Document Outline

has been used as a standard antioxidant, and ORAC values are
expressed in trolox-equivalent units. When antioxidants are
present, the ﬂuorescence decay time course shows a delay or
shift, and the magnitude of the shift was converted into ORAC
values with computer-aided analysis

[12]

The objective of this study was to develop an ORAC-like assay

that directly quantiﬁes the antioxidant’s scavenging capacity
against AAPH-derived free radicals. This method employed the
electron paramagnetic resonance (EPR) spin trapping method,
and we named it ORAC–EPR. The spin trapping method is an estab-
lished technique for free radical identiﬁcation and quantiﬁcation,
where unstable free radicals are trapped and stabilized via free

0003-2697/$ - see front matter Ó 2008 Elsevier Inc. All rights reserved.
doi:10.1016/j.ab.2008.12.022

Corresponding author. Fax: +81 11 612 3617.
E-mail address:

hgfujii@sapmed.ac.jp

(H. Fujii).

Abbreviations used: ORAC, oxygen radical absorbance capacity; AAPH, 2,2-

azobis(2-amidinopropane) dihydrochloride; USDA, US Department of Agriculture;
ORAC–FL, ﬂuorescence-based ORAC assay; EPR, electron paramagnetic resonance;
ORAC–EPR, ORAC assay employing EPR spin trapping method; UV, ultraviolet; GSH,
glutathione; CYPMPO, 5-(2,2-dimethyl-1,3-propoxy cyclophosphoryl)-5-methyl-1-
pyrroline-N-oxide; DMSO, dimethyl sulfoxide; HFSC, hyperﬁne splitting constant;
AIBN, azobisisobutylnitrile.

Analytical Biochemistry 386 (2009) 167–171

Contents lists available at

ScienceDirect

Analytical Biochemistry

j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / y a b i o

radical trapping compounds (spin traps)

[20–22]

. The stabilized

product (spin adduct or free radical adduct) is identiﬁed and quan-
tiﬁed with EPR spectrometry. ORAC–EPR determines the type and
level of AAPH-derived free radicals in the presence and absence
of antioxidants; therefore, a constant and reproducible amount of
free radicals from AAPH must be produced. Although ORAC–FL
used thermal decomposition to generate AAPH free radicals, we
employed in situ photolysis of AAPH and were able to produce a
reproducible amount of free radicals with ultraviolet (UV)/visible
light illumination. In the presence of antioxidant, the free radical
level was decreased due to the antioxidant’s scavenging activity,
and ORAC–EPR values can be calculated following a simple kinetic
equation. Using ORAC–EPR, we determined ORAC values for several
natural antioxidants and human serum, and we critically evaluated
the method in comparison with ORAC–FL.

Materials and methods

Materials

AAPH (

) was purchased from Wako Pure Chemicals (Osaka,

Japan). Ascorbic acid (vitamin C), N-acetylcysteine, glutathione
(GSH), caffeic acid, genistein, rutin, quercetin, catechin, uric acid,
trolox (

), and sodium phosphate were purchased from Wako

Pure Chemicals (Osaka, Japan). The spin trap CYPMPO (5-(2,2-di-
methyl-1,3-propoxy

cyclophosphoryl)-5-methyl-1-pyrroline-N-

oxide) (

) was obtained from Radical Research (Tokyo, Japan).

Blood was obtained from healthy adult volunteers according to the
protocol approved by the institutional human subject use commit-
tee at Sapporo Medical University.

A light illuminator (cat. no. RUVF-203SR, Radical Research) was

employed to produce free radicals from AAPH. The light source was
either a 200-W Xe arc lamp (San-ei Electronics, Osaka, Japan),
which delivers mainly visible light with a wavelength of 400 to
700 nm, or a 200-W Hg-Xe arc lamp (San-ei Electronics), which
delivers mainly UV light with a wavelength of 200 to 400 nm.
The illuminator was equipped with a quartz ﬁber-optic guide,
and its far end was mechanically ﬁtted to the hole in front of the
EPR cavity. The illuminator is equipped with a computer-controlla-
ble mechanical shutter, and the illumination period can be set in
the range of 0.1 to 100 s with precision of ±0.01 s. The illumination
conditions were optimized so that AAPH was efﬁciently photode-
composed to generate free radicals

