Antioxidant Actions of Phenolic Compounds Found in Dietary Plants on Low-Density Lipoprotein and Erythrocytes in Vitro

Original Research
Antioxidant Actions of Phenolic Compounds Found in
Dietary Plants on Low-Density Lipoprotein and
Erythrocytes in Vitro
Rosanna Y.Y. Lam, MPhil, Anthony Y.H. Woo, PhD, Po-Sing Leung, PhD, and Christopher H.K. Cheng, PhD
Department of Biochemistry (R.Y.Y.L., A.Y.H.W., C.H.K.C.), Center of Novel Functional Molecules (C.H.K.C.) and Department
of Physiology (P.S.L.), The Chinese University of Hong Kong, Shatin, N.T., Hong Kong, CHINA
Key words: gingerol, aloe-emodin, barbaloin, rhapontin, anthraquinone derivatives, stilbenes, low-density lipoprotein,
erythrocytes, erythrocyte membranes, oxidative stress, peroxyl radicals, lipid peroxidation, Na?/K?-ATPase, Ca2?-
ATPase, protein sulfhydryl groups
Objective: There is increasing interest in the study of the antioxidant actions of plant phenolic compounds
as evidence shows that consumption of plant products rich in these compounds contributes to protection from
a number of ailments including cardiovascular diseases. In the present study, the antioxidant effects of selected
phenolic compounds from dietary sources, namely barbaloin, 6-gingerol and rhapontin, were investigated.
Methods: Low-density lipoprotein (LDL), erythrocytes and erythrocyte membranes were subjected to
several in vitro oxidative systems. The antioxidant effects of the phenolic compounds were assessed by their
abilities in inhibiting hemolysis and lipid peroxidation of LDL and erythrocyte membranes, and in protecting
ATPase activities and protein sulfhydryl groups of erythrocyte membranes.
Results: 6-Gingerol and rhapontin were found to exhibit strong inhibition against lipid peroxidation in LDL
induced by 2,2?-azobis(2-amidinopropane) hydrochloride (AAPH) and hemin while barbaloin possessed weaker
effects. A similar order of antioxidant potencies among the three compounds was observed on the lipid
peroxidation of erythrocyte membranes in a tert-butylhydroperoxide (tBHP)/hemin oxidation system. On the
other hand, barbaloin and rhapontin were comparatively stronger antioxidants than 6-gingerol in preventing
AAPH-induced hemolysis of erythrocytes. Among the three compounds, only barbaloin protected Ca2?-ATPase
and protein sulfhydryl groups on erythrocyte membranes against oxidative attack by tBHP/hemin. Interestingly,
rhapontin demonstrated protective actions on Na?/K?-ATPase in a sulfhydryl group-independent manner under
the same experimental conditions.
Conclusions: In view of their protective effects on LDL and erythrocytes against oxidative damage, these
phenolic compounds might have potential applications in prooxidant state-related cardiovascular disorders.
INTRODUCTION
from fruits and vegetables is associated with a reduced risk for
cardiovascular diseases [1,2]. The phenolic constituents in
Reactive oxygen species (ROS) and other free radicals are
plants have been suggested to play a role [3]. Phenolic com-
characterized by their ability in causing oxidative damage to
pounds are widely distributed in plants. They are comprised of
the body. They contribute to the etiology of a number of
several classes of compounds such as flavonoids, anthraquino-
pathological conditions including cardiovascular diseases. Ep-
nes, stilbenoids, and their derivatives. The antioxidant proper-
idemiological studies suggest that dietary intake of antioxidants
ties of these phenolic compounds are related to their abilities to
Address reprint requests to: Christopher H.K. Cheng, PhD, Department of Biochemistry, The Chinese University of Hong Kong, Shatin, N.T., HONG KONG. E-mail:
chkcheng@cuhk.edu.hk
R.Y.Y.L. and A.Y.H.W. contributed equally to the work.
Abbreviations: AAPH ? 2,2?-azobis(2-amidinopropane) hydrochloride, DMSO?dimethylsulfoxide, DW ? distilled water, Hepes ? 4-(2-hydroxyethyl)piperazine-1-
ethanesulfonate, LDL ? low-density lipoprotein, PBS ? phosphate-buffered saline, Pi ? inorganic phosphate, ROS ? reactive oxygen species, TBARS ? thiobarbituric
acid-reactive substance, tBHP ? tert-butylhydroperoxide.
Journal of the American College of Nutrition, Vol. 26, No. 3, 233–242 (2007)
Published by the American College of Nutrition
233

