate tests and expressed as micromoles of
Trolox equivalents per gram of soybean flour. Means þ SD are shown (n = 3). Genotypes marked by the same letter are not significantly different (P e 0.05).

Figure 4. Hydroxyl radical scavenging capacity of colored seed coat soybean extracts. Values are based on triplicate tests and expressed as micromoles of
Trolox equivalents per gram of soybean flour. Means þ SD are shown (n = 3). Genotypes marked by the same letter are not significantly different (P e 0.05).

used: TPC, ORAC, DPPH radical scavenging capacity, and isoflavone mixture, at three different concentrations shown in
FRAP (5, 6, 26, 31). Worthy of note, the vast majority of Table 1. The isoflavone mixture contained daidzein, glycitein, and
isoflavones have been shown to be found in the cotyledon and genistein at the same levels that were present in the Pi 88788
germ, not the seed coat (10), and some studies have found fewer soybean extract, which had the highest total isoflavones among
isoflavones in black seed coat soybeans as compared to other all tested soybean samples.
colors (10,26). The difference in antioxidant activity arises at least Figure 5A shows the comparison of cell number after 48 h of
in part from the presence of anthocyanins in the seed coats, which treatment by unhydrolyzed extracts, whose isoflavones are there-
only black seed coats contain (11). They serve as both pigment fore mostly in their original glycoside forms on the same per
and antioxidant. In addition, previous research has shown that soybean flour weight basis. Figure 5B shows cell number after 48 h
brown and black seed coat soybeans contain proanthocya- treatment by the hydrolyzed extracts, which are therefore in their
nins (11), which may contribute to the overall health properties aglycone forms. As seen in Figure 5A, Peking soybean extract at the
of black and brown soybeans, although they were not measured high dose of 15 mg of soy flour equiv/mL is the only unhydrolyzed
in the present study. extract to significantly (P e 0.05) suppress HT-29 cell proliferation
Antiproliferative Activity against HT-29 Cells. In the present after 48 h of treatment, whereas the isoflavone mixture and
study, two black soybean samples (Peking and Pi 88788), one unhydrolyzed Pi 88788 black soybean extract at the high testing
brown soybean (MD 0304 WN 46-1), one green soybean dose did not produce a statistically significant reduction in cell
(Emerald), and one yellow soybean sample (MD 0304 WN-46) number until 72 h of treatment time (data not shown). Among the
were selected because of their stronger antioxidant activities hydrolyzed extracts, the highest treatment level (15 mg of flour
within their same seed coat color groups and investigated for equiv/mL) of black Peking and Pi 88788, yellow MD 0304 WN-46,
their potential antiproliferative effect using HT-29 human colon and brown MD 0304 WN-46-1 soybeans, and the medium treat-
cancer cells. The antiproliferative effects of five soybean extracts ment level (5 mg of flour equiv/mL) of Peking and MD 0304 WN-46
were compared to the control containing only solvent and to the significantly (P e 0.05) reduced cell number at 48 h of treatment
11182 J. Agric. Food Chem., Vol. 57, No. 23, 2009 Slavin et al.

Figure 5. Antiproliferative effects of hydrolyzed and unhydrolyzed soy extracts. Data were collected at 48 h of treatment: (A) treated by glycoside forms of
soybean extracts; (B) treated by aglycone forms of soybean extracts (hydrolyzed). Isoflavone treatment contains genistein, daidzein, and glycitein. Low,
medium, and high doses of soybean extracts represent 1.5, 5, and 15 mg of flour equiv/mL treatment, with isoflavone concentrations of each shown in Table 1.
The isoflavone standard was designed to mimic the concentration of isoflavones in Pi 88788. Relative luminescence is proportionate to the number of viable
cells. Values are based on triplicate tests, with mean þ SD shown (n = 3). Columns marked by an asterisk are significantly different from the control (P e 0.05).

