Chemical studies on antioxidant mechanism of garcinol: Analysis of radical reaction products of garcinol and their antitumor activities

Article abstract of DOI:10.1016/S0040-4020(01)01015-8

Garcinol (1), a polyisoprenylated benzophenone, purified from Garcinia indica fruit rind, displays antioxidant properties and is thought to act as an antioxidant in biological systems. However, the mechanisms of its antioxidant reactions remain unknown. The objective of this study was to characterize the reaction products of garcinol with a stable radical, 2,2-diphenyl-1-picrylhydrazyl (DPPH). Structural elucidation of these products can provide insights into specific mechanisms of antioxidant reactions. Two major reaction products, GDPPH-1 (2) and GDPPH-2 (3), were isolated and identified for the first time. Their structures were determined on the basis of detailed high field 1D and 2D spectral analysis. The identification of these products provides the first unambiguous proof that the principal sites of antioxidant reactions are on the 1,3-diketone and the phenolic ring part of 1. The induction of apoptosis in human leukemia HL-60 cells, the inhibition of NO generation, the effects on the activity of MMP, and the inhibitory effects on H₂O₂ production of TPA-stimulate HL-60 cells by 1 and its two oxidant products (2 and 3) were investigated and would also be discussed in this paper.

Full text of DOI:10.1016/S0040-4020(01)01015-8

TETRAHEDRON
Pergamon
Tetrahedron 57 02001) 9931±9938
Chemical studies on antioxidant mechanism of garcinol:
analysis of radical reaction products of garcinol and their
antitumor activities
Shengmin Sang,a Min-Hsiung Pan,b Xiaofang Cheng,c Naisheng Bai,a Ruth E. Stark,c
Robert T. Rosen,a Shoei-Yn Lin-Shiau,b Jen-Kun Linb and Chi-Tang Hoa,p
aDepartment of Food Science, Center for Advanced Food Technology, Rutgers University, 65 Dudley Road,
NewBrunswick, NJ 08901-8520, USA
bInstitute of Biochemistry and Toxicology, College of Medicine, National Taiwan University, 1 Section 1, Jen-ai Road, Taipei, Taiwan, ROC
cDepartment of Chemistry, College of Staten Island, City University of NewYork, NewYork, NY 10314, USA
Received 7 August 2001; accepted 18 September 2001
AbstractÐGarcinol 01), a polyisoprenylated benzophenone, puri®ed from Garcinia indica fruit rind, displays antioxidant properties and is
thought to act as an antioxidant in biological systems. However, the mechanisms of its antioxidant reactions remain unknown. The objective
of this study was to characterize the reaction products of garcinol with a stable radical, 2,2-diphenyl-1-picrylhydrazyl 0DPPH). Structural
elucidation of these products can provide insights into speci®c mechanisms of antioxidant reactions. Two major reaction products, GDPPH-1
02) and GDPPH-2 03), were isolated and identi®ed for the ®rst time. Their structures were determined on the basis of detailed high ®eld 1D
and 2D spectral analysis. The identi®cation of these products provides the ®rst unambiguous proof that the principal sites of antioxidant
reactions are on the 1,3-diketone and the phenolic ring part of 1. The induction of apoptosis in human leukemia HL-60 cells, the inhibition of
NO generation, the effects on the activity of MMP, and the inhibitory effects on H2O2 production of TPA-stimulate HL-60 cells by 1 and its
two oxidant products 02 and 3) were investigated and would also be discussed in this paper. q 2001 Elsevier Science Ltd. All rights reserved.
1. Introduction
relative importance of many competing biological reactions.
Humans have evolved with antioxidant systems to protect
against free radicals. These systems include some anti-
oxidants produced in the body 0endogenous) and others
obtained from the diet 0exogenous). Over the past decade,
evidence has been accumulated that plant polyphenols are
an important class of defense antioxidants. These
compounds are widespread virtually in all plant foods,
often at high levels, and include phenols, phenolic acids,
¯avonoids, tannins, and lignans.