---

Antioxidant Mechanism of Phenolic Compounds
donate electrons and to act as free radical scavengers by the
formation of stable phenoxyl radicals [4]. Many structure-
activity relationship studies of the antioxidant mechanisms of
phenolic compounds have been conducted using purely chem-
ical model systems [5,6]. These include assays to assess the
scavenging effects of the phenolic compounds on superoxide
radicals, hydroxyl radicals, 2,2?-azinobis-(3-ethylbenzo-thiazo-
line-6-sulfonic acid) radicals 2,2-diphenyl-1-picrylhydrazyl
radicals, and other free radicals. Simple biological models such
as the lipid peroxidation systems of unsaturated fatty acids [7]
and liver microsomes [8] are also commonly applied. The
simplicity of these systems allows a large number of agents to
be tested conveniently. However, these models also have the
drawback of having low physiological relevance. Although the
copper-catalysed low-density lipoprotein (LDL) oxidation sys-
tem [9] is considered more physiological in this respect and has
recently gained wide acceptance, care should be taken when
applying this system to study the antioxidant mechanisms of
phenolic compounds. It is because the antioxidant effects are
often complicated by the metal-ion chelating effects of the test
compounds [10,11]. In this regard, the development of robust in
vitro model systems with high physiological relevance is still in
need for the study of the antioxidant mechanisms of phenolic
compounds.
In the present study, the antioxidant actions of a selected
number of natural phenolics were studied in several carefully
designed model systems of LDL and erythrocytes which allow
the antioxidant mechanisms to be elucidated. The compounds
studied originated from three common dietary plants, viz. gin-
Fig. 1. Chemical structures of 6-gingerol (1), rhapontin (2), aloe-
ger (Zingiber officinale Roscoe), aloe (Aloe vera (L.) Burm f.)
emodin (3), and barbaloin (4).
and rhubarb (Rheum rhabarbarum L. or R. undulatum L.).
They include 6-gingerol (5-hydroxy-1-(4-hydroxy-3-methoxy-
phenyl)-3-decanone) from ginger, rhapontin (rhaponticin or
18]. In sickle cell anemia, studies have shown that the sickled
3,3?,5-trihydroxy-4?-methoxystilbene 3-?-D-glucoside) from
erythrocytes produce more ROS [19]. In addition, serum anti-
rhubarb, aloe-emodin (1,8-dihydroxy-3-(hydroxymethyl)an-
oxidants such as ascorbate and ?-carotene are found to be
thraquinone) from rhubarb or aloe, and barbaloin (aloin or
lower in sickle cell anemic subjects [20,21]. Exposure of eryth-
1,8-dihydroxy-10-(?-D-glucopyranosyl)-3-(hydroxymethyl)-9-
rocytes to this and other prooxidant conditions can lead to a
anthracenone) from aloe. Their chemical structures are shown
number of membrane changes including lipid peroxidation
in Fig. 1. The content of rhapontin in rhubarb root varies
[22,23], protein crosslinking [23] and sulfhydryl group oxida-
between 3.1% to 4.3% in different seasons [12]. Ginger was
tion [24], resulting subsequently in membrane damage and
found to contain 11% of gingerols including 5% 6-gingerol
hemolysis [22,25]. Membrane enzymes such as ATPases are
[13]. Barbaloin and its aglycone aloe-emodin are the major
also targets of free radical attack [23,26]. Decrease in erythro-
ingredients of aloe leaves. The content of barbaloin in the juice
cyte ATPase activities has been found to coincide with patho-
of aloe leaves was reported to be 15– 40% [14].
logical changes of other clinical parameters in coronary heart
LDL and erythrocytes are important blood components that
disease [27]. Dietary antioxidants, which can reduce the oxi-
are exposed to increased oxidative stress under certain patho-
dative damage on LDL and erythrocytes, might thus confer
physiological conditions. In familial hypercholesterolemia, a
protective effects in prooxidant state-related cardiovascular dis-
defect characterized by reduced cellular uptake of LDL, LDL
orders.
stays in the circulation for an extended period of time. This will
In the present investigation, the protective actions of the test
result in increased LDL oxidation [8] and a higher incidence of
compounds on LDL was studied by determining the extent of
atherosclerosis as considerable evidence has implicated the role
lipid peroxidation in LDL induced by 2,2?-azobis(2-amidino-
of oxidized LDL in atherogenesis [15]. The mechanism of LDL
propane) hydrochloride (AAPH) or hemin. Moreover, AAPH-
oxidation in vivo is not fully understood. Both cell-mediated
induced hemolysis was used to assess the gross damaging
and cell-independent mechanisms have been proposed [16 –
effects of free radicals on erythrocyte membranes and the
234
VOL. 26, NO. 3