(Figure 5B). Data in Figure 5 suggest that soybeans with different as compared to 3 h by the actual Pi 88788 aglycone extract. Also,
seed coat color may differ in their antiproliferative components. by 72 h the Pi 88788 aglycone extract had essentially eliminated
Data in Figure 5 also suggest that the hydrolyzed, aglycone forms of viable cells to 0% of the vehicle, whereas the isoflavone mixture
soybean extracts might be more effective in suppressing HT-29 reduced cell numbers to only 60% of the vehicle.
colon cancer cell growth than their corresponding glycosides on the Interestingly, the green Emerald soybean had isoflavone levels
same per molar concentration basis. This may be partially explained similar to or greater than those of the yellow MD 0304 WN-46
by the greater cellular availability of the aglycones because of the and brown MD 0304 WN-46-1 soybeans, but had no detectable
reduced water solubility and polarity, rendering them better able to antiproliferative activity at any dose at any time under the
cross the cell membrane and exert effects intracellularly. experimental conditions (Figure 6F), whereas both yellow MD
The hydrolyzed, aglycone form of the Peking, Pi 88788, MD 0304 WN-46 and brown MD 0304 WN-46-1 soybean extracts
0304 WN-46, and MD 0304 WN 46-1 extracts and the soy significantly (P e 0.05) suppressed HT-29 cell growth (Figures 5B
isoflavone mixture suppressed HT-29 cancer cell proliferation and 6D,E). In addition, the MD 0304 WN-46 soybean extract had
in a dose- and time-dependent manner (Figure 6). At the high the lowest total isoflavones, which was approximately 25% of
treatment dose, the aglycone forms of black Peking and Pi 88788, that in the isoflavone mixture, but its aglycone form elicited
brown MD 0304 WN46-1, and yellow MD 0304 WN-46 soybean significant reduction in cell number after 3 h of treatment at both
flour extracts significantly (P e 0.05) reduced cell number after 3 the medium and high doses, whereas the isoflavone mixtures did
h of treatment (Figure 6B-E). The aglycone forms of Peking not significantly (P e 0.05) inhibit cell proliferation until 72 h of
soybean extracts also reduced cell number significantly (P e 0.05) treatment at the high dose. These data suggested that soybean
in the medium dose (5 mg of defatted flour equiv/mL) at 3 h lines and cultivars with the same seed coat color might differ in
of treatment, whereas the only other medium dose to significantly their antiproliferative activities and that other factors in soybeans
(P e 0.05) affect antiproliferative capacity was the yellow MD beyond isoflavones and cyanidin may contribute to the overall
0304 WN-46 after 24 h of treatment (Figure 6B,E). No treatment antiproliferative activity of soybean flour extracts.
displayed any inhibitory effect at the low dose at any time under Finally, the isoflavone mixture was broken down into its
the experimental conditions. individual components to test which may be contributing to the
The isoflavone standard mixture (Figure 6A) did not produce antiproliferative activity. Additionally, cyanidin was introduced
as effective a reduction in cell number as Pi 88788, whose to the mixture to mimic the concentration of the predominant
isoflavone concentration it was designed to mimic (Figure 6C), anthocyanidin in addition to the isoflavones of the two black
nor were its effects as fast. The isoflavone standard mixture soybean extracts present after hydrolysis. Concentrations of
produced a statistically significant reduction in cell number by 72 h, treatment media are shown in Table 2. Results at 96 h are
Article J. Agric. Food Chem., Vol. 57, No. 23, 2009 11183

Figure 6. Time and dose effects of soy isoflavones and hydrolyzed soybean extracts on HT-29 cell proliferation:(A) soy isoflavones consisting of genistein,
daidzein, and glycitein; (B) black Peking; (C) black Pi 88788; (D) brown MD 0304 WN46-1; (E) yellow MD 0304 WN-46; and (F) green Emerald extracts were
hydrolyzed with base prior to cell treatment to convert all isoflavones to aglycone form. Relative luminescence is proportionate to the number of viable cells.
Values are based on triplicate tests, with mean values shown (n = 3). Low, medium, and high doses of soybean extracts represent 1.5, 5, and 15 mg of flour
equiv/mL treatment, with isoflavone concentrations of each shown in Table 1. The isoflavone standard was designed to mimic the concentration of isoflavones
in Pi 88788.

