Food Sci. Technol. Res., 15 (2), 171–178, 2009

Technical paper

Development and Validation of a Microplate-based β-carotene Bleaching Assay and

Comparison of Antioxidant Activity (AOA) in Several Crops Measured by β-carotene

Bleaching, DPPH and ORAC Assays

Ichiho MikaMi1*, Minako YaMaguChi

1, Hiroshi shiNMoTo2 and Tojirou Tsushida

1

1 National Food Research Institute, 2-1-12, Kannondai, Tsukuba, Ibaraki 305-8642, Japan2 Tamagawa University, School of Agriculture, 6-1-1 Tamagawa-gakuen, Machida, Tokyo 194-8267, Japan

Received August 15, 2008; Accepted November 19, 2008

The reproducibility and reliability of the micro-β-carotene-linoleic acid bleaching (BCB) assay have been improved, enabling comparison to the antioxidant activity (AOA) of extracts from eleven kinds of crops evaluated by the oxygen radical absorbance capacity (ORAC) and 1,1-diphenyl-2-picrylhydrazyl (DPPH) assays. All assays were conducted using 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid (Trolox) as a calibrator, and AOAs were expressed as micromoles of Trolox equivalent per gram of weight. The coefficient variation (CV) of the BCB assay using a microplate was 14.40%, which was comparable to that of the ORAC assay (H-ORAC (14.10%), L-ORAC (18.76 %)).

Keywords: AOA, β-carotene bleaching assay, DPPH assay, ORAC assay, Trolox, plate reader, tocopherol

*To whom correspondence should be addressed.

E-mail: ichiho@affrc.go.jp

IntroductionAntioxidants are believed to play an important role in

preventing chronic illnesses such as heart disease, stroke,

cancer, Alzheimer’s disease, rheumatoid arthritis, and cata-

racts. As people have become more health-conscious, re-

search on antioxidants has increased. Recently, reviews of

AOA and measuring assays have been published (Koleva

et al., 2002; Prior et al., 2005; Roginsky and Lissi 2004).

These studies compared several commonly used assays for

measuring AOA, and summarized the reaction mechanisms.

Multiple reaction characteristics and mechanisms, as well as

different phase localizations, were usually involved. Thus, no

single assay accurately reflects all the radical sources or all

the antioxidants in a system (Prior et al., 2005). In addition,

antioxidants may respond differently to different radical or

oxidant sources. Therefore, it is important at the outset to un-

derstand that AOA cannot be measured accurately and quan-

titatively by any simple universal assay (Prior et al., 2005).

Antioxidants can deactivate radicals by two major mech-

anisms: hydrogen atom transfer (HAT), and single electron

transfer (SET) (Huang et al., 2005). Therefore, we used the

oxygen radical absorbance capacity (ORAC) assay (a HAT-

based method) and 1,1-diphenyl-2-picrylhydrazyl (DPPH)

radical-scavenging assay (a SET-based method) to evaluate

the AOA of crop extracts from Ibaraki Prefecture, Japan. We

used the β-carotene bleaching (BCB) assay to evaluate the

AOA of crop extracts and compared it to the AOA measured

by the ORAC and DPPH assays. The BCB assay is based on

the competitive bleaching of β-carotene during the antioxi-

dation of linoleic acid in aqueous emulsion, and is monitored

as decay of absorbance in the visible region (Miller, 1971).

However, this assay is poorly quantified since the AOA

is only given as an inhibition percentage and it is difficult

to obtain repeatable data (Roginsky and Lissi, 2004). We

therefore, for the first time, tried to express the AOA of the

BCB assay as the Trolox equivalent. All three assays were

conducted using the Trolox calibrator (positive control), and

AOA was expressed as micromoles of the Trolox equivalent

per gram of weight. In this study, we also ascertained the re-

producibility and reliability of this BCB assay. Furthermore,

the relation between AOA and solvent difference was also

discussed.

Materials and MethodsChemicals and Apparatus 1,1-Diphenyl-2-picrylhydra-

zyl (DPPH), butylated hydroxyanisole (BHA), 2-morpho-

linoethanesulfonic acid (MES), and fluorescein (sodium salt)

(FL) were obtained from Sigma-Aldrich (St. Louis, MO,

USA). 6-Hydroxy-2,5,7,8 -tetramethyl chroman-2-carbox-

ylic acid (Trolox) was purchased from Merck (Darmstadt,

Germany). Methyl-β-cyclodextrin (MCD) was obtained

from Junsei Chemical Co., Ltd. (Tokyo, Japan). 2,2’-Azobis

(2-aminopropane) dihydro-chloride (AAPH), α-tocopherol,

polyoxyethylene (20) sorbitan mono-palmitate, linoleic

acid, and other reagents were purchased from WAKO Pure

Chemical Industries, Ltd. (Osaka, Japan). A Powerscan HT

Multi-Detection Microplate Reader (Dainippon Pharmaceuti-

cal, Osaka, Japan) was used for the ORAC and BCB assays.

Microplates (96 well, Falcon 3072), purchased from Bec-

ton, Dickinson, and Company (NJ, USA), were used for the

ORAC and BCB assays.

Materials Eleven different crops were purchased from

retail stores in Ibaraki Prefecture, Japan (Table 1). The edible

parts of the crops were lyophilized, powdered using a mill,

and stored at -30℃ until used. The crop powder was used in

this study.