---

Antioxidant Mechanism of Phenolic Compounds
protective actions of the test compounds. In addition, lipid
could affect the determination of ATPase activities in the
peroxidation inhibition, ATPase protection and sulfhydryl
subsequent assays.
group protection were also performed on isolated erythrocyte
membranes exposed to tert-butylhydroperoxide (tBHP)/hemin
Lipid Peroxidation of LDL
oxidative damage to further characterize the different aspects of
the protective actions of these compounds on the cell mem-
In the present investigation, oxidation of LDL was induced
branes. Hemin was used to catalyze the generation of peroxyl
by AAPH, a water-soluble free radical generator [25]; or by
radicals and alkoxyl radicals by tBHP (reactions 1 and 2 below)
hemin, a trivalent ferric oxidant which also exists in vivo [17].
[23], subsequently producing the oxidative damage on the
AAPH undergoes thermal decomposition to produce carbon-
erythrocyte membranes.
centered radicals which are converted into peroxyl radicals in
the presence of oxygen. The peroxyl radicals formed could then
HX-FeIII ? tBu-OOH 3 HX-FeII ? tBu-OO ? ? H ?
initiate lipid peroxidation chain reactions [25]. AAPH was used
in the present study such that the antioxidant effects of free
(1)
radical scavengers in aqueous phase could be assessed. In the
HX-FeII ? tBu-OOH 3 HX-FeIII ? tBu-O ? ? OH ?
absence of exogenous hydroperoxides, hemin (HX-FeIII) prop-
agates lipid peroxidation chain reactions by catalyzing the
(2)
formation of lipid peroxyl (LOO?) or alkoxyl (LO?) radicals
from endogenous lipid hydroperoxides (LOOH) in LDL (reac-
tions 3–5 below) [29,30].
MATERIALS AND METHODS
HX-FeIII ? LOOH 3 HX-FeII ? LOO ? ? H ?
(3)
Reagents
HX-FeIII ? LOOH 3 HX-?FeIV ? O?2? ? LO ? ? H ?
Human LDL suspended in phosphate-buffered saline (PBS)
with EDTA was purchased from Calbiochem (La Jolla, CA).
(4)
Trolox, 6-gingerol (?95%) and AAPH were supplied by Wako
HX-FeII ? LOOH 3 HX-FeIII ? LO ? ? OH ?
(5)
Chemical Company (Japan). Aloe-emodin (?95%), barbaloin
(?97%), rhapontin (?99%), tBHP and hemin were purchased
Therefore a compound that could inhibit lipid peroxidation
from Sigma-Aldrich (St. Louis, MO). Dimethylsulfoxide
induced by hemin is likely to be a chain-breaking antioxidant in
(DMSO) was purchased from Merck (Germany). All other
lipid. By employing both oxidation systems, the relative po-
reagents used were of analytical grade.
tency of a test compound in resembling a free radical scavenger
and/or a chain-breaking antioxidant could be evaluated.
The condition of the AAPH-induced lipid peroxidation of
Preparation of Rat Erythrocytes
LDL was 37°C for 2 h in a final volume of 1 mL containing
About 10 mL of blood from an adult male Sprague-Dawley
PBS (pH 7.4), 0.5 mM EDTA, 50 ?L of the test compound
rat was obtained from the Laboratory Animal Service Center of
dissolved in DMSO, 0.1 mg/mL LDL and 20 mM AAPH. The
The Chinese University of Hong Kong. The blood sample,
condition of the hemin-induced lipid peroxidation of LDL was
collected in heparinized tubes, was centrifuged at 1500g for 10
37°C for 2 h in a final volume of 1 mL containing PBS (pH
min at 4°C. The erythrocytes collected were resuspended
7.4), 0.5 mM EDTA, 50 ?L of the test compound dissolved in
gently with 5 parts of physiological saline and centrifuged
DMSO, 0.1 mg/mL LDL and 1 ?M hemin. Trolox was used as
again at 1500g for 10 min at 4°C. The washing procedure was
the reference antioxidant compound in the assays.
repeated two more times. In the last wash, the centrifugal force
was lowered to 1000g. The erythrocytes were finally suspended
tBHP/hemin-Induced Lipid Peroxidation of
to a hematocrit of 0.20 in PBS.
Erythrocyte Membranes
The reaction was performed at 37°C for 4 h in a final
Preparation of Erythrocyte Membranes
volume of 1 mL containing PBS (pH 7.4), 50 ?L of the test
compound dissolved in DMSO, 0.05 mg/mL erythrocyte mem-
Approximately 100 mL of blood was obtained from a rabbit
branes, 0.5 mM tBHP and 1.5 ?M hemin. Trolox was used as
provided by the Laboratory Animal Service Center of The
the reference antioxidant compound in the assay.
Chinese University of Hong Kong. The blood was collected in
tubes with 60 mM EDTA in saline as anticoagulant (blood:
Determination of Lipid Peroxidation by
anticoagulant ? 9:1 v/v). After washing the erythrocytes, he-
Fluorescence Measurement of TBARS
moglobin-free erythrocyte membranes were prepared accord-
ing to a hypotonic lysis procedure [28]. Great care was taken to
Lipid peroxidation was determined by measuring the thio-
avoid contamination by potassium and phosphate ions which
barbituric acid-reactive substance (TBARS) formed [31]. The
JOURNAL OF THE AMERICAN COLLEGE OF NUTRITION
235