[23]

. The light source was se-

lected to avoid possible photodecomposition of antioxidants. We
used mainly attenuated Xe arc lamp light and illuminated it for

60 s to a sodium phosphate buffer solution (100 mM, pH 7) con-
taining 10 mM AAPH and 10 mM CYPMPO. When an Hg-Xe arc
lamp was used, the AAPH concentration was 1 mM and the illumi-
nation time was 5 s. Under these conditions, AAPH showed the EPR
signal with signal-to-noise ratio larger than 100 (

). Quercetin

and catechin are insoluble in buffer; therefore, 1% dimethyl sulfox-
ide (DMSO) phosphate buffer solution was used as a solvent. The
ORAC–EPR value for this solvent was approximately 0.001 relative
to CYPMPO; thus, the effect on ﬁnal results was considered to be
negligible. For EPR measurements, a JEOL RE-1X X-band ESR spec-
trometer (Tokyo, Japan) equipped with 100 kHz ﬁeld modulation
and Win-Rad operation software (Radical Research) was employed.
Typical spectrometer settings were as follows: ﬁeld modulation
width, 0.1 mT; microwave power, 6 mW; ﬁeld scan width/rate,
±7.5 mT/2 min; time constant, 0.1 s.

ORAC–EPR measurement

A typical ORAC–EPR procedure was as follows. A phosphate buf-

fer solution of AAPH, antioxidant, and CYPMPO was loaded into a
disposable borosilicate ﬂat sample cell (Radical Research) and
placed inside the EPR cavity. In situ illumination was carried out,
and the EPR signal was recorded immediately after the illumina-
tion was stopped. The intensity of the selected EPR line was re-
corded as a function of antioxidant concentration and analyzed
following the procedures described in the next section.

Data analysis

The peak-to-peak intensity of the selected EPR line of the free

radical adduct was monitored in the presence or absence of the se-
lected antioxidant. The kinetic formulation of the competitive reac-
tion between the spin trap (CYPMPO in this case) and the
antioxidant against free radical has been published elsewhere

[24]

. Brieﬂy, in the presence of the spin trap (ST) and the antioxi-

dant (AOx), the following free radical (R) scavenging reactions
should occur:

Fig. 1. Structures of AAPH, trolox, and CYPMPO.

Fig. 2. Top panel: EPR spectrum of CYPMPO free radical adduct. Aqueous solution of
AAPH (1 mM) and CYPMPO (10 mM) was illuminated with Hg-Xe arc for 5 s, and
the EPR spectrum was recorded. Bottom panel: Computer-simulated EPR spectrum
for the recorded spectrum. HFSCs used for the simulation were as follows:
A(H) = 1.24, A(N) = 1.40, and A(P) = 4.70. The free radical species was assigned to RO
radical adduct, where R is the H

N(HN)C-C(CH

)

- group (see Results section).

168

ORAC-like assay / S. Kohri et al. / Anal. Biochem. 386 (2009) 167–171

R þ ST ! R-adduct rate constan : tk

ð1Þ

R þ AOx ! product rate constant : k

AOx

ð2Þ

When I

and I are the EPR peak heights in the presence of ST

alone and ST + AOx, respectively, the amount of the product in
Eq.

(2)

will be I

– I. Thus, I and I

I are proportional to k

[R][ST]

and k

AOx

[R][AOx], respectively, and

ðI

IÞI ¼ k

AOx

½R½AOx=k

½R½ST:

ð3Þ

Nearly negligible portions of ST and AOx will be consumed in

this reaction, we can safely replace [ST] and [AOx] with [ST]

and

[AOx]

, respectively, where []

denotes the initial concentration

of the component. Then Eq.