Preparation of crop extracts A freeze-dried crop sample

(100 mg) was added to a 10 mL screw-cap tube with 4 mL

of hexane and shaken overnight by an orbital shaker, centri-

fuged at 3,000 rpm for 10 min, and the supernatant was then

collected. Residues were mixed twice with 2 mL of hexane

and were vortexed for 30 seconds, and the supernatant was

collected by centrifugation at 3,000 rpm for 10 min. These

hexane fractions were dried by nitrogen gas and dissolved

in 0.4 mL of acetone. This solution was then used for the

lipophilic ORAC (L-ORAC) assay after having been ap-

propriately diluted with 7% RMCD solvent (w/v) made in

50% acetone-water mixture (v/v) (Huang et al., 2002a). The

residual was dried by using nitrogen gas. Four milliliters of

acetone/water/acetic acid (70:29.5:0.5, v/v/v; AWA) was then

added, and the solution was shaken overnight by an orbital

shaker. After being centrifuged at 3,000 rpm for 10 min, the

supernatant was used for the hydrophilic ORAC (H-ORAC)

assay. Samples that were directly extracted with AWA (H-

ORAC without extraction with lipophilic solvent; W/O

Hex) were also measured. For the DPPH and BCB assays,

80% EtOH was used for the solvent (Table 3). Addition-

ally, to investigate the effects of solvents, the AOA of ginger

extracts by hexane, AWA (W/O Hex) and AWA (with Hex;

after extracted by hexane) (as used for ORAC assay), were

also prepared for the DPPH and BCB assays, and compared

with those of the ORAC assay (Fig. 3). Hexane fractions

were evaporated to dryness using nitrogen gas, dissolved in

DMSO and used as the lipophilic sample for the DPPH and

BCB assays.

DPPH assay The DPPH assay of the crop extracts was

examined by modifying the method of Oki et al. (2001),

Maeda et al. (2006) and Oki (2008). Briefly, the crop extract

was diluted with solvent (80% EtOH) in six stages, and 300

µL of the diluted crop extract was added to a test tube. The

reaction was initiated by adding 900 µL of 20% EtOH/0.2 M

MES buffer (pH 6.0)/400 µM of DPPH in EtOH (1:1:1(v/v/v);

Mix solution). In the case of using solvents other than 80%

EtOH for the assay, the concentration of the organic solvent

(acetone and DMSO) in the reaction mixture was equal. The

reaction mixture was left to stand for 20 min at room temper-

ature, and its absorbance at 520nm was measured using a UV

spectrophotometer (UV-1800, SHIMADZU, Kyoto, Japan).

DPPH scavenging activity was expressed as µmol of Trolox

equivalent/g of fresh weight using the standard Trolox curve.

Three measurements of duplicate data were used to calculate

mean expression.

Measurement of AOA by the β-carotene-linoleic acid

bleaching (BCB) assay Prevention of the autooxidation of

emulsified linoleic acid in crop extracts was determined by

i. MikaMi et al.

Sample No.

Japaneas name English name Scientific name

1 Yacon Yacon Samallanthus sonchifolius2 Red andesu Potato Solanum tuberosum L．3 Satsumaimo Sweet potato Ipomoea batatas Poir．4 Bottyan-kabotya Pumpkin Cucurbita maxima 5 Ahrusu-melon Melon Cucumis melo L．(var. Earl’s Favorite)6 Kyoho Kyoho grapes Hybrids (Vitis vinifera x Vitis labrusca )7 Molokheiya Mulukhiya Corchorus olitorius L．8 Tamanegi Onion Allium cepa L．9 Shoga Ginger Zingiber officinale Rosc．10 Renkon Lotus root Nelumbo nucifera Gaertn．11 Kokabu Turnip Brassica rapa

Table 1. Crops used as experiment materials, obtained from retail stores in Ibaraki Prefecture, Japan.

172

modifying the method of Tsushida et al. (1994). The assay

was adapted to microplate format for convenience. Briefly, 8

µL of crop extract was added to the microplate (Falcon 3072)

in duplicate. Ten milligrams of linoleic acid, 50 mg of Tween

40, and 0.25 mg of β-carotene dissolved in the chloroform

were then placed in a flask. The chloroform was removed

immediately using nitrogen gas, and 22.5 mL of distilled hot

water (50℃) and 2.5 mL of phosphate buffer at pH 6.8 were

then added to the flask and shaken well. Next, 200 µL of lin-

oleic acid-β-carotene solution was transferred rapidly (200

µL/well) to the reaction plate using a multi-channel pipette.

The microplate was immediately placed in the Powerscan

HT Multi-Detection Microplate Reader maintained at 50℃.

The absorbance at 470 nm was recorded every 1 min for 50

min. Because the reaction was temperature-sensitive, close

temperature control throughout the plate was essential in

this assay; therefore, only the 60 inner wells were used (Fig.

1). In addition, when colored crops were measured, it was

necessary to subtract the absorbance of the sample itself. In

our calculation, we applied the slope factors, as in the DPPH

assay, and measurement of α-glucosidase inhibitory activity

(Oki et al., 2001; Yamaki and Mori, 2006). As indicated in

Fig. 1, multiple dilutions of the same sample needed to be

run, and four samples could be used in duplicate per plate.

Samples without the crop extract and with 250 µmol of Tro-

lox, diluted by solvent in five stages, were used as the control

sample (blank) and the standard substance. Dilution of the

sample extracts was done by solvents used for sample extrac-

tion (80% EtOH, AWA or DMSO). Fig. 2 presents the typical

evaluation of AOA by this micro-BCB assay. The absorben-

cies of the sample and Trolox at 5 min are S5 min and T5 min,

respectively, and the absorbencies of the sample and Trolox

at 50 min are S50 min and T50 min, respectively. The differences

in absorbance (470 nm) at 5 min and 50 min of different con-

centrations of Trolox standards [⊿Trolox (T5 min-T50 min)] are

plotted as a function of concentration (Fig. 2B). The differ-

ences in absorbance (470 nm) at 5 min and 50 min of differ-

ent concentrations of samples [⊿Sample (S5 min-S50 min)] are

also plotted (Fig. 2C). The resultant slope and that of Trolox

were used to calculate the AOA of the BCB assay as Trolox

equivalents. Three or more measurements of duplicate data

were expressed as µmol of Trolox equivalents (TE) per gram

of fresh sample weight.