---

Antioxidant Mechanism of Phenolic Compounds
sample mixture containing oxidized LDL or erythrocyte mem-
with 1 mL of malachite green reagent [34] and 450 ?L of 1%
branes was mixed with one part of 20% trichloroacetic acid and
polyvinyl alcohol. Absorbance readings were compared against
one part of 0.8% thiobarbituric acid. The mixture was heated
a preconstructed calibration curve using potassium dihydrogen
for 1 h at 95°C. After cooling, it was centrifuged at 3000g for
phosphate as standard. The amount of Pi released within a
5 min. The supernatant was extracted with butan-1-ol and the
defined period (10 min) was calculated by subtracting the
organic layer was saved for subsequent TBARS determination.
reading of an aliquoted sample at zero-time from the one after
TBARS was measured fluorometrically at an excitation wave-
10 min incubation. The enzyme activity of Ca2?-ATPase or
length of 515 nm and an emission wavelength of 553 nm.
Na?/K?-ATPase in the membranes expressed as nmole Pi/mg/
TBARS formation in the presence of the test compound was
min was calculated by the difference in the activities deter-
expressed as a percentage of the control to indicate the extent
mined in assay buffers with and without the activation ions
of inhibition of lipid peroxidation.
(calcium ions for Ca2?-ATPase, and potassium ions for Na?/
K?-ATPase).
AAPH-Induced Hemolysis Assay
Sufhydryl Group Protection Assay
The assay was performed as described by Jimenez et al. [32]
with minor modifications. The reaction mixture was made up
The oxidation reaction was performed in a final volume of
by combining 500 ?L of rat erythrocyte suspension, 250 ?L of
800 ?L containing 50 mM Hepes buffer (pH 7.4), the test
the test compound dissolved initially in DMSO and diluted
compound, 0.3 mg/mL erythrocyte membrane, 0.5 mM tBHP
with PBS (pH 7.4), and 250 ?L of 400 mM AAPH in a 1.5 mL
and 10 ?M hemin. The mixture was incubated for 4 h at 37°C
microfuge tube. Trolox was used as the reference antioxidant
with gentle agitation. It was then centrifuged at 20,000g for 5
compound in the assay. The final concentration of DMSO was
min at 4°C. The supernatant was discarded and the membrane
2.5%. Preliminary experiments had shown that this concentra-
pellet was kept on ice. The erythrocyte membranes were resus-
tion of DMSO had minimal effects on the assay. All assay
pended in 400 ?L of 75 mM phosphate buffer (pH 7.4) with 2.7
mixtures were incubated at 37°C for 3 h with gentle rotation.
mM EDTA. Afterwards, 500 ?L of 10% sodium dodecyl
After the incubation period, aliquoted samples of each reaction
sulfate was added to denature the proteins and 100 ?L of 1 mM
mixture were mixed with 9 parts of PBS or distilled water
5,5?-dithiobis(2-nitrobenzoic acid) was also added for protein
(DW), respectively. All samples were mixed well and centri-
thiol determination [35]. Color was allowed to develop for 15
fuged at 1500g for 10 min. Absorbance readings of the super-
min before absorbance measurement at 412 nm. Net absor-
natants at 540 nm were measured. The percentage of erythro-
bance for each sample was obtained by subtracting the blank
cyte lysed in each sample was calculated according to the
(containing an equivalent amount of membrane proteins). The
following equation:
amount of free sulfhydryl groups in the membrane proteins of
the sample was determined from a standard curve constructed
Lysis % ? A540 in PBS/A540 in DW ? 100%
with a series of known amounts of reduced glutathione.
ATPase Protection Assay
Protein Assay
The oxidation reaction was performed in a final volume of
Protein assay was conducted according to the method of
800 ?L containing 50 mM sodium 4-(2-hydroxyethyl)pipera-
Lowry et al. [36] using bovine serum albumin as standard.
zine-1-ethanesulfonate (Hepes) buffer (pH 7.4), the test com-
pound, erythrocyte membranes (0.3 mg/mL for Ca2?-ATPase,
Statistical Analyses
and 1 mg/mL for Na?/K?-ATPase), 0.5 mM tBHP, and hemin
(2 ?M for Ca2?-ATPase, and 1.5 ?M for Na?/K?-ATPase).
Statistical analyses of the results were performed using
The mixture was incubated for 4 h at 37°C with gentle agita-
one-way ANOVA followed by Dunnett’s test. A P value
tion. It was then centrifuged at 20,000g for 5 min at 4°C. The
of ? 0.05 was considered statistically significant.
supernatant was discarded and the membrane pellet was kept
on ice. The erythrocyte membranes were resuspended in 450
?L of assay buffer according to Reinila et al. [33] except that
RESULTS
imidazole buffer was used instead of Tris buffer, with 0.01%
Triton X-100 (to release the ATPase from the membranes)
The antioxidant actions of the phenolic compounds were
followed by incubation at 37°C for 10 min before commence-
studied in a multitude of experimental systems. Results of the
ment of the ATPase assay. At suitable time intervals after the
initial screening are shown in Table 1. For those compounds
addition of ATP (1 mM final concentration), aliquots (50 ?L)
with positive results, dose-dependent studies were subse-
of samples were drawn for inorganic phosphate (Pi) determi-
quently performed to quantitate their potencies. The results are
nation. The amount of Pi in each sample was determined by
shown in Fig. 2, Fig. 3, Fig. 4, Fig. 5, Fig. 6, Fig. 7, Fig. 8. It
absorbance measurement at 650 nm after incubation for 30 min
was found that aloe-emodin is inactive in six out of the seven
236
VOL. 26, NO. 3