(3)

can be rewritten as

I 1 ¼ k

AOx

½AOx

½ST

ð4Þ

In conclusion, a plot of I

/I – 1 against [AOx]

/[ST]

will give a

zero-crossing line with the slope of k

AOx

. When the same spin

trap and the free radical-generating condition are used, k

be-

comes common for all antioxidants. If k

AOx

for each sample is

normalized using that of the standard trolox, the result will be ex-
pressed in trolox-equivalent units. Such ORAC–EPR values can be
directly compared with corresponding ORAC–FL values.

Results

Identiﬁcation of free radicals produced from AAPH

An EPR spectrum shown in

(top) was obtained when an

aqueous solution containing AAPH and CYPMPO was photolyzed
with either an Xe arc or Hg-Xe arc lamp. Exactly the same spec-
trum pattern was recorded when the AAPH solution was heated
at 37 °C for 10 min (data not shown). The overall spectral intensity
was dependent on the AAPH concentration and the illumination or
heating time, but the spectral pattern did not depend on those con-
ditions. This spectrum was reproduced using computer spectrum
simulation (

, bottom), gaining the following hyperﬁne split-

ting constants (HFSCs): A(H) = 1.24 mT, A(N) = 1.36 mT, and
A(P) = 4.80 mT. Based on the comparison with HFSCs of AAPH rad-
ical adduct of the analogous spin trap DEPMPO

[25]

(A(H) = 1.27

mT, A(N) = 1.34 mT, and A(P) = 4.63 mT)

[26]

, the spectrum shown

is assignable to the alkyl-oxy (RO) radical adduct, where R

represents the H

N(HN)C–C(CH

)

-group.

ORAC–EPR values for compounds with antioxidant activity and human
serum

Fig. 3

shows the dependence of EPR signal intensity when the

trolox concentration was increased stepwise. The intensity I

or I

of the line marked with an asterisk () in

Fig. 3

was plotted accord-

ing to Eq.

(4)

(

Fig. 4

). The slope of the line (i.e., k

TROLOX

CYPMPO

)

passing through each point and the origin were calculated and
averaged, and the standard deviation was calculated. The k

TROLOX

CYPMPO

value was used to calculate trolox-equivalent ORAC–EPR

values. In

Fig. 4

, the plot for the antioxidant GSH is also illustrated.

Trolox shows a larger slope than GSH, indicating that the ORAC–
EPR value of trolox is larger than that of GSH.

lists trolox-equivalent ORAC–EPR values of various anti-

oxidants. ORAC–FL analysis was performed for most of these anti-
oxidants using ﬂuorescein (ORAC

) or b-phycoerythrin (ORAC

)

as the ﬂuorescence probe. Those values are listed in

ORAC–EPR values for human serum are listed in

Table 2

with pre-

vious ORAC–FL results. Because the amount of individual antioxi-
dant in serum is unknown, the numbers are shown in units of
millimolar (mM) trolox equivalent (i.e., mmol trolox equivalent/L
serum).

Discussion

The photochemical production of AAPH free radicals and the

EPR spin trapping technique were combined in ORAC–EPR, and
we obtained ORAC–EPR values for various antioxidants and hu-
man serum in a much shorter time than in ORAC–FL. The photol-
ysis of AAPH was so efﬁcient that we were able to use a much
lower concentration (e.g., 1 or 10 mM) of AAPH than in ORAC–
FL (typically 75 mM)

[8]

; this may be beneﬁcial in eliminating

the possibility of the solute aggregation. A possible drawback
of ORAC–EPR is photochemical decomposition of the sample
antioxidant. By using a relatively short illumination of Xe arc
light (wavelength longer than 400 nm), the photochemical
decomposition may be minimized. The use of Hg-Xe arc light
has made the illumination time very short (e.g., <5 s), and this
may be beneﬁcial for thermally unstable specimens. Other mer-
its of ORAC–EPR include the simplicity of the analysis, whereas
in ORAC–FL computer-aided calculation was necessary

[12]

Moreover, the selection of experimental conditions in ORAC–
EPR was not strict. First, we obtained similar results (slopes)
by changing the absolute concentrations of the antioxidant and
the spin trap while keeping the ratio constant (

Fig. 4

). Second,

the light intensity, illumination time, and AAPH concentration
can be modiﬁed within a reasonable range. Third, in contrast
to thermal ORAC–FL, the start and stop of AAPH radical produc-
tion was controllable simply by turning the light on and off.
Although we employed a commercially available illuminator, this
is not a critical requirement given that a suitable illuminator
could be improvised from a commercially available Xe arc lamp
and a mechanical shutter. Although the use of a new spin trap
(CYPMPO) facilitated data collection, this is not a critical factor
and other spin traps such as DEPMPO

[26]

could be used.