presented in Figure 7. Every combination of two or three range of isoflavone type and dose, in combination with cyanidin,
isoflavones produced significant reduction in HT-29 cell prolif- may elicit the same ultimate response by cells.
eration at 96 h. Glycitein was the only individual isoflavone to Previous work has suggested that isoflavones and anthocya-
produce significant cell reduction in 96 h. This is likely because the nidins independently inhibit cancer cell growth in their aglycone
glycitein concentration was approximately twice that of genistein form in in vitro studies (32-34). Taking into account the data in
or daidzein in the treatments (designed to mimic the concentra- Figures 5-7, the present study suggests that soybeans contain
tions seen in Pi 88788). However, it may be important to note that antiproliferative components, although their antiproliferative
although all of the combination treatments were successful at activities may not be fully explained by their isoflavone content,
inhibiting cell proliferation versus the control, they were not cyanidin content, and seed coat color.
significantly different from each other (P e 0.05). In other words, This study is fundamentally limited in its analysis of only one
the glycitein treatment contained less total isoflavones than any of crop year. It is well-accepted that environmental conditions
the combination treatments, but it was similarly effective at (temperature, solar radiation levels, soil composition, water
reducing cell proliferation. A difference was seen in the time availability, etc.) affect the nutritional and phytochemical com-
required to significantly (P e 0.05) reduce cell proliferation;the position of crops, including soybeans. For example, soy isofla-
Peking and Pi 88788 mimics with isoflavones and cyanidin elicited vone levels have been shown to be depressed during growth at
a significant inhibition of cells by 48 h, whereas the mixture of higher temperatures (35). For this reason, additional studies
three isoflavones and glycitein alone exhibited significant inhibi- analyzing multiple crop years may serve to identify if the trends
tion in 72 h. These data suggested that variations within a certain noted here hold true across a variety of growing conditions.
11184 J. Agric. Food Chem., Vol. 57, No. 23, 2009 Slavin et al.
(12) Trostle, R. Fluctuating food commodity prices: a complex issue with
no easy answers. Amber Waves 2008, Nov.
(13) Xu, B. J.; Chang, S. K. C. A comparative study on phenolic profiles
and antioxidant activities of legumes as affected by extraction
solvents. J. Food Sci. 2007, 72, S159–S166.
(14) Slavin, M.; Cheng, Z.; Luther, M.; Kenworthy, W.; Yu, L. Anti-
oxidant properties and phytochemical composition of Maryland-
grown soybean lines with special fatty acid profiles. Food Chem.
2009, 114 (1), 20–27.
(15) Yu, L.; Adams, A.; Watkins, B. Comparison of commercial supple-
ments containing conjugated linoleic acids. J. Food Compos. Anal.
2003, 16, 419–428.
(16) Hao, Z.; Parker, B.; Knapp, M.; Yu, L. L. Simultaneous quantifica-
tion of R-tocopherol and four major carotenoids in botanical
materials by normal phase liquid chromatography-atmospheric
Figure 7. Effects of cell isoflavone treatments at 96 h. Daid, daidzein; Gly, pressure chemical ionization-tandem mass spectrometry. J. Chro-
glycitein; Gen, genistein. Treatment concentrations are shown in Table 2, matogr., A 2005, 1094, 83–90.
with high doses shown here. All treatments, including control, contained (17) Yu, L.; Haley, S.; Perret, J.; Harris, M. Antioxidant properties of
1.2% DMSO. Values are based on triplicate tests, with mean ( SD shown. extracts from hard winter wheat. Food Chem. 2002, 78, 457–461.
Columns marked by an asterisk are significantly different from the control (18) Lee, Y. W.; Kim, J. D.; Zheng, J.; Row, K. H. Comparisons of
(P e 0.05). isoflavones from Korean and Chinese soybean and processed
products. Biochem. Eng. J. 2007, 36, 49–53.