---

Antioxidant Mechanism of Phenolic Compounds
Table 1. Results of Initial Screening of Compounds on All Assays
Effectiveness, maximum concentration tested (?M)
Compound
A
B
C
D
E
F
G
6-Gingerol
?, 1000
?, 50
?, 500
?, 70
?, 1000
?, 1000
?, 1000
Rhapontin
?, 1000
?, 50
?, 500
?, 70
?, 1000
?, 2000
?, 1000
Aloe-emodin
?, 20
?, 20
?, 5
?, 20
?, 15
?, 15
?, 15
Barbaloin
?, 1000
?, 100
?, 500
?, 1000
?, 2000
?, 2000
?, 2000
Column A: inhibition of AAPH-induced lipid peroxidation in LDL; Column B: inhibition of hemin-induced lipid peroxidation in LDL; Column C: inhibition of
AAPH-induced hemolysis; Column D: inhibition of lipid peroxidation of erythrocyte membranes; Column E: protection of Ca2?-ATPase activity in erythrocyte
membranes; Column F: protection of Na?/K?-ATPase activity in erythrocyte membranes; Column G: protection of protein sulfhydryl groups in erythrocyte membranes.
The effectiveness of the test compounds is indicated by the following symbols: ?, active; ?, inactive.
Fig. 2. The inhibitory actions of Trolox, 6-gingerol, rhapontin and
barbaloin on AAPH-induced lipid peroxidation in LDL. Results are
expressed as a percentage of the control. Data represent mean values ?
S.D. from four independent experiments.
Fig. 4. The inhibitory actions of Trolox, 6-gingerol, rhapontin and
barbaloin on AAPH-induced hemolysis. Results are expressed as a
percentage of erythrocytes lysed. Data represent mean values ? S.D.
from four independent experiments.
Fig. 3. The inhibitory actions of Trolox, 6-gingerol, rhapontin and
barbaloin on hemin-induced lipid peroxidation in LDL. Results are
expressed as a percentage of the control. Data represent mean values ?
Fig. 5. The inhibitory actions of Trolox, 6-gingerol, rhapontin and
S.D. from four independent experiments.
barbaloin on lipid peroxidation of erythrocyte membranes. Results are
expressed as a percentage of the control. Data represent mean values ?
S.D. from four independent experiments.
assays (Table 1). In the hemin-induced LDL lipid peroxidation
system (Column B in Table 1), the only assay showing some
activity for aloe-emodin, the effectiveness of this compound
AAPH-induced LDL lipid peroxidation assay, compared
was found to be much lower than those of the other compounds
against Trolox as the positive control. Fig. 3 shows the results
tested, being more than 100 times less potent than Trolox (data
in which LDL oxidation was induced by an alternative method,
not shown).
the hemin system. It was observed that the effective concen-
The inhibitory actions of the active phenolic compounds
trations of all the test compounds in the hemin system were
against LDL oxidation were studied. Fig. 2 shows the antiox-
about two orders of magnitude lower than those in the AAPH
idant actions of 6-gingerol, rhapontin and barbaloin in the
system. This is probably due to the low level of endogenous
JOURNAL OF THE AMERICAN COLLEGE OF NUTRITION
237