The discrepancy between ORAC–FL and ORAC–EPR (

Tables 1

and 2

) is not surprising given that, within ORAC–FL data, 1.9 to

3.5 times difference has been reported when the ﬂuorescent probe
was switched from b-phycoerythrin (ORAC

) to ﬂuorescein (OR-

)

[15]

(

). A comparative study by Cao and Prior using

the two ORAC–FL assays indicated signiﬁcant discrepancies be-
tween them

[13]

(

Table 2

). The difference of ORAC–FL values

depending on the selection of the ﬂuorescence probe indicates that
ORAC–FL values might not be antioxidant intrinsic constants, sug-
gesting that the direct comparison of ORAC–FL and ORAC–EPR val-
ues might not be meaningful. We believe that ORAC–EPR is
especially suitable for thermally labile biological specimens be-
cause the sample is not required to be heated for free radical pro-
duction as it is in ORAC–FL. Photochemical production of AAPH
radicals may be a less perturbing means of free radical production
than other methods such as heating and a metal ion-based free
radical generating system (e.g., the Fenton reaction).

Current ORAC assays are indirect methods that do not specify

the type of oxygen radicals to be scavenged by antioxidants. The
spin trapping technique has made possible the identiﬁcation of
AAPH-derived oxygen radicals as alkyl-oxy (RO) radicals (

);

thus, we clearly deﬁne that ORAC–EPR is an evaluation method
for alkyl-oxy radical scavenging capacity. ORAC–FL is considered
as a scavenging assay for alkyl-peroxy (ROO) radicals, as described
in most publications

[8]

. This is because AAPH has been shown to

be an excellent lipid peroxidation agent in aqueous media

[27,28]

Indeed, in nonaqueous media, ROO radicals have been shown to be
spin-trapped in thermal decomposition of a lipid-soluble azo-rad-
ical initiator, azobisisobutylnitrile (AIBN)

[23]

. However, previous

spin trapping studies have shown that free radical species pro-
duced

the

thermal

decomposition

AAPH

were

N(HN)C-C(CH

)

-O) radicals but not ROO radicals

[23,26]

The same RO radicals were spin-trapped in the photolysis of AAPH

ORAC-like assay / S. Kohri et al. / Anal. Biochem. 386 (2009) 167–171

169

in this study (

), indicating that the AAPH-based ORAC assay is

likely to be the assay for RO radical scavenging capacity.

In summary, we have developed a rapid, simple, and depend-

able method to measure ORAC values of antioxidants and biologi-
cal specimens by using photochemical AAPH radical production
and EPR spin trapping. By selecting photochemical free radical pre-
cursors, this method may be applicable to the determination of the
antioxidant’s scavenging capacity against other oxygen radicals.

Fig. 3. Changes in the level of the alkyl-oxy adduct of CYPMPO as a function of the concentration of the antioxidant trolox. In situ Xe-arc light photolysis was carried out for
60 s to the sodium phosphate buffer (100 mM) solution of AAPH (10 mM), CYPMPO (10 mM), and various concentrations of trolox. The peak-to-peak intensity of the line
marked with an asterisk () was adopted as I

or I in Eq.

(4)

to calculate relative alkyl-oxy radical scavenging rates (k

AOx

Fig. 4. Plot of I

/I – 1 against [AOx]

/[CYPMPO]

, where I

and I denote EPR intensity

in the absence and presence of trolox or GSH. [AOx]

and [CYPMPO]

represent the

initial concentrations of each antioxidant and CYPMPO, respectively. The slopes of
the zero-crossing lines represent k

AOx

for trolox and GSH (