ORAC assay The ORAC value of the extracts was

mainly evaluated according to the method of Cao et al.

(1995), Huang et al. (2002a), and Prior et al. (2003). In

summary, 81.6 nM of fluorescein sodium salt (Sigma, St.

Louis, MO, USA) solution and 200 mM of AAPH 2,2-azobis

(2-amidinopropane) solution were prepared by the phos-

phate buffer solution (pH 7.0). After suitable dilution with

assay buffer, the sample solution was used for the ORAC

assay. Twenty-five microliters of diluted sample and 150 µL

of fluorescein solution were then applied to the microplate

and set in the Multi-Detection Microplate Reader equipped

with a temperature-controlled incubation chamber and an

injection pump. The temperature of the incubator was set at

37℃. After 10 min incubation, 25 µL of AAPH solution was

added by the injection pump. After mixing 10 sec, fluores-

cence (excitation, 485 nm; emission, 528 nm) was recorded

every 2 min for 1.5 hours by the Powerscan HT Gen5 Soft-

ware. Three or more measurements of duplicate data were

expressed as µmol of Trolox equivalents (TE) per gram of

weight (µmol of TE/g of FW).

Results and DiscussionMicro-BCB assay We modified the previously reported

BCB assay to express the AOA of the BCB assay as the Tro-

lox equivalent using a microplate. The BCB assay was first

reported by Marco (1968) and Miller (1971) as a spray re-

agent of Thin Layer Chromatography (TLC) for antioxidants.

Development and Validation of a Microplate-based β-carotene Bleaching Assay

Fig. 1. Display of the results obtained by the Powerscan HT micro-plate reader (micro-BCB assay).Outer wells (A, H1-12, B1, B12, C1, C12, D1, D12, E1, E12, F1, F12, G1, G12) were filled with 200 µL of water to provide a large thermal mass. Only the inner 60 wells were used for experimen-tal determinations. B2 to B11 correspond to the 80% EtOH and β-carotene: linoleic acid emulsion (see Materials and Methods). Columns 2 and 3 (wells C2 to G2 and C3 to G3) correspond to the mixture of β-carotene: linoleic acid emulsion and Trolox at differ-ent concentrations (C2 to C3, 0.2 nM; D2 to D3, 0.4 nM; E2 to E3, 0.6 nM; F2 to F3, 0.8 nM; G2 to G3, 1 nM). Columns 4, 5, 10 and 11 (wells C4 to G4, C5 to G5, C10 to G10, and C11 to G11) cor-respond to the mixture of β-carotene: linoleic acid emulsion and ginger extracts (C4 to C5, 0.005 µL/well; D4 to D5, 0.01 µL/well; E4 to E5, 0.015 µL/well; F4 to F5, 0.02 µL/well; G4 to G5, 0.025 µL/well; C10 to C11, 0.008 µL/well; D10 to D11, 0.016 µL/well; E10 to E11, 0.024 µL/well; F10 to F11, 0.032 µL/well; G10 to G11, 0.04 µL/well). C6 to G6 and C7 to G7 correspond to the mixture of β-carotene: linoleic acid emulsion and extracts from lotus root (C6 to C7, 0.016 µL/well; D6 to D7, 0.32 µL/well; E6 to E7, 0.48 µL/well; F6 to F7, 0.64 µL/well; G6 to G7, 0.8 µL/well). C8 to G8 and C9 to G9 correspond to the mixture of β-carotene: linoleic acid emulsion and extracts from turnip (C8 to C9, 0.8 µL/well; D8 to D9, 1.6 µL/well; E8 to E9, 2.4 µL/well; F8 to F9, 3.2 µL/well; G8 to G9, 4 µL/well).

173

Later, Tsushida et al. (1994) improved upon it, creating a

reliable assay for measuring antioxidants in a test tube. The

BCB assay was adaptable to high-throughput methodology

because no special equipment or reagents were required, and

a microplate and reader device were available in the recent

study (Dapkevicius et al., 1998; Koleva et al., 2002; Kulisic

et al., 2004; Breksa and Manners, 2006). However, these

evaluations of AOA normally used percentage inhibition of

oxidation at steady state or concentrations sufficient to inhib-

it 50% of oxidation (IC50). Percentage inhibition of oxidation

at steady state yielded a better correlation than the values

relative to the IC50 value (Castro et al., 2006), but this assay

did not make it possible to obtain repeatable data (Roginsky

and Lissi, 2004). Furthermore, it is difficult to correlate the

AOA of natural products measured by various assays and

units. Hence, we modified the display of the AOA measured

by the BCB assay as µmol of Trolox equivalents (TE) for the

first time, reflecting results ob-

tained using the ORAC and DPPH

assays (Table 3).

Linearity and Reproducibility

of the BCB assay In order to as-

certain the reproducibility and re-

liability of this micro-BCB assay,

we examined linearity, range, pre-

cision, and accuracy as follows.

(a) Linearity. The correla-

tion between the difference in

absorbance (470 nm) at 5 min

and that at 50 min (⊿470 nm (5

- 50)) for the antioxidant and its

concentration was evaluated us-

ing Trolox and α-tocopherol (Fig.