Free radicals and reactive oxygen species play an important
physiological role, they are essential for production of
energy, synthesis of biologically essential compounds, and
phagocytosis, a critical process of immune system.1 They
also play a vital role in signal transduction. On the other
hand, there is now increasing evidence that these reactive
oxygen species may play a causative role in a variety of
diseases including heart disease, aging and cancer.2,3 Free
radicals and reactive oxygen species attack lipids, sugars,
proteins, and DNA and induce their oxidation, which may
result in oxidative damage such as membrane dysfunction,
protein modi®cation, enzyme inactivation, and breakage of
DNA strands and modi®cation of its bases.1 Consequently,
the role of antioxidants has received increased attention
during the past decade. It is important to elucidate the
mechanisms and dynamics of the oxidative damage in
order to understand its biological signi®cance and develop
strategies for its prevention. Both active oxygen species and
antioxidants are double-edged swords and the balance of
their bene®cial and toxic effects is determined by the
Garcinol 01) is a polyisoprenylated benzophenone deriva-
tive from Garcinia indica and other species.4±8 The dried
rind of G. indica 0cv. Kokum) which is used as a garnish for
curry and in some of the folklore medicine in India contains
2±3% garcinol.4,5 Garcinol is structurally similar to a well-
known antioxidant-curcumin, which contains both phenolic
hydroxyl groups and a b-diketone moiety. Garcinol has
been reported to possess antibiotic activities,8 antiulcer
activity,10 suppressed colonic aberrant crypt foci 0ACF)
formation,11 and induction of apoptosis through cytochrome
c release and activation of caspases in human leukemia
HL-60 cells.12 It also showed strong antioxidant activity.
In the H2O2±NaOH±DMSO system, garcinol suppressed
superoxide anion, hydroxyl radical, and methyl radical; in
the Fenton reaction system, garcinol suppressed hydroxyl
Keywords: garcinol; antioxidant mechanism; antitumor activities.
Corresponding author. Tel.: 11-732-932-9611x235; fax: 11-732-932-
6776; e-mail: ho@aesop.rutgers.edu
p
0040±4020/01/$ - see front matter q 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020001)01015-8

9932
S. Sang et al. / Tetrahedron 57 62001) 9931±9938
radical more strongly than dl-a-tocopherol; in the
hypoxanthine±xanthine oxidase system, emulsi®ed garcinol
suppressed superoxide anion to almost the same extent as
dl-a-tocopherol by weight.9 It also showed nearly three
times greater 2,2-diphenyl-1-picrylhydrazyl 0DPPH) free
radical scavenging activity than dl-a-tocopherol by
weight.10 However, the antioxidant mechanisms of garcinol
remain unclear. To evaluate the antioxidant mechanisms of
garcinol, we reacted garcinol with DPPH. The DPPH
system is a stable radical-generating procedure.13 Because
it can accommodate a large number of samples in a short
period, and is sensitive enough to detect active principles at
low concentrations, it was used in the present study. The
antioxidant process of this reaction is thought to be divided
into the following two stages:
inhibitory effects on H2O2 production of TPA-stimulate
HL-60 cells by 1 and its two oxidation products 02 and 3)
have also been evaluated.
2. Results and discussion
2.1. Structural elucidation of two oxidation products !2
and 3)
Reaction of garcinol 01) with DPPH was carried out and two
major oxidation products, GDPPH-1 02) and GDPPH-2 03),
were isolated and identi®ed on the basis of their spectral
data.
DPPHz 1 AH O DPPH±H 1 Az
ꢀ1†
ꢀ2†
Compound 2, a yellow amorphous solid, was assigned the
molecular formula of C38H48O6 determined by positive-ion
HRFAB-MS 0[M1H]1 at m/z 601.3527), as well as from its
13C NMR data. The molecular formula indicated 15 degrees
of unsaturation, which showed that 2 had one more
Az 1 Xz ! nonradical materials
AH is the antioxidant, Az is the antioxidant radical, and Xz is
another radical species or the same species as Az.1,14
Although the ®rst stage is a reversible process, the second
stage is irreversible and must produce stable radical
terminated compounds. Structural information about these
nonradical products would afford important contributions to
the antioxidant mechanism studies.
