Inhibition of cyclooxygenase and prostaglandin E2 synthesis by γ-mangostin, a xanthone derivative in mangosteen, in C6 rat glioma cells

Article abstract of DOI:10.1016/S0006-2952(01)00810-3

The fruit hull of mangosteen, Garcinia mangostana L., has been used for many years as a medicine for treatment of skin infection, wounds, and diarrhea in Southeast Asia. In the present study, we examined the effect of γ-mangostin, a tetraoxygenated diprenylated xanthone contained in mangosteen, on arachidonic acid (AA) cascade in C6 rat glioma cells. γ-Mangostin had a potent inhibitory activity of prostaglandin E₂ (PGE₂) release induced by A23187, a Ca²⁺ ionophore. The inhibition was concentration-dependent, with the IC₅₀ value of about 5 μM. γ-Mangostin had no inhibitory effect on A23187-induced phosphorylation of p42/p44 extracellular signal regulated kinase/mitogen-activated protein kinase or on the liberation of [¹⁴C]-AA from the cells labeled with [¹⁴C]-AA. However, γ-mangostin concentration-dependently inhibited the conversion of AA to PGE₂ in microsomal preparations, showing its possible inhibition of cyclooxygenase (COX). In enzyme assay in vitro, γ-mangostin inhibited the activities of both constitutive COX (COX-1) and inducible COX (COX-2) in a concentration-dependent manner, with the IC₅₀ values of about 0.8 and 2 μM, respectively. Lineweaver-Burk plot analysis indicated that γ-mangostin competitively inhibited the activities of both COX-1 and -2. This study is a first demonstration that γ-mangostin, a xanthone derivative, directly inhibits COX activity.

Full text of DOI:10.1016/S0006-2952(01)00810-3

Biochemical Pharmacology 63 (2002) 73–79
Inhibition of cyclooxygenase and prostaglandin E2 synthesis by
␥-mangostin, a xanthone derivative in mangosteen, in C6 rat
glioma cells૾
Keigo Nakatania, Norimichi Nakahatab,*, Tsutomu Arakawac, Hideyuki Yasudac,
Yasushi Ohizumia
aDepartment of Pharmaceutical Molecular Biology, Graduate School of Pharmaceutical Sciences, Tohoku University, Aoba, Aramaki, Aoba-ku, Sendai
980-8578, Japan bDepartment of Cellular Signaling, Graduate School of Pharmaceutical Sciences, Tohoku University, Aoba, Aramaki, Aoba-ku, Sendai
c
980-8578, Japan Central Laboratory, Lotte Co., Ltd., 3–1-1 Numakage, Urawa, Saitama 336-0027, Japan
Received 10 February 2001; accepted 14 September 2001
Abstract
The fruit hull of mangosteen, Garcinia mangostana L., has been used for many years as a medicine for treatment of skin infection,
wounds, and diarrhea in Southeast Asia. In the present study, we examined the effect of ␥-mangostin, a tetraoxygenated diprenylated
xanthone contained in mangosteen, on arachidonic acid (AA) cascade in C6 rat glioma cells. ␥-Mangostin had a potent inhibitory activity
of prostaglandin E2 (PGE2) release induced by A23187, a Ca2ϩ ionophore. The inhibition was concentration-dependent, with the IC50 value
of about 5 ␮M. ␥-Mangostin had no inhibitory effect on A23187-induced phosphorylation of p42/p44 extracellular signal regulated
kinase/mitogen-activated protein kinase or on the liberation of [14C]-AA from the cells labeled with [14C]-AA. However, ␥-mangostin
concentration-dependently inhibited the conversion of AA to PGE2 in microsomal preparations, showing its possible inhibition of
cyclooxygenase (COX). In enzyme assay in vitro, ␥-mangostin inhibited the activities of both constitutive COX (COX-1) and inducible
COX (COX-2) in a concentration-dependent manner, with the IC values of about 0.8 and 2 ␮M, respectively. Lineweaver–Burk plot
50
analysis indicated that ␥-mangostin competitively inhibited the activities of both COX-1 and -2. This study is a first demonstration that
␥-mangostin, a xanthone derivative, directly inhibits COX activity. © 2002 Elsevier Science Inc. All rights reserved.
Keywords: ␥-Mangostin; Cyclooxygenase; PGE2; C6 rat glioma cells; Competitive inhibition
1. Introduction
medicine for the treatment of skin infections, wounds, and
diarrhea for many years [1]. Thus, mangosteen may be
expected to contain an anti-inflammatory drug. ␥-Mangos-
tin, a constituent of the fruit hull, is a tetraoxygenated
diprenylated xanthone derivative. Xanthone derivative has
been reported to possess several pharmacological activities,
such as antimalarial activity [2], antimicrobial activity
against methicillin-resistant Staphylococcus aureus
(MRSA) [3], and inhibitory activity for monoamine oxidase
[4].
Mangosteen, Garcinia mangostana L. (Guttifereae), is a
tree that is fairly widespread in Thai, India, Srilanka, and
Myanmar and is known for its medicinal properties. The
fruit hull of this plant has been in use in Thai indigenous
૾ This work was partly supported by a Grant-in-Aid for Scientific
Research from the Ministry of Education, Science, Sport and Culture of
Japan.
In the brain, prostaglandin E2 (PGE2) levels are very low
or undetectable in normal conditions, but can rise during
inflammatory processes, multiple sclerosis, and AIDS-asso-
ciated dementia [5,6]. High levels of PGE2 can affect the
activities of several cell types, including neurons, glial, and
endothelial cells, and can regulate microglia/macrophage
and lymphocyte functions during inflammatory and immune
processes [7]. Therefore, the interplay between PGE2 and
* Corresponding author. Tel.: ϩ8122217-6809; fax: ϩ8122217-6809.
E-mail address: nakahata@mail.pharm.tohoku.ac.jp (N. Nakahata)
1 Abbreviations: AA, arachidonic acid; BSA, bovine serum albumin;
COX, cyclooxygenase; COX-1, constitutive COX; COX-2, inducible
COX; cPLA2, cytosolic PLA2; EMEM, Eagle’s minimum essential me-
dium; ERK, extracellular signal regulated kinase; HEPES, 2-[4-(2-Hy-
droxyethyl)-1-piperazinyl]ethanesulfonic acid; MAPK, mitogen-activated
protein kinase;PGE2, prostaglandin E2; sPLA2, secretory phospholipase A2;
TBST, Tris-buffered saline containing 0.05% Tween 20.
0006-2952/02/$ – see front matter © 2002 Elsevier Science Inc. All rights reserved.
PII: S0006-2952(01)00810-3

74
K. Nakatani et al. / Biochemical Pharmacology 63 (2002) 73–79
other local factors, including pro- and anti-inflammatory
cytokines, is likely to influence the outcome of inflamma-
tory and immune responses in the central nervous system
(CNS). There is considerable evidence linking the genera-
tion of prostaglandins (PGs) with inflammation, pain, and
fever. Glial cells, which outnumber neurons by about 10 to
1 in the brain, provide both mechanical and metabolic sup-
port for neurons. Glial cells are assumed to be an important
source of PGs in the CNS [8]. It is suggested that regulation
of arachidonic acid (AA) metabolism, particularly PGE2
production, appears beneficial in patients with inflammatory
conditions [9].
medium (EMEM) was purchased from Nissui Pharmaceu-