2). Fig. 2A illustrates the BCB

curves in the presence of Trolox

or α-tocopherol. Figs. 2B and 2C

depict the linear response between

the concentration of Trolox and

α-tocopherol, and the difference

in absorbance (⊿470 nm (5 -

50)) (r > 0.99).

(b) Range. For Trolox, the lin-

ear response between concentra-

tion and difference of absorbance

was 15.6 µmole to 125 µmol (from

0.2 to 1 nmol/assay).

(c) Precision and Accuracy.

Table 2 summarizes the precision

and accuracy of the BCB assay

using α-tocopherol as a candidate

compound. The precision, expressed as the coefficient of

variation (% CV) of the intermean for the BCB value relative

to Trolox, was ± 8.71%. The accuracy of the assays varied

from 86.96% to 112.58% within individual batches. On the

other hand, the accuracy varied from 93.38% to 104.81%

among all the batches.

Comparison of the AOA using three different assays In

this study, we used the BCB, ORAC, and DPPH assays as

in vitro assays to measure AOA. The ORAC and DPPH as-

says had two major mechanisms, HAT and ET, respectively

(Huang et al., 2005). We evaluated the AOA of crop extracts

from Ibaraki Prefecture using these assays and compared the

results to that of the BCB assay (Table 3 and 4).

The intra-assay coefficient of variation (CV) of this BCB

assay was 7.20% and the inter-assay CV was 14.40%, and

this result was almost identical to that of the ORAC (H-

ORAC (9.74 and 14.10%) and L-ORAC (12.95 and 18.76%))

i. MikaMi et al.

Fig. 2. Typical evaluation of AOA by the micro-BCB assay.(A) Rate of BCB in control without antioxidant (-), in the presence of Trolox (1 nM (■), 0.2 nM (▲) per assay), and α-tocopherol (0.01 nmol (◇), 0.02 nmol (●), 0.03 nmol (△) and 0.04 nmol (○) per assay). T-0.2 and T-1 indicate the differences in absorbance (470 nm) at 5 min and 50 min of the standard solution of Trolox (0.2 nM and 1 nM). A to C and D are the differ-ences in absorbance (470 nm) at 5 min and 50 min of the standard solution of a sample solution (α-tocopherol). (B) The differences in absorbance (470 nm) at 5 min and 50 min (absorbance ⊿470 nm) of different concentrations of Trolox standards are plotted as a function of concen-tration. The subsequent curve is then used to interpolate the AOA of unknown samples. (C) Absorbances (⊿470 nm) of different concentrations of samples (α-tocopherol) are plotted. The resultant determinations are expressed as Trolox equivalents.

0.2

0.4

0.6

0.8

1

1.2

0 10 20 30 40 50Time (min)

Abs

orba

nce

(470

nm

)

D AB

T-0.2 T-1

C

(A)

y = - 0.1353 x + 0.3582

R2 = 0.9928

0.0

0.1

0.2

0.3

0.4

0 0.2 0.4 0.6 0.8 1

⊿47

0nm

(5-5

0)

Trolox

T-0.2

Trolox（nmol/assay)

(B)

y = - 2.7657 x + 0.397

R2 = 0.9916

0.0

0.1

0.2

0.3

0.4

0.00 0.02 0.04 0.06

?-tocopherol

A B C D

α-tocopherol (nmol/assay)

(C)

⊿47

0nm

(5-5

0)

T-1

174

Development and Validation of a Microplate-based β-carotene Bleaching Assay

Columns 3 and 10 Columns 4 and 9 Columns 5 and 8 Columns 6 and 7 Between batches

Run1

Intrameana 15.94 16.82 17.50 17.33 16.90

SDb 2.61 1.32 0.88 1.28 0.70

CVc 16.40 7.87 5.00 7.39 4.16

%RECd 102.53 108.20 112.58 111.51 108.72N 2 2 2 2 8

Run 2Intramean 13.52 14.70 15.25 15.76 14.81

SD 3.30 1.49 0.37 1.42 0.96CV 24.45 10.13 2.44 9.03 6.50

%REC 86.96 94.60 98.14 101.38 95.29N 2 2 2 2 8

Run 3Intramean 14.09 14.89 16.12 14.60 14.92

SD 1.71 0.44 1.31 0.40 0.86CV 12.15 2.93 8.11 2.73 5.78

%REC 90.64 95.78 103.71 93.97 96.04N 2 2 2 2 8

sehctabehtllAsnurdelooPIntermean 14.51 15.47 16.29 15.90 15.54

SD 1.26 1.17 1.13 1.37 0.76CV 8.71 7.57 6.94 8.61 4.91

%REC 93.38 99.53 104.81 102.28 100N 6 6 6 6 24

Columns 2 and 11(wells B2-G2 and B11-G11) were used for Trolox standard, and remaining columns (wells B3-G3,

B4-G4, B5-G5, B6-G6, B7-G7, B8-G8, B9-G9, and B10-G10) were used for α-tocopherol.a Relative BCB value of α-tocopherol (compared to the Trolox value of 1; µmol-α-tocopherol equivalent/µmol-Trolox).b Standard deviation. c Relative standard deviation (CV, coefficient of variation). d Recovery

Table 2. Precision and accuracy of the micro-BCB assay.