unsaturation than 1. The H NMR spectrum of 2 showed
1
the presence of two singlets aromatic protons at d 6.96 and
8.44 ppm, respectively; three isopropylidine groups 0three
one-hydrogen triplets at d 5.00, t, Jˆ6.0 Hz, 4.76, t,
Jˆ6.0 Hz, and 4.66, t, Jˆ6.0 Hz, and six singlet methyl
groups at d 1.72, 1.64, 1.58, 1.54, 1.44, and 1.09, respec-
tively), one isopropenyl group 0two singlets of 2H at d 4.41
and 4.31, together with a singlet methyl signal at d 1.56),
and two methyl groups on a saturated carbon 0two singlets
of 3H each at d 1.29 and 1.22), in addition to methylene and
methine protons 0a complex multiplet of 12H in the regions
In this paper, we will report the isolation and structure
elucidation of two major oxidation products, GDPPH-1
and GDPPH-2 02 and 3) formed by the reaction of garcinol
01) with the stable radical DPPH. The induction of apoptosis
in human leukemia HL-60 cells, the inhibition of NO
generation, the effects on the activity of MMP, and the
1
of d 3.00±1.45). Thus the signi®cant difference in the H
NMR spectrum of 2 compared to 1 is the loss of one
aromatic proton. As we mentioned earlier, 2 had one more
Figure 1. Structures of Garcinol 01) and its oxidant products GDPPH-1 02) and GDPPH-2 03).

S. Sang et al. / Tetrahedron 57 62001) 9931±9938
9933
unsaturation than 1. All of these mean that there was one
more ring in 2 than in 1. The above ®ndings were in agree-
ment with the 13C NMR spectrum of 2. It also showed the
presence of three isopropylidine groups 0three methine
carbons of trisubstituted ole®nic groups at d 123.8, 123.0,
and 118.8, and six methyl groups at d 17.9, 18.0, 18.5, 25.9,
26.0, and 26.1), one isopropenyl group 0d 113.9 for a term-
inal methylene carbon, and d 18.2 for a methyl signal), and
two methyl groups on a saturated carbon 0d 23.5 and 27.4).
Two methine carbons for the aromatic ring at d 102.5 and
110.9; and three oxygen-substituted aromatic carbons at d
144.7, 150.7, 152.9, which indicated that one aromatic
proton of garcinol was substituted by oxygen in 2. Since
both of the aromatic protons in 2 were singlets, it must be
H-16 that was substituted by oxygen. Furthermore, in the
13C spectral data, one of the enolized 1,3-diketone in 1 was
changed from d 194.0 0or 195.2) to d 175.6 in 2. The 13C
spectral data of the carbonyl 0C-10) was also changed from
d 199.1 in garcinol to d 173.1 in 2. All of these indicated
that the oxygen of one of the enolized 1,3-diketone was
attached at C-16. The HMBC correlation between Cd175.6
and H-17 0d 2.82 and 2.89), Cd193.4 and H-7 0d 1.99 and
2.44), and H-29 0d1.79 and 2.28); Cd173.5 and H-12 0d 8.44),
and H-15 0d 6.96) 0Fig. 2) suggested that d 175.6 was
assigned to C-1, d 193.4 was to C-3, and d 173.5 was to
C-10, respectively. This con®rmed that the carbonyl at C-3
was enolized and the oxygen was attached to C-16. Thus,
the structure of 2 was deduced as shown 0Fig. 1). The
complete interpretation of the NMR data was based on the
results of COSY, TOCSY, HMQC and HMBC.
25.8, and 25.9), one isopropenyl group 0d 112.2 ppm for a
terminal methylene carbon, and d 18.3 for a methyl signal),
and two methyl groups on a saturated carbon 0d 27.3 and
29.9), two methine carbons for the aromatic ring at d 102.5
and 110.6; and three oxygen-substituted aromatic carbons at
d 144.6, 150.2, 152.8. The data of C1±C3 0d 114.5, 178.2,
and 193.5) and C10 0d 173.7) were also similar with those of
2. All of these indicated that 3 was an isomer of 2. This was
con®rmed by the cross peaks in the HMBC spectrum. In the
HMBC of 3, the correlations between Cd193.5 and H-17 0d
2.59 and 2.71), Cd178.2 and H-7 0d 2.15 and 2.38), and H-29
02.01 and 2.24); Cd173.5 and H-12 0d 8.38), and H-15 0d
6.91) 0Fig. 2) suggested that d 178.2 was assigned to C-1,
d 193.5 was to C-3, and d 173.7 was to C-10, respectively.
Thus, the structure of 3 was identi®ed as shown 0Fig. 1). For
compound 2 the carbonyl at C-3 was enolized and the
oxygen was attached to C-16, while for compound 3 the
carbonyl at C-1 was enolized and the oxygen was attached
to C-16. The complete interpretation of the NMR data was
based on the results of COSY, TOCSY, HMQC and HMBC.