There has been a question if the bioactivity and prevention of (19) Lee, J. H.; Kang, N. S.; Shin, S.-O.; Shin, S.-H.; Lim, S.-G.; Suh,
diseases by soybeans have been falsely attributed to any indivi- D.-Y.; Baek, I.-Y.; Park, K.-Y.; Ha, T. J. Characterisation of
dual phytochemical or component. Data from this study suggest anthocyanins in the black soybean (Glycine max L.) by HPLC-
that whereas black soybeans may have higher levels of isofla- DAD-ESI/MS analysis. Food Chem. 2009, 112, 226–231.
vones, anthocyanins, and antioxidant activity, their health ben- (20) Choung, M.-G; Baek, I.-Y.; Kang, S.-T.; Han, W.-Y.; Shin, D.-C.;
Moon, H.-P.; Kang, K.-H. Isolation and determination of antho-
eficial property as a whole may not increase proportionately with
cyanins in seed coats of black soybean (Glycine max (L.) Merr.).
these levels. More studies are necessary comparing the effects of J. Agric. Food Chem. 2001, 49, 5848–5851.
isolated soy components with those of the whole bean or bean (21) Moore, J.; Luther, M.; Cheng, Z.; Yu, L. Effects of baking condi-
part. tions, dough fermentation, and bran particle size on antioxidant
properties of whole-wheat pizza crusts. J. Agric. Food Chem. 2009,
LITERATURE CITED 57, 832–839.
(1) McCue, P.; Shetty, K. Health benefits of soy isoflavonoids and (22) Moore, J.; Hao, Z.; Zhou, K.; Luther, M.; Costa, J.; Yu, L.
strategies for enhancement: a review. Crit. Rev. Food Sci. 2004, 44, Carotenoid, tocopherol, phenolic acid, and antioxidant properties
361–367. of Maryland-grown soft wheat. J. Agric. Food Chem. 2005, 53, 6649–
(2) Valachovicova, T.; Slivova, V.; Sliva, D. Cellular and physiological 6657.
effects of soy flavonoids. Mini-Rev. Med. Chem. 2004, 4, 881–887. (23) Ou, B.; Hampsch-Woodill, M.; Prior, R. L. Development and
(3) Dixon, R. A.; Ferreira, D. Genistein. Phytochemistry 2002, 60, 205– validation of an improved oxygen radical absorbance capacity assay
211. using fluorescein as the fluorescent probe. J. Agric. Food Chem. 2001,
(4) Takahashi, R.; Ohmori, R.; Kiyose, C.; Momiyama, Y.; Ohsuzu, F.; 49, 4619–4626.
Kondo, K. Antioxidant activities of black and yellow soybeans (24) Moore, J.; Yin, J.-J.; Yu, L. Novel fluorometric assay for hydroxyl
against low density lipoprotein oxidation. J. Agric. Food Chem. 2005, radical scavenging capacity (HOSC) estimation. J. Agric. Food
53, 4578–4582. Chem. 2006, 54, 617–626.
(5) Xu, B.; Yuan, S. H.; Chang, S. K. C. Comparative analyses of (25) Wang, C. Y.; Wang, S. Y.; Yin, J.-J.; Parry, J.; Yu, L. L. Enhancing
phenolic composition, antioxidant capacity, and color of cool season antioxidant, antiproliferation, and free radical scavenging activities
legumes and other selected food legumes. J. Food Sci. 2007, 72, S167– in strawberries with essential oils. J. Agric. Food Chem. 2007, 55,
S177. 6527–6532.
(6) Takahata, Y.; Ohnishi-Kameyama, M.; Furuta, S.; Takahashi, M.; (26) Xu, B.; Chang, S. K. C. Total phenolics, phenolic acids, isoflavones,
Suda, I. Highly polymerized procyanidins in brown soybean seed and anthocyanins and antioxidant properties of yellow and black
coat with high radical-scavenging activity. J. Agric. Food Chem. soybeans as affected by thermal processing. J. Agric. Food Chem.
2001, 49, 5843–5847. 2008, 56, 7165–7175.
(7) Astadi, I. R.; Astuti, M. A.; Santoso, U.; Nugraheni, P. S. In vitro (27) Kim, H. J.; Tsoy, I.; Park, J. M.; Chung, J. L.; Shin, S. C.; Chang,
antioxidant activity of anthocyanins of black soybean seed coat in K. C. Anthocyanins from soybean seed coat inhibit the expression of
human low density lipoprotein (LDL). Food Chem. 2009, 112, 659– TNF-R-induced genes associated with ischemia/reperfusion in en-
663. dothelial cell by NF-κB-dependent pathway and reduce rat myocar-
(8) Rho, S. J.; Park, S.; Ahn, C. W.; Shin, J. K.; Lee, H. G. Dietetic and dial damages incurred by ischemia and reperfusion in vivo. FEBS
hypocholesterolaemic action of black soy peptide in dietary obese Lett. 2006, 580, 1391–1397.