TAC 3

Interassay 5 Intra-

assay 5 InterassayIntra-assay

InterassayIntra-assay

Interassay InterassayIntra-assay

InterassayIntra-assay

1 Yacon 10.90 ± 1.35 （12.37） 2.16 0.14 ± 0.04 （28.46） 16.53 6.19 ± 0.55 ( 8.95) 8.18 6.33 ± 0.59 （ 9.34） 5.95 ± 0.35 （ 5.84） 12.40 1.25 ± 0.04 （3.25） 2.91

2 Potato 7.07 ± 1.08 （15.31） 4.41 0.12 ± 0.05 （43.72） 10.41 5.07 ± 0.65 （12.91） 8.49 5.20 ± 0.71 （13.60） 4.71 ± 0.63 （10.82） 7.36 0.46 ± 0.04 （9.32） 6.29

3 Sweet potat 10.61 ± 2.48 （23.33） 15.72 0.54 ± 0.09 （ 7.64） 5.07 6.77 ± 1.00 （14.82） 9.44 7.30 ± 1.10 （15.20） 7.21 ± 0.58 （ 4.40） 1.19 0.96 ± 0.04 （3.96） 9.19

4 Pumpkin 32.50 ± 4.54 （13.97） 2.00 1.19 ± 0.13 （11.18） 12.72 9.58 ± 2.38 （24.80） 10.54 10.77 ± 2.51 （23.29） 9.91 ± 0.62 （ 6.25） 10.42 2.52 ± 0.19 （7.43） 2.77

5 Melon 6.56 ± 1.03 （15.77） 9.93 0.02 ± 0.01 （27.48） 11.39 1.94 ± 0.52 （20.37） 1.36 1.96 ± 0.52 （26.69） 1.71 ± 0.03 （ 1.53） 5.16 0.38 ± 0.02 （5.85） 5.13

6 Kyoho Grape 7.48 ± 1.33 （17.76） 6.61 0.01 ± 0.00 （13.16） 9.12 3.28 ± 0.80 （15.66） 20.03 3.29 ± 0.80 （24.36） 3.05 ± 0.22 （ 7.10） 15.35 0.41 ± 0.01 （3.42） 2.61

7 Mulukhiyy 181.93 ± 15.21 （ 8.36） 2.27 2.98 ± 0.68 （22.85） 15.56 83.45 ± 13.33 （12.28） 7.93 86.43 ± 14.01 （16.21） 102.95 ± 6.47 （ 6.29） 8.18 26.60 ± 1.29 （4.85） 6.05

8 Onion 6.12 ± 0.85 （13.81） 6.33 0.08 ± 0.01 （22.72） 6.51 6.70 ± 1.19 （10.72） 3.23 6.78 ± 1.21 （17.77） 8.58 ± 1.47 （10.68） 3.49 0.42 ± 0.01 （1.38） 3.11

9 Ginger 291.21 ± 21.98 （ 7.55） 12.17 516.82 ± 58.15 （ 8.57） 26.78 187.44 ± 14.50 （ 7.74） 17.46 704.27± 72.65 （10.32） 405.44 ± 30.94 （ 7.63） 10.62 91.00 ± 2.17 （2.39） 2.26

10 Lotus root 15.88 ± 2.27 （14.32） 10.33 0.30 ± 0.02 （ 4.55） 17.05 76.95 ± 11.85 （14.70） 11.33 77.24 ± 11.87 （15.36） 81.72 ± 8.12 （ 9.94） 5.05 17.79 ± 1.58 （8.87） 3.52

11 Turnip 0.24 ± 0.04 （15.83） 7.30 0.10 ± 0.01 （16.06） 11.33 1.81 ± 0.22 （12.17） 9.10 1.91 ± 0.23 （12.06） 1.68 ± 0.18 （10.55） 3.78 0.52 ± 0.01 （1.50） 4.71

1 AOA obtained by beta-carotene bleaching (BCB) assay (see materials and methods). 2 L-ORAC denotes Lipophilic ORAC, and H-ORAC denotes Hydrophilic ORAC. 3 Total antioxidant capacity (TAC) = L-ORAC + H-ORAC.

4 W/O Hex, without extraction with lipophilic solvent (hexane). 5 Experiment data on three or more analyses of average ± SD (CV(%) of interassay) and %CV of intraassay were reported.

All the data were expressed as micromoles of Trolox equivalents per gram of fresh weight (µmol of TE/g of FW).

SampleNo.

Samplename

AOA (µmol TE/g of FW)

BCB 1 DPPHH-ORAC (W/O Hex) 4H-ORAC 2L-ORAC 2

Table 3. Antioxidant activities (AOAs) of crops from Ibaraki Prefecture, Japan.

assays (Table 3). Huang et al. (2002b) and Prior et al. (2003)

also reported that the CV of the H-ORAC assay in a 96-well

plate was 13.4%, and both intraday and interday % CV val-

ues were less than 15%. In addition, the symmetric sample

layout in the microplate might predominate for reproducible

results in the BCB assay (Table 2). These results indicate this

micro-BCB assay is useful and convenient, and has good

reliability. However, the CV of the DPPH assay was the low-

est (4.41% and 4.75%), and reproducibility was high (Table

3) in these three assays. This result was consistent with the

report that the DPPH assay indicated no differences among

runs in comparison with the ORAC assay (Thaipong et al.,

2006).

Judging from the AOAs in Table 3, compared with the

correlation between the AOAs measured by the DPPH and

ORAC assays from crop extracts (0.948 < r < 0.999), that of

the BCB and other assays was low (0.840 < r < 0.939) (Table

4). Although both the BCB and ORAC assays are HAT-based

methods, AOA measured by the BCB assay has a poor rela-

tion to AOAs measured by the ORAC and DPPH assays.

175

This contradiction might be due to the difference in solvents

used. In Table 3, 80% EtOH extracts from several crops were

used for the BCB and DPPH assays. On the other hand, those

of hexane and AWA extracts were used for the ORAC assay.

Therefore, further experimentation was performed as fol-

lows.