2.2. Induction of apoptosis by garcinol !1), GDPPH-1 !2),
and GDPPH-2 !3) in human leukemia HL-60 cells
A sub-G1 0sub-2N) DNA peak, which has been suggested to
be the apoptotic DNA was detected in cells that were treated
with 1±3, washed, and stained with propidium iodide. As
shown in Fig. 3, the percentage of apoptotic HL-60 cells
were 9.85, 66.34, and 66.33% after 12 h of incubation with
1±3 010 mM), respectively. Among them, 2 appeared to be
more potent, with an EC50 of 8.4 mM and induced dose-
dependent on the cell apoptosis. When the cells were treated
with the same concentration 020 mM) of compounds, the
apoptotic potency was the same. These data are consistent
with DNA fragmentation.
Compound 3 was isolated as a yellow amorphous solid. The
positive-ion HRFAB-MS of 3 display a molecular ion peak
at m/z [M1H]1 601.3527, supporting a molecular formula
of C38H48O6, which was the same as the molecular formula
of 2. The 1H NMR spectrum of 3 displayed a signal pattern
similar to that of 2. It also showed the signals of two singlet
aromatic protons at d 6.91 and 8.38, respectively; three
isopropylidine groups 0three one-hydrogen triplets at d
4.97, t, Jˆ6.0 Hz, 4.75, t, Jˆ6.0 Hz, and 4.64, t,
Jˆ6.0 Hz, and six singlet methyl groups at d 1.72, 1.62,
1.56, 1.49, 1.47 and 0.95, respectively), one isopropenyl
group 0two singlets of 2H at d 4.11 and 4.14, together
with a singlet methyl signal at d 1.54), and two methyl
groups on a saturated carbon 0two singlets of 3H each at d
1.14 and 1.03). The 13C NMR spectrum of 3 also indicated
the presence of three isopropylidine groups 0three methane
carbons of trisubstituted ole®nic groups at d 123.7, 122.5,
and 119.7, and six methyl groups at d 17.7, 18.1, 18.5, 22.6,
2.3. Inhibition of NO generation by garcinol !1),
GDPPH-1 !2), and GDPPH-2 !3)
Compounds 1±3 were examined to determine whether
they affect NO production in macrophages activated with
LPS for 24 h. Of these compounds tested, 1 inhibited
LPS-stimulated NO generation the most strongly; however,
1±3 all markedly reduced NO generation in a concentration-
dependent manner 0Fig. 4). At concentrations of 2.5, 5, and
10 mM, 1 inhibited NO generation by 49, 87, and 92%,
respectively. At concentrations of 2.5, 5, and 10 mM, 2
inhibited NO generation by 45, 75, and 78%, respectively.
At concentrations of 2.5, 5, and 10 mM, 3 inhibited NO
Figure 2. Signi®cant HMBC 0H!C) correlations of GDPPH-1 02) and GDPPH-2 03).

9934
S. Sang et al. / Tetrahedron 57 62001) 9931±9938
Figure 3. Inducing apoptosis of HL-60 cells by compounds 1±3 0garcinol, GDPPH-1, and GDPPH-2). Cells were harvested 12 h after treatment, and apoptosis
was quanti®ed by ¯ow cytometry. The method of ¯ow cytometry used is described under Experimental.
generation by 19, 62, and 73%, respectively. Inhibition of
NO production was not due to cytotoxicity, as determined
with the trypan blue exclusion assay.
2.5. Effects of garcinol !1), GDPPH-1 !2), and GDPPH-2
!3) on H2O2 production of TPA-stimulated HL-60 cells
Studies of the effects of 1±3 on H2O2 production by TPA-
stimulate HL-60 cells were analyzed by ¯ow cytometry.
The results show that garcinol was the most potent inhibitor
of TPA-induced H2O2 production in HL-60 cells at 5 mM
0Fig. 6). The inhibitory potency was estimated in the follow-
ing order: 1.3.2 at 5 mM. The effects of 1±3 on H2O2
production in TPA-stimulated HL-60 cells are possibly
involved in the signal transduction of the PKC activation
pathway that is stimulated by TPA.
2.4. Effects of garcinol !1), GDPPH-1 !2), and GDPPH-2
!3) on the activity of MMP
MMPs expression has been correlated with metastatic
potential. We determined the effects of 1±3 on the activity
of MMP by zymographic analysis. The same number of
cells were seeded, and the MMP activity of conditioned
medium from the cells were measured with three
compounds for 24 h. As shown in Fig. 5, the inhibitory
potency was estimated in the following order: 1.3.2
at 2 mM. The inhibition of MMP activity was not due
to toxicity, as determined by trypan blue exclusion
assay.