rats. J. Sci. Food Agric. 2007, 87, 908–913. (28) Tsoyi, K.; Park, H. B.; Kim, Y. M.; Chung, J. I.; Shin, S. C.; Lee,
(9) Kim, J.-M.; Kim, J.-S.; Yoo, H.; Choung, M.-G.; Sung, M.-K. W. S.; Seo, H. G.; Lee, J. H.; Chang, K. C.; Kim, H. J. Anthocyanins
Effects of black soybean [Glycine max (L.) Merr.] seed coats and its from black soybean seed coats inhibit UVB-induced inflammatory
anthocyanidins on colonic inflammation and cell proliferation in cylooxygenase-2 gene expression and PGE production through
vitro and in vivo. J. Agric. Food Chem. 2008, 56, 8427–8433. regulation of the nuclear factor-κB and phosphatidylnositol 3-
(10) Xu, B.; Chang, S. K. C. Antioxidant capacity of seed coat, dehulled kinase/Akt pathway. J. Agric. Food Chem. 2008, 56, 8969–8974.
bean, and whole black soybeans in relation to their distributions of (29) Kim, H. J.; Bae, I. Y.; Ahn, C.-W.; Lee, S.; Lee, H. G. Purification
total phenolics, phenolic acids, anthocyanins, and isoflavones. and identification of adipogenesis inhibitory peptide from black
J. Agric. Food Chem. 2008, 56, 8365–8373. soybean protein hydrolysate. Peptides 2007, 28, 2098–2103.
(11) Todd, J. J.; Vodkin, L. O. Pigmented soybean (Glycine max) seed (30) Kwon, S.-H; Ahn, I.-S; Kim, S.-O; Kong, C.-S.; Chung, H.-Y.; Do,
coats accumulate proanthocyanidins during development. Plant M.-S.; Park, K.-Y. Anti-obesity and hypolipidemic effects of black
Physiol. 1993, 102, 663–670. soybean anthocyanins. J. Med. Food 2007, 10, 552–556.
Article J. Agric. Food Chem., Vol. 57, No. 23, 2009 11185
(31) Furuta, S.; Takahashi, M.; Takahata, Y.; Nishiba, Y.; Oki, T.; content in soybean seeds: changes in isoflavones, sapon-
Masuda, M.; Kobayashi, M.; Suda, I. Radical-scavenging activities ins, and composition of fatty acids at different temperatures
of soybean cultivars with black seed coats. Food Sci. Technol. Res. during seed development. J. Agric. Food Chem. 1995, 43, 1184–
2003, 9, 73–75. 1192.
(32) Zhang, Y.; Vareed, S.; Nair, M. Human tumor growth inhibition by
nontoxic anthocyanidins, the pigments in fruits and vegetables. Life Received for review July 27, 2009. Revised manuscript received
Sci. 2005, 76, 1465–1472. October 15, 2009. Accepted October 23, 2009. This research was
(33) Yu, Z.; Li, W.; Liu, F. Inhibition of proliferation and induction of supported by grants from the Maryland Soybean Board in 2007, 2008,
apoptosis by genistein in colon cancer HT-29 cells. Cancer Lett.
and 2009, grants from the Maryland Grain Producers Utilization Board
2004, 215, 159–166.
(MGPUB) with MGPUB grant proposals of 208198 and 209198, a
(34) Guo, J.; Xiao, B.; Liu, D.; Grant, M.; Zhang, S.; Lai, Y.; Guo, Y.;
Liu, Q. Biphasic effect of daidzein on cell growth of human colon grant from the National Science Foundation with a federal grant number
cancer cells. Food Chem. Toxicol. 2004, 42, 1641–1646. of CBET-0650650, and a grant from JIFSAN. MD lines with reduced
(35) Tsukamoto, C.; Shimada, S.; Igita, K.; Kudou, S.; Kokubun, 16:0 and 18:3 were developed as part of a separate grant funded by the
M.; Okubo, K.; Kitamura, K. Factors affecting isoflavone United Soybean Board and Maryland Soybean Board to W.K.