Effectiveness of the solvent difference In order to exam-

ine the above-mentioned contradiction that might be due to

solvent differences, AOA in ginger extracts (newly prepared

by 80% EtOH, hexane and AWA) was measured by the BCB,

ORAC and DPPH assays, and results were compared (Fig. 3).

In this case, there was good correlation between the AOAs

measured by the BCB assay and those of the ORAC assay

(0.95). Compared with this, poor correlations were shown,

not only between the AOAs measured by the BCB and

DPPH assays (0.67), but also between the AOAs measured

by the DPPH and ORAC assays (0.74). Therefore, the differ-

ences in the AOAs of ginger extracts measured by the three

different assays arose mainly from solvent differences, rather

than the differentiation of mechanisms.

In addition to this, to resolve questions regarding the ef-

fectiveness of the solvent in dissolving the sample, we also

compared L-ORAC and H-ORAC with/without extraction

with lipophilic solvent (Table 3). The ORAC value of ginger

from direct extraction by AWA (405.44 ± 30.94 µmol TE/g)

was 21.6% lower than that after lipophilic extraction (516.82

± 58.15 µmol TE/g), and 42.4% lower than TAC (704.27

± 72.65 µmol TE/g). These data clearly indicate that some

lipophilic antioxidants might have been extracted with AWA

by direct extraction, but the ORAC values measured by hy-

drophilic extraction alone did not reflect the total AOA, as

indicated by Wu et al. (2004b).

Furthermore, to examine whether this phenomenon could

be observed in AOA measured by the BCB and DPPH as-

says, ginger extracts (prepared as for the ORAC assay) were

used for the BCB and DPPH assays (Fig. 3). A ginger sample

was newly purchased and prepared (due to the lack of an ex-

isting one) for use in this test. Therefore, the AOA of ginger

extracts shown in Fig. 3 was different from that of Table 3,

indicating sample variation. As shown in Fig. 3, in extracts

prepared using the same solvents, the AOAs determined by

the BCB assay were correlated with those of ORAC. Fig. 3

shows that the AOA of ginger using direct extraction with

AWA (AWA (W/O Hex)) was about 45% higher than that

after lipophilic extraction ( 2 AWA (with Hex)), but much

lower than TAC (total antioxidant capacity: 1 + 2 ). On

the other hand, measurements using the DPPH assay were

greatly different, with the AOA of ginger from direct extrac-

tion with AWA (; AWA (W/O Hex)) being about 70% higher

than residue AWA ( 2 ), and TAC was lower than AWA (with

Hex). For the elucidation of this phenomenon, further inves-

tigation will be needed. However, this is a digression from

the main subject.

AOAs of Trolox and α-tocophero l Trolox and

α-tocopherol were expected to possess similar AOA because

they have structural similarities. However, Hung et al. (2002)

found that the ORAC value of α-tocopherol was about half

that of Trolox. Therefore, we examined these two compo-

nents using the micro-BCB assay and compared AOAs using

the DPPH and ORAC assays. Table 5 presents the results

i. MikaMi et al.

Table 4. Pearson’s correlation coefficients of AOAs of crops from Ibaraki Prefecture of Japan measured by BCB ORAC, and DPPH assays (#1 to 6).

Fig. 3. AOAs of ginger extracts prepared by different solvents as determined by the BCB, ORAC and DPPH assays.Sample directly extracted with AWA showed AWA (W/O Hex). AWA (with Hex) shows the hydrophilic sample after extracted lipo-philic extract by Hexane. Total antioxidant capacity (TAC) was cal-culated by summing the hydrophilic and lipophilic samples ( 1 + 2

). See Materials and Methods for further details. Data are expressed as mean ± SD of three determinations.

1 presents the H-ORAC directory extracted with AWA, without extractionwith lipophilic solvent. AOAs of crop extracts were correspond to Table 3.

BCB

DPPH

ORAC

Trolox

1

1

1

α-tocopherol

0.5 ± 0.02 c

a BCB assay was repeated more than three times, and average data wasshowed as relative AOA (µmol-Trolox equivalent/µmol-α-tocopherol).

b

a0.064

0.946

b Suda et al. (2003). c Huang et al. (2002a).

Table 5. Relative AOAs of α-tocopherol measured by BCB, DPPH, and ORAC assays.

176

of the AOAs of Trolox and α-tocopherol measured by the

micro-BCB, DPPH, and ORAC assays. The values indicated

by µmol of Trolox equivalent/µmol of α-tocopherol were

significantly different. These two compounds had structural

similarities but different affinities to each reaction or sub-

strate. Castro et al. (2006) also found that the AOA of similar

compounds changed significantly according to the substrate

used to conduct the evaluation, the parameter adopted to

compare the substances in the same assay, and the form ad-

opted to express the antioxidant concentration.

Accumulated AOA data and its utilization Ginger, mu-

lukhiya, and lotus root indicated high AOAs using all assays

(Table 3). Tsushida et al. (1994) and Ikeba and Kashima

(2006) found AOA by comparing BHA in the extracts of

various vegetables. They discovered four groups of activ-

ity: very high (more than 100 mg BHA equivalent/100 g),

high (50 to 100 mg BHA equivalent/100 g), medium (5 to 25

mg BHA equivalent/100 g), and low (less than 5 mg BHA

equivalent/100 g) using a test tube and a UV spectrometer.

This time, the micro-BCB assay confirmed that these broadly

grouped AOA data were consistent with this experiment (Ta-

ble 3). In addition, the differences between the AOAs in the

samples were clearly shown as a µmol of Trolox equivalent.