2.6. Discussions
Two compounds 02 and 3) were isolated from the radical
reaction products of garcinol 01) and DPPH. From our
elucidation of the chemical structures of these two
Figure 4. Effects of compounds 1±3 0garcinol, GDPPH-1, and GDPPH-2) on LPS-induced nitrite production in RAW 264.7 Cells. The cells were treated with
different concentrations of compounds and LPS 0100 mg mL21) for 24 h. Nitrite were determined by Griess reaction, as described in Experimental. The values
were expressed as means^SE of triplicate tests.

S. Sang et al. / Tetrahedron 57 62001) 9931±9938
9935
Figure 5. Gelatin-based zymography 0MMP activity) of the culture medium of B16 cells. B16 cells were treated with various concentrations of compounds 1±
3 0garcinol, GDPPH-1, and GDPPH-2) for 24 h, and MMP activity of culture medium were determined by zymographic assay. The values were expressed as
means^SE of triplicate tests.
compounds, we proposed the antioxidant mechanism of
garcinol as illustrated in Fig. 7. As shown in Fig. 7, garcinol
donated an H atom from the hydroxyl group of the enolized
1,3-diketone to form a pair of resonance. If the H atom of
the hydroxyl group at C-3 was donated, 2 would be formed,
and 3 was given when the H atom of the hydroxyl group at
C-1 was trapped by the DPPH radical.
inhibitory effects on H2O2 production of TPA-stimulate
HL-60 cells indicated that like 1, both 2 and 3 showed
inhibitory effects on these assay. More importantly, 2
appeared to be more potent than 1, with an EC50 of
8.4 mM and induced dose-dependent on the cell apoptosis.
The potency of these compounds in apotosis-induction may
vary with the different cell lines. These ®ndings might
suggest possible chemopreventive ability of garcinol and
its two oxidation products. Thus, analysis of these products
could prove a unique tool for assessing the contribution of
antioxidant reactions to the disease preventive effects of
garcinol.
The antitumor test of garcinol and its two reaction products
0GDPPH-1 and GDPPH-2) on the induction of apoptosis in
human leukemia HL-60 cells, the inhibition of NO
generation, the effects on the activity of MMP, and the
Figure 6. Inhibition of TPA-stimulated H2O2 production in HL-60 cells. Studies of the effects of compounds 1±3 0garcinol, GDPPH-1, and GDPPH-2) on
H2O2 production by TPA-stimulated HL-60 cells were performed as described in the Experimental.

9936
S. Sang et al. / Tetrahedron 57 62001) 9931±9938
Figure 7. Proposed mechanism for the formation of oxidation product GDPPH-1 02) and GDPPH-2 03).
Our future work will focus on whether reaction products of
garcinol and free radicals in lipid bilayers in vitro and in
biological membranes in vitro and in vivo are similar to
those observed in this study in a homogeneous solution
system. We are also very interested in the studies that it is
garcinol or its oxidation products or both of them that play
the functional role in the biological systems.
Table 1. dH 0600 MHz) NMR spectral data of compounds 2, 3 and dC
0150 MHz) NMR spectral data of compounds 1±3 0CD3OD) 0d in ppm, J
in Hz)
1, dC
2
3
dC
dH
dC
dH
3. Experimental
1
2
3
4
5
6
7
194.0 s 193.4 s
116.0 s 119.8 s
195.2 s 175.6 s
178.2 s
114.5 s
193.5 s
71.7 s
50.1 s
46.3 d
42.3 t
3.1. General procedures
69.9 s
49.8 s
47.0 d
42.7 t
64.0 s
49.9 s
46.8 d
43.2 t
1H 0600 MHz), 13C 0150 MHz) and all 2D NMR spectra
were run on Varian AM-600 NMR spectrometer, with
TMS as internal standard. FT-IR was performed on a
Magna 550 spectrometer. HRFAB-MS was run on JEOL
HX-110 double focusing mass spectrometer. The APCI
MS was performed on a Fisons/VG Platform II mass spec-
trometer. Thin layer chromatography was performed on
Sigma±Aldrich TLC plates 0250 mm thickness, 2±25 mm
particle size), with compounds visualized by spraying
with 5% 0v/v) H2SO4 in ethanol solution. DPPH was
purchased from Sigma Chemical Co. CD3OD was
purchased from Aldrich Chemical Co.