---

Antioxidant Mechanism of Phenolic Compounds
Fig. 6. The protective actions of barbaloin on Ca2?-ATPase activity
Fig. 8. The protective actions of barbaloin on sulfhydryl groups against
against oxidative damage in erythrocyte membranes. Results are ex-
oxidative damage in erythrocyte membranes. Results are expressed as
pressed as enzyme unit in nmol Pi/mg/min. Data represent mean
nmol SH/mg protein. Data represent mean values ? S.D. from four
values ? S.D. from four independent experiments. C: control without
independent experiments. C: control without the test compound, T: 1
the test compound, T: 1 mM Trolox. Results are compared against the
mM Trolox. Results are compared against the control with oxidative
control with oxidative damage by one-way ANOVA followed by
damage by one-way ANOVA followed by Dunnett’s test. *, P ?0.05;
Dunnett’s test. *, P ?0.05.
***, P ?0.001.
actions of the phenolic compounds on erythrocyte membranes
were also investigated (Fig. 5). In order to compare the anti-
oxidant potencies of the phenolic compounds in these four
assays, the IC
value of each compound was determined and
50
further transformed into the relative potency as compared with
Trolox, and the results are listed in Table 2. These results show
that 6-gingerol and rhapontin exhibit stronger antioxidant ac-
tions than Trolox in inhibiting lipid peroxidation in all the
systems studied while barbaloin possesses a much weaker
effect. 6-Gingerol is about two times stronger than rhapontin in
inhibiting hemin-induced LDL lipid peroxidation and the he-
min/tBHP-induced erythrocyte membrane lipid peroxidation.
On the other hand however, 6-gingerol exhibits a weaker effect
than rhapontin in inhibiting AAPH-induced LDL lipid peroxi-
dation. Rhapontin and barbaloin are comparable in protecting
erythrocytes from AAPH-induced hemolysis. They are almost
two times more potent than Trolox while 6-gingerol was only
Fig. 7. The protective actions of rhapontin on Na?/K?-ATPase activity
about half as potent as Trolox in this assay. These results
against oxidative damage in erythrocyte membranes. Results are ex-
suggest that 6-gingerol exhibits a strong antioxidant action
pressed as enzyme unit in nmol Pi/mg/min. Data represent mean
against lipid peroxidation but this compound possesses little
values ? S.D. from four independent experiments. C: control without
activity in scavenging AAPH-derived radicals. The fact that
the test compound, T: 1 mM Trolox. Results are compared against the
rhapontin exhibits stronger antioxidant actions than 6-gingerol
control with oxidative damage by one-way ANOVA followed by
in all the AAPH-induced oxidation systems suggests that it is a
Dunnett’s test. *, P ?0.05; ***, P ?0.001.
stronger AAPH-derived radical scavenger than 6-gingerol.
lipid peroxides in LDL, and therefore a small amount of anti-
Comparing the IC
values of the phenolic compounds in
50
oxidant is sufficient to inhibit lipid peroxidation. The antioxi-
assays A, C and D in Table 2, it is interesting to note that in the
dant actions of the test compounds on AAPH-induced hemo-
AAPH-induced hemolysis inhibition assay (Column C in Table
lysis are shown in Fig. 4. The lipid peroxidation inhibitory
2), the IC
value of 6-gingerol is about 5 times higher than that
50
238
VOL. 26, NO. 3

---

Antioxidant Mechanism of Phenolic Compounds
Table 2. Relative Effectiveness of 6-Gingerol, Rhapontin and Barbaloin in Inhibiting Lipid Peroxidation and Hemolysis as
Compared with Trolox
IC
(?M) (Relative potency)
50
Compound
A
B
C
D
Trolox
40.9 (1.00)
0.195 (1.00)
59.2 (1.00)
224.4 (1.00)
6-Gingerol
19.1 (2.14)
0.054 (3.61)
95.3 (0.62)
20.7 (10.84)
Rhapontin
14.1 (2.90)
0.104 (1.87)
31.9 (1.86)
38.3 (5.86)
Barbaloin
142.4 (0.29)
2.845 (0.07)
31.7 (1.87)
71.0 (3.16)
The IC
values of the phenolic compounds and Trolox in different assays (Column A: inhibition of AAPH-induced lipid peroxidation in LDL; Column B: inhibition of
50
hemin-induced lipid peroxidation in LDL; Column C: inhibition of AAPH-induced hemolysis; Column D: inhibition of lipid peroxidation of erythrocyte membranes) were
calculated from the dose-dependence curves in Fig. 2 to Fig. 5 by Sigmaplot. The relative potency (values shown in parentheses) for each compound was calculated by