Previously, a considerable amount of AOA data has been

accumulated using various assays and units: over 100 dif-

ferent kinds of US foods including fruits, vegetables, nuts,

dried fruits, spices, and infant food and others have been

measured by the ORAC assay (Wu et al., 2004a), 200 agri-

cultural products from the Tohoku area of Japan have been

determined by the DPPH and SOD assays (Kimura et al.,

2002), 110 fruits and vegetables cultivated in the Okinawa

region of Japan have been measured by the DPPH assay (Suda

et al., 2005; Maeda et al., 2006), and 76 kinds of vegetables

produced in Ibaraki Prefecture, Japan, have been determined

by the BCB assay (Ikeba and Kashima, 2006). Nevertheless,

because of the divergence in units (i.e., IC50, SOD units/g,

mg-Gallic acid equivalent/g, milli-mole- Gallic acid equiva-

lent/g), comparison of these data has proved difficult. Here,

for the first time, the AOAs of several crops measured by

three different methods are presented using the same units

(µmol TE/g).

As for the dissimilarities between the BCB assay and

the other two assays, they are as follows. 1) The BCB assay

uses a model lipid substrate (linoleic acid) in an emulsified

form, and the method lies in between methods employing

only model substrates (e.g. the DPPH and ORAC assays)

and those using real lipids (Koleva et al., (2002)). 2) The

BCB assay utilizes carotenoid bleaching via autoxidation,

oxidation induced by light or heat, or oxidation induced by

peroxyl radical. 3) This decolorization can be diminished

or prevented by classic antioxidants that donate hydrogen

atoms to quench radicals (known as the HAT-based method).

The ORAC assay is also a HAT-based method; on the other

hand, the DPPH assay is a SET-based method. 4) β-carotene

decolorization at 470 nm can occur by multiple pathways:

initiation, progression, and termination reactions. The initia-

tion reaction corresponds to the HAT-based method; there-

fore, interpretation of results can be complicated (Prior et

al., (2005)). Instead of differences among the three assays

as mentioned above, when comparisons were made between

ginger extracts using the same solvents, there was, intrigu-

ingly, high correlation between the AOAs measured by the

BCB and ORAC assays (Fig. 3). However, the most decisive

evidence is that the AOAs measured by the three different

methods can be displayed with identical units (µmol TE/g).

Of course, several factors (e.g. digestion and absorption

from the gastrointestinal (GI) tract) influence the AOA of

foods, and the relationship between the structure of antioxi-

dant compounds and their AOAs measured by assays is note-

worthy (Burda and Oleszek, 2001; Sakakibara et al., 2003).

However, the outcomes of this study would facilitate the in-

terpretation of the vast accumulation of AOA data measured

by different methods. It is expected that this knowledge

would be utilized in making wise food choices and ingesting

antioxidants with every meal for a healthy life.

Acknowledgements The authors thank Dr. T. Oki (National Ag-

ricultural Research Center for Kyushu Okinawa Region, National

Agricultural Research Organization) and Mr. K. Yamasaki (Taiyo

Kagaku Co., Ltd.) for their technical advice. This study was per-

formed with financial support from the Ministry of Agriculture,

Forestry and Fisheries of Japan.

ReferencesBreksa, A.P.III and Manners, G.D. (2006). Evaluation of the Anti-

oxidant Capacity of Limonin,Nomilin,and Limonin Glucoside.

J.Agric.Food Chem., 54, 3827-3831.

Burda, S, and Oleszek, W. (2001). Antioxidant and antiradical ac-

tivities of flavonoids. J. Agric. Food Chem., 49, 2774-2779.

Cao, G., Verdon, C.P., Wu, A.H., Wang, H. and Prior, R.L. (1995).

Automated assay of oxygen radical absorbance capacity with the

COBAS FARA II. Clin. Chem., 41, 1738-1744.

Castro, I.A., Rogero, M.M., Junqueira, R.M. and Carrapeiro, M.M.

(2006). Free radical scavenger and antioxidant capacity correla-

tion of α-tocopherol and Trolox measured by three in vitro meth-

odologies. International Journal of Food Sciences and Nutrition.,

57,75-82.

Decker, E.A., Warner, K., Richards, M.P. and Shahidi, F. (2005).

Measuring antioxidant effectiveness in food. J. Agric. Food

Chem., 53, 4303-4310.

Development and Validation of a Microplate-based β-carotene Bleaching Assay 177

Dapkevicius, A., van Beek, T.A., Linssen, J.P.H. and Venskutonis,

R. (1998). Rapid spectroscopic screening for antioxidant activ-

ity in sage, thyme and oregano isolates with the beta-carotene-

linoleic acid model system. In “Natural Product Analysis”, ed. by

P. Schreier, M. Herderich, H.U. Humpf and W. Schwab. Vieweg,

Braunschweig, pp.235-237.

Huang, D., Ou, B., Hampsch-Woodill, M., Flanagan, J.A. and

Deemer, E.K. (2002a). Development and validation of oxygen

radical absorbance capacity assay for lipophilic antioxidants us-

ing randomly methylated beta- cyclodextrin as the solubility en-

hancer. J. Agric. Food Chem., 50, 1815-1821.

Huang, D., Ou, B., Hampsch-Woodill, M., Flanagan, J.A. and Prior,

R.L. (2002b). High-Throughput Assay of Oxygen Radical Absor-

bance Capacity (ORAC). J. Agric. Food Chem., 31, 4437-4444.

Huang, D., Ou, B. and Prior, R.L. (2005). The chemistry behind an-

tioxidant capacity assays. J. Agric. Food Chem., 53, 1841-1856.

Ikeba, T. and Kashima, K. (2006). Evaluation of antioxidant activ-

ity of raw and cooked vegetable extracts produced in Ibaraki pre-

fecture. Ibaraki Nougyo Sougou Senter Engei Kenkyusyo kenkyu

Houkoku, 14, 27-33.