1.45 m
1.99 m
2.24 m
1.50 m
2.15 m
2.38 m
8
9
58.1 s 62.0 s
207.1 s 207.9 s
55.6 s
207.5 s
173.7 s
114.5 s
110.6 d 8.38 s
150.2 s
152.8 s
10 199.1 s 173.5 s
11 127.8 s 117.7 s
12 114.4 d 110.9 d 8.44 s
13 143.9 s 150.7 s
14 149.9 s 152.9 s
15 116.6 d 102.5 d 6.96 s
16 120.2 d 144.7 s
17 27.2 t
102.5 d 6.91 s
144.6 s
26.0 t
2.82 dd, 6.6, 14.4 26.1 t
2.89 dd, 6.6, 14.4
2.59 dd, 6.6, 14.4
2.71 dd, 6.6, 14.4
18 122.8 d 118.8 d 4.76 t, 6.0
19 135.5 s 135.3 s
119.7 d 4.75 t, 6.0
134.9 s
26.2 q
18.5 q
22.6 q
27.3 q
29.9 t
3.2. Preparation of garcinol
20 26.2 q
21 18.4 q
22 22.9 q
23 27.2 q
24 29.1 t
26.1 q
18.5 q
23.5 q
27.4 q
29.5 t
1.58 s
1.72 s
1.29 s
1.22 s
1.69 m
1.93 m
1.56 s
1.72 s
1.14 s
1.03 s
1.60 m
1.96 m
Garcinol was prepared from G. indica dried fruit rind that
was extracted with ethanol. The extract was subjected to
silica gel column eluted with hexane±ethyl acetate solvent
system.15 The fractions containing garcinol were concen-
trated and dried in vacuo. The residue was dissolved in
hexane, and remained in the freezer for 24 h to get yellow
amorphous powder from the solution and washed with cold
hexane on a glass ®lter. After recrystallization, the powder
was dissolved in hot acetonitrile at room temperature; pale
yellow crystal needles were obtained from the solvent. The
crystals were identi®ed as garcinol 04,7) from the following
spectral data: 1H NMR 0CDCl3): 6.93 01H, dd, Jˆ7.8,
2.0 Hz), 6.91 01H, d, Jˆ2.0 Hz), 6.60 01H, d, Jˆ7.8 Hz),
4.92, 5.02, 5.06 01H each, t, Jˆ6.0 Hz), 4.38, 4.34 01H each,
s), 2.80±1.46 012H, m, methylene and methine), 1.77, 1.71,
1.67, 1.63, 1.57, 1.52, 1.50, 1.14, 0.98 03H each, s); 13C
NMR 0CDCl3): Table 1; Positive APCI-MS: m/z 603
[M1H]1.
25 123.9 d 123.8 d 4.66 t, 6.0
26 133.1 s 133.2 s
123.7 d 4.64 t, 6.0
133.1 s
27 25.9 q
28 18.1 q
29 36.3 t
25.9 q
17.9 q
37.6 t
1.44 s
1.09 s
1.79 m
2.28 m
2.74 m
25.8 q
17.7 q
36.4 t
1.47 s
0.95 s
2.01 m
2.24 m
2.38 m
30 43.7 d 43.9 d
31 148.2 s 149.2 s
44.2 d
147.2 s
32 112.9 t 113.9 t 4.31 s
4.41 s
112.2 t 4.11 s
4.14 s
33 17.8 q
34 32.8 t
18.2 q
32.4 t
1.56 s
1.97 m
2.00 m
18.3 q
33.2 t
1.54 s
2.00 m
2.16 m
35 124.2 d 123.0 d 5.00 t, 6.0
36 132.2 s 132.9 s
122.5 d 4.97 t, 6.0
132.5 s
37 26.0 q
38 18.1 q
26.0 q
18.0 q
1.64 s
1.54 s
25.9 q
18.1 q
1.62 s
1.49 s

S. Sang et al. / Tetrahedron 57 62001) 9931±9938
9937
3.3. Oxidation of 1 and isolation of reaction products 2
and 3
3.6. DNA extraction and electrophoresis analysis
HL-60 cells 05£105 cells/mL) were harvested, washed
with PBS, and then lysed with digestion buffer containing
0.5% sarkosyl, 0.5% mg mL21 proteinase K, 50 mM
tris0hydroxymethyl)aminomethane 0pH 8.0), and 10 mM
EDTA at 568C for 3 h and treated with RNase A
00.5 mg mL21) for another 2 h at 568C The DNA was
extracted by phenol±chloroform±isoamyl 025:24:1) before
loading and analyzed by 1.8% agarose gel electrophoresis.