the IC
of Trolox over the IC
of the test compound.
50
50
in assays A and D. In the case of barbaloin, however, the
and protein sulfhydryl groups, and of rhapontin towards Na?/
reverse is true in that assay C exhibits the lowest IC
value. A
K?-ATPase, were found to be comparable to that of Trolox
50
closer look at Fig. 4 shows that there is a dramatic decrease in
(Table 3). The percentage recovery at 1 mM instead of the IC50
the percentage of hemolysis from 80 ?M of 6-gingerol onwards
value was used to calculate the relative potency because the
but there is virtually no protection below 80 ?M. This result
maximal protective effects afforded by Trolox in these assays
suggests that 6-gingerol does not directly eliminate AAPH-
was observed at 1 mM in the initial studies and the percentage
derived radicals in aqueous solution nor does it prevent the
recoveries afforded by the test compounds in some of these
initial attack of AAPH on the erythrocyte membranes. It prob-
assays could not reach 50%. From the results, it could be
ably acts by protecting the erythrocytes from lysis by terminat-
concluded that protection of protein sulfhydryl groups may be
ing the lipid peroxidation chain reactions. For barbaloin, the
related to the protective actions of barbaloin on Ca2?-ATPase
lower IC
value in the hemolysis inhibition assay C than the
but not the protective actions of rhapontin on Na?/K?-ATPase.
50
lipid peroxidation inhibition assays A and D (Table 2) suggests
The ineffectiveness of 6-gingerol in these three assays also
that antioxidant mechanisms other than inhibition of lipid per-
suggests that it is a poor tBHP-derived radical scavenger.
oxidation might contribute to the protection of the erythrocytes
from lysis. Considering the lower potency of barbaloin in
inhibiting lipid peroxidation, it is possible that protection of
membrane proteins is an important mechanism for the action of
DISCUSSION
barbaloin in inhibiting hemolysis.
The protective actions of these phenolic compounds on
The ortho-hydroxyl-methoxyl structure in the phenyl rings
membrane proteins were thus studied by the ATPase protection
of rhapontin and 6-gingerol probably contributes to the high
assays and sulfhydryl group protection assay. The results show
antioxidant activities of the two phenolic compounds, due to
that barbaloin is effective in protecting Ca2?-ATPase (Fig. 6)
the increase in stability of the phenoxyl radical formed by
and protein sulfhydryl groups (Fig. 8) against oxidative attack
electron delocalization to the adjacent methoxyl group. 6-Gin-
by tBHP/hemin while other phenolic compounds tested have no
gerol is a hydrophobic molecule due to the presence of a long
effects (Columns E and G in Table 1). Moreover, rhapontin
alkyl chain. This increases its solubility in lipid and the chance
protects Na?/K?-ATPase from oxidative attack (Fig. 7) in the
of encountering lipid peroxyl radicals in LDL or erythrocyte
absence of sulfhydryl group protection (Column F in Table 1).
membranes, thus explaining its strong chain-breaking action in
The protective potencies of barbaloin towards Ca2?-ATPase
lipid. However, its lipophilicity may also be the cause of its
Table 3. Relative Effectiveness of Rhapontin and Barbaloin in Protecting ATPases and Protein Sulfhydryl Groups in Erythrocyte
Membranes as Compared with Trolox
% Recovery at 1 mM of compound (Relative potency)
Compound
E
F
G
Trolox
46.5% (1.00)
55.1% (1.00)
52.1% (1.00)
Rhapontin
N.A.
35.4% (0.64)
N.A.
Barbaloin
42.3% (0.91)
N.A.
86.1% (1.65)
The percentage recovery in Ca2?-ATPase activity (Column E), Na?/K?-ATPase activity (Column F) or free sulfhydryl group content (Column G) by 1 mM of the phenolic
compounds or Trolox was calculated by regarding the difference in the determinants of the control with and without oxidant treatment to be equivalent to 100% recovery.
The relative potency (values shown in parentheses) for each compound was calculated by the percentage recovery of the test compound over the percentage recovery of
Trolox. N.A. stands for not applicable as the compound is inactive in that assay.
JOURNAL OF THE AMERICAN COLLEGE OF NUTRITION
239

---

Antioxidant Mechanism of Phenolic Compounds
poor scavenging effect on the highly water-soluble AAPH-
radical scavenger in aqueous solution is particularly effective in
derived peroxyl radical (Fig. 9) as this radical contains two
the protection of membranes since its presence could restrict
amine groups. On the other hand the glucosyl moiety increases
free radical damage on lipids and proteins. The resultant pres-
the hydrophilicity of rhapontin, thus explaining the stronger
ervation of membrane structure and function is essential for
scavenging activity of rhapontin on the AAPH-derived peroxyl
maintaining membrane fluidity and flexibility as well as ionic
radicals in solution than 6-gingerol.
balance between the intracellular and extracellular compart-
Lipid peroxidation could be initiated by free radical gener-
ments. These processes are crucial for the survival of a cell.
ating compounds/systems such as AAPH, tBHP, Fe2?/H O
ATPase inhibition by oxidative attack has been suggested to
2
2
and Fe3?/ascorbate. At the AAPH concentration used in the
involve mechanisms including protein crosslinking [23], sulf-
hemolysis assay, there would be an abundant supply of lipid
hydryl group oxidation [38,39] and lipid peroxidation [38,40].
peroxidation initiators in the system. As a result, the lipid
From our data, it is conceivable that protection of protein
peroxidation chain reactions can still propagate if chain-termi-
sulfhydryl groups could explain, in part at least, the protective
nators such as 6-gingerol are present below a certain threshold
actions of barbaloin on Ca2?-ATPase. However, our data also
in the erythrocyte membranes. This is reflected in the require-
show that while barbaloin at 1 mM recovers 86% of the
ment of a critical concentration of 6-gingerol to produce an
sulfhydryl groups in erythrocyte membranes oxidized by 0.5
inhibitory action on hemolysis (Fig. 4). On the other hand,
mM tBHP and 10 ?M hemin (Fig. 8), the same concentration
antioxidants that could scavenge AAPH-derived radicals in
of barbaloin could only recover 42% of the Ca2?-ATPase
aqueous phase such as rhapontin could reduce the degree of
activity (Fig. 6) and is completely ineffective in recovering
lipid peroxidation by reducing the number of chain initiators in
Na?/K?-ATPase activity (Column F in Table 1) in the pres-
the first place. Thus, a dose-dependent inhibition of hemolysis
ence of 0.5 mM tBHP and 1.5–2 ?M hemin. These results
was evident starting at a low concentration (Fig. 4).
suggest that mechanisms other than sulfhydryl group oxidation
The only structural difference between aloe-emodin and
are involved in the tBHP/hemin-induced inactivation of AT-
barbaloin is the presence of an extra glucosyl moiety in barba-
Pases as this part of inactivation is irrecoverable by barbaloin.
loin. The huge difference in their antioxidant activities suggests
Of particular interest is the finding that rhapontin protects
that substitution of this glucosyl group is important for the
Na?/K?-ATPase against free radical attack independent of
observed antioxidant activities of barbaloin. Solubility may be
sulfhydryl group protection (Table 3). In addition, although
one of the factors. The absence of a glucosyl group in aloe-
rhapontin inhibits TBARS formation in erythrocyte membranes
emodin increases the difficulty of dissolving this compound in
at micromolar concentrations (Fig. 5), its protective effect on
an aqueous medium. It is possible that the antioxidant actions
Na?/K?-ATPase against the same tBHP/hemin system could
of aloe-emodin were not fully revealed because of the limita-
only be observed at the millimolar range. While we cannot rule
tion imposed by its low water solubility. Of particular interest
out the possibility that the enzyme could be inhibited by lipid-
is the identification of 10-hydroxyaloins A and B as the main in
derived oxidants in the absence of TBARS formation, these
vitro oxidation products of barbaloin diastereomers [37]. This
results suggest that mechanisms other than sulfhydryl group
suggests that the C-10-glucosylation in barbaloin renders the
protection and lipid peroxidation inhibition might be involved
molecule a distinct electron-donating site as compared with its
in the protective effect of rhapontin on Na?/K?-ATPase.
parent molecule aloe-emodin which has a C-10-carbonyl group
Oral administration of ginger, aloe and rhubarb have been
instead. This may be the reason for a different and, perhaps, an
reported to produce beneficial effects on experimental animals
enhanced free radical scavenging property for barbaloin as
in vivo. Ginger has been demonstrated to decrease levels of
compared with aloe-emodin as indicated in the present study.
lipid peroxidation in rabbits with experimental atherosclerosis
Barbaloin is shown in the present study to inhibit lipid
[41], in Apo E-deficient mice [42], and in hyperlipidemic rats
peroxidation (Fig. 5) and to afford protection of protein sulf-
[43]. Aloe vera gel extract has been found to exhibit various
hydryl groups (Fig. 8) and Ca2?-ATPase (Fig. 6) in erythrocyte
antioxidant effects in streptozotocin-induced diabetic animals
membranes challenged with tBHP-derived free radicals. This
[44 – 46] and kainic acid-induced neurotoxicity in mice [47].
wide spectrum of antioxidant actions may provide the basis for
Oral treatment with Aloe vera leaf juice also significantly
its high potency in inhibiting AAPH-induced hemolysis (assay
decreased malondialdehyde formation, improved superoxide
C in Table 2). It is possible that the scavenging of tBHP- or
dismutase and catalase activities, and reduced glutathione lev-
AAPH-derived radicals in the aqueous phase may be involved
els in various tissues of gamma-irradiated rats [48]. Aloe vera
in the protective mechanisms of barbaloin. An efficient free
leaf pulp extract could also increase the hepatic enzyme activ-
ities of glutathione S-transferase, DT-diaphorase, superoxide
dismutase, catalase, glutathione peroxidase and glutathione re-
ductase, and decrease hepatic malondialdehyde formation and
lactate dehydrogenase activity in normal mice [49]. A rhubarb
preparation has been shown to increase the enzyme activities of
Fig. 9. Chemical structure of peroxyl radical derived from AAPH.
superoxide dismutase, catalase and glutathione peroxidase in
240
VOL. 26, NO. 3

---

Antioxidant Mechanism of Phenolic Compounds
erythrocytes and to reduce lipid peroxidation in the brain, liver
4. Croft KD: Antioxidant effects of plant phenolic compounds. In
and whole blood of aging mice [50]. Results of the present
Basu TK, Temple NJ, Garg ML (eds): “Antioxidants in Human
investigation and other in vitro studies [51–57] have attributed
Health and Diseases,” New York: CABI Publishing, pp 109–121,
these antioxidant activities to the phenolic compounds from
1999.
these plants. Ginger, aloe and rhubarb are regarded as func-
5. Heim KE, Tagliaferro AR, Bobilya DJ: Flavonoid antioxidants:
tional foods in China as well as in some other countries. Their
chemistry, metabolism and structure-activity relationships. J Nutr
Biochem 13:572–584, 2002.
consumption on a long-term basis is considered beneficial to
6. Gao Z, Huang K, Yang X, Xu H: Free radical scavenging and
human health. The present study illuminates a practical guide in
antioxidant activities of flavonoids extracted from the radix of
nutritional medi