Kimura, T., Yamagishi, K., Suzuki, M. and Shinmoto, H. (2002).

Relative estimation of the radical scavenging activities of agri-

cultural products. Nippon Shokuhin Kagaku Kogaku Kaishi, 49,

257-266 (in Japanese).

Koleva, I.I., van Beek, T.A., Linssen, J.P., de Groot, A. and Evsta-

tieva, L.N. (2002). Screening of plant extracts for antioxidant ac-

tivity: a comparative study on three testing methods. Phytochem.

Anal., 13, 8-17.

Kulisica, T., Radonicb, A., Katalinicc, V. and Milos, M. (2004). Use

of different methods for testing antioxidative activity of oregano

essential oil. Food Chemistry, 85, 633-640.

Maeda, G., Takara, K., Wada, K., Oki, T., Masuda, M., Ichiba, T.,

Chuda, Y., Ono, H. and Suda, I. (2006). Evaluation of antioxidant

activity of vegetables from Okinawa prefecture and determination

of some antioxidative compounds. J. Food Sci. Technol. Res., 12,

8-14.

Marco, G.J. (1968). Rapid method for evaluation of antioxidants.

JAOCS., 45, 594.

Miller, H.E. (1971). A simplified method for the evaluation of anti-

oxidants. JAOCS., 48, 91.

Moreno, C.S. (2002). Review: Methods Used to Evaluate the Free

Radical Scavenging Activity in Foods and Biological Systems.

Food Science and Technology International, 8, 121-137.

Oki, T., Masuda, M., Kobayashi, M., Nishiba, Y., Furuta, S., Suda,

I. and Sato, T. (2001). Radical scavenging activity of fried chips

made from purple-fleshed sweet potato. Nippon Shiokuhin

Kagaku Kougaku Kaishi, 48, 926-932 (in Japanease).

Oki, T. (2008). Measurement of DPPH· scavenging activity. In

“Shokuhin Kinousei Hyouka Manyualusyu II,” ed. by Nippon

Shokuhin Kagaku Kougakukai, pp.71-78 (in Japanease).

Prior, R.L., Hoang, H., Gu, L., Wu, X., Bacchiocca, M., Howard,

L., Hampsch-Woodill, M., Huang, D., Ou, B. and Jacob, R.

(2003). Assays for hydrophilic and lipophilic antioxidant capac-

ity (oxygen radical absorbance capacity (ORAC(FL))) of plasma

and other biological and food samples. J. Agric. Food Chem., 51,

3273-3279.

Prior, R.L.,Wu, X. and Schaich, K. (2005). Standardized meth-

ods for the determination of antioxidant capacity and phenolics

in foods and dietary supplements. J.Agric. Food Chem., 53,

4290-4302.

Sakakibara, H., Honda, Y., Nakagawa, S., Ashida, H. and

Kanazawa, K. (2003). Simultaneous determination of all poly-

phenols in vegetables, fruits, and teas. J. Agric. Food Chem., 51,

571-581.

Sannaveerappa, T., Westlund, S., Sandberg, A.S., and Undeland, I.

(2007). Changes in the antioxidative property of herring (Clupea

harengus) press juice during a simulated gastrointestinal diges-

tion. J. Agric. Food Chem., 26, 10977-10985.

Suda, I., Masuda, M., Furuta, S., Oki, T. and Kobayashi, M. (2003).

DPPH Radical-Scavenging Activity of Representative Antioxi-

dant. Kyushyu Nogyou Kenkyu., 65, 55.

Suda, I., Oki, T., Nishiba, Y., Masuda, M., Kobayashi, M., Nagai,

S., Hiyane, R. and Miyashige, T. (2005). Polyphenol contents and

radical-scavenging activity of extracts from fruits and vegetables

in cultivated in Okinawa, Japan. Nippon Shokuhin Kagaku

Kogaku Kaishi, 52, 462-471 (in Japanese).

Roginsky, V. and Lissi. E.A. (2004). Review of methods to deter-

mine chain-breaking antioxidant activity in food. Food Chem.,

92, 235-254.

Thaiponga, K., Boonprakoba, U., Crosbyb, K., Cisneros-Zevallosc,

L. and Byrnec, D. H. (2006). Comparison of ABTS, DPPH,

FRAP, and ORAC assays for estimating antioxidant activity from

guava fruit extracts. Journal of Food Composition and Analysis,

19, 669-675.

Tsushida, T., Suzuki, M. and Kurogi, M. (1994). Evaluation of anti-

oxidant activity of vegetable extracts and determination of some

antioxidants, Nippon Shiokuhin Kagaku Kougaku Kaishi, 41,

611-618 (in Japanese).

Wu, X., Beecher, G.R., Holden, J.M., Haytowitz, D.B., Gebhardt,

S.E. and Prior, R.L. (2004a). Lipophilic and Hydrophilic Antioxi-

dant Capacities of Common Foods in the United States. J. Agric.

Food Chem., 52, 4026-4037.

Wu, X., Gu, L., Holden, J., Haytowitz, D.B., Gebhardt, S.E., Beech-

er, G. and Prior, R.L. (2004b). Development of a database for

total antioxidant capacity in foods: a preliminary study. Journal

of Food composition and analysis, 17, 407-422.

Yamaki, K. and Mori, Y. (2006). Evaluation of α-glucosidase Inhib-

itory Activity in Colored Foods: a Trial Using Slope Factors of

Regression Curves. Nippon Shiokuhin Kagaku Kougaku Kaishi,

53, 229-231 (in Japanese).

i. MikaMi et al.178