The agarose gels were run at 50 V for 120 min in Tris±
borate/EDTA electrophoresis buffer 0TBE). Approximately
20 mg of DNA was loaded in each well and visualized under
UV light and photographed.
Garcinol 01) 01.0 g, 1.67 mmol) was allowed to react with
DPPH 01.3 g, 3.34 mmol) in acetone at room temperature
under darkness for 24 h. After evaporation of the solvent in
vacuo, the residue was ®rst applied to silica gel column
eluting with chloroform ®rst to get rid of the DPPH and 1
060 mg) and then with chloroform±acetone 010:1) to get a
mixture of two reaction products. The mixture of two reac-
tion products was subjected to RP-C18 silica gel column
eluting by 90% methanol±water solvent system to give
90 mg compound 2 and 30 mg compound 3.
Compound 2 was isolated as a pale amorphous substance:
IR 0®lm) nmax: 3300, 2918, 2849, 1732, 1681, 1616, 1512,
1
1465, 1390, 1293, 1190, 1143 cm21; H NMR 0CD3OD,
3.7. Nitrite assay
600 MHz): see Table 1; 13C NMR 0CD3OD, 150 MHz):
see Table 1; positive HRFAB-MS m/z 601.3527 [M1H]1
0calcd for C38H49O6, 601.3529).
The nitrite concentration in the culture medium was
measured as an indicator of NO production, according to
the Griess reaction. One hundred microliters of each super-
natant was mixed with the same volume of Griess reagent
01% sulfanilamide in 5% phosphoric acid and 0.1%
naphthylethylenediamine dihydrochloride in water).
Absorbance of the mixture at 550 nm was determined
with an enzyme-linked immunosorbent assay plate reader
0Dynatech MR-7000; Dynatech Labs, Chantilly, VA).
Compound 3 was isolated as a pale amorphous substance:
IR 0®lm) nmax: 3300, 2910, 2845, 1725, 1680, 1615, 1510,
1465, 1390, 1285, 1140 cm21
;
1H NMR 0CD3OD,
600 MHz): see Table 1; 13C NMR 0CD3OD, 150 MHz):
see Table 1; positive HRFAB-MS m/z 601.3527 [M1H]1
0calcd for C38H49O6, 601.3529).
3.4. Cell culture and chemicals
3.8. Zymographic assay for metalloproteinase
Human promyelocytic leukemia 0HL-60) cells obtained
from American Type Culture Collection 0Rockville, MD)
were grown in 90% RPMI 1640 and 10% fetal bovine serum
0GIBCO BRL, Grand Island, NY), supplemented with
2 mM glutamine 0GIBCO BRL), 1% penicillin±strepto-
The amount of gelatinase in the conditioned media 0serum
free) was estimated and quanti®ed by cell numbers. Con-
ditioned media were concentrated, and then equal amounts,
based on cell numbers determined at the time of harvest,
were electrophoresed in 10% SDS-PAGE containing 0.1%
0w/w) gelatin. The gel was washed twice, each time for
30 min at room temperature in solution containing 2.5%
0v/v) Triton X-100, and subsequently transferred to a
substrate buffer containing 10 mM CaCl2, and 500 mM
Tris±HCl, pH 8.0, at 378C with shaking for 18 h. Finally,
the gel was stained for 1 h with 0.25% 0w/v) Coomassie blue
in 10% acetic acid 0v/v) and 50% methanol and the
destained in 8% acetic acid and 5% methanol 0v/v).
21
mycin 010,000 units of penicillin/mLand 10 mg mL
streptomycin). RAW 264.7 cells were cultured in RPMI-
1640 0without phenol red) supplement with 10% endo-
toxin-free heat-inactivated fetal calf serum 0GIBCO,
Grand Island, NY). Mouse B16 melanoma cells were
maintained in Dulbecco's modi®ed Eagle's medium
0DMEM) containing 10% heat-inactivated fetal bovine
serum. Lipopolysaccharide 0LPS) 0Escherichia coli 0127:
B8), 12-O-tetradecanoylphorbol-13-acetate 0TPA), and
propidium iodide were purchased from Sigma 0St Louis,
MO). 20,70-Dichloro¯uorescein diacetate 0DCF-DA) was
purchased from 0Molecular Probes, Eugene, Oregon, USA).
3.9. Determination of hydrogen peroxide by DCF-DA in
the ¯owcytometer
3.5. Flowcytometry
DCFH-DA is a non¯uorescent probe, which becomes ¯uor-
escent in the presence of hydrogen peroxide. This dye is
stable nonpolar compound that readily diffuses into the
cells and is hydrolyzed by intracellular esterase to yield
dichlorodihydro¯uoresecin 0DCFH), which is trapped
within the cell. Hydrogen peroxide or low molecular weight
peroxides, produced during oxidative respiratory burst,
oxidize DCFH to highly ¯uorescent 20,70-dichloro¯uores-
cein 0DCF). HL-60 01£106) cells were stimulated with the
addition of 100 ng mL21 TPA and the desired concentration
of tested compounds at 378C for 30 min. The cells were
treated with 30 mM for a further 30 min and suspended in
PBS and were analyzed in the FASCAN laser ¯ow
cytometry 0Becton Dickenson, San Jose, CA).
HL-60 cells 02£105) were cultured in 60 mm Petri dishes
and incubated for 12 h. Then cells were harvested, washed
with PBS, resuspended in 200 mLof PBS, and ®xed in
800 mLof iced 100% ethanol at 2208C. After being left
to stand overnight, the cell pellets were collected by centri-
fugation, resuspended in 1 mLof hypotonic buffer 00.5%
Triton X-100 in PBS and 0.5 mg mL21 RNase), and incu-
bated at 378C for 30 min. Then 1 mLof propidium iodide
solution 050 mg mL21) was added, and the mixture was
allowed to stand on ice for 30 min. Fluorescence emitted
from the propidium iodide±DNA complex was quantitated
after excitation of the ¯uorescent dye by FACScan
cytometry 0Becton Dickinson, San Jose, CA).

9938
S. Sang et al. / Tetrahedron 57 62001) 9931±9938
Kobayashi, Y.; Miyauchi, K.; Shimano, R. Biol. Pharm.
Bull. 1996, 19, 311±314.
References
9. Yamaguchi, F.; Ariga, T.; Yoshimura, Y.; Nakazawa, H.
J. Agric. Food Chem. 2000, 48, 180±185.
1. Papas, A. M. Antioxidant Status, Diet, Nutrition, and Health;
CRC: Florida, USA, 1999 pp 3±20.
2. Slater, T. Biochem. J. 1984, 222, 1±15.
10. Yamaguchi, F.; Saito, M.; Ariga, T.; Yoshimura, Y.;
Nakazawa, H. J. Agric. Food Chem. 2000, 48, 2320±2325.
11. Tanaka, T.; Kohno, H.; Shimada, R.; Kagami, S.; Yamaguchi,
F.; Kataoka, S.; Ariga, T.; Murakami, A.; Koshimizu, K.;
Ohigashi, H. Carcinogenesis 2000, 21, 1183±1189.
12. Pan, M. H.; Chang, W. L.; Lin-Shiau, S. Y.; Ho, C. T.; Lin,
J. K. J. Agric. Food Chem. 2001, 49, 1464±1474.
13. Deby, C.; Magotteau, G. CR Soc. Biol. 1970, 164, 267±268.
14. Frankel, E. N. Lipid Oxidation; The Oily Press: Dundee,
Scotland, 1998 pp 129±160.
3. Yagi, K. Chem. Phys. Lipids 1987, 45, 337±351.
4. Krishnamurthy, N.; Lewis, Y. S.; Ravindranath, B.
Tetrahedron Lett. 1981, 22, 793±796.
5. Krishnamurthy, N.; Ravindranath, B. Tetrahedron Lett. 1982,
23, 2233±2236.
6. Bakana, P.; Claeys, M.; Totte, J.; Pieters, L. A. C.; Van Hoof,
L.; Van Den Berghe, D. A.; Vlie-Tink, A. J.
J. Ethnopharmacol. 1987, 21, 75±84.
7. Sahu, A.; Das, B.; Chatterjee, A. Phytochemistry 1989, 28,
1233±1235.
8. Iinuma, M.; Tosa, H.; Tanaka, T.; Kanamaru, S.; Asai, F.;
15. Still, W. C.; Kahn, M.; Mitra, A. J. Org. Chem. 1978, 43,
2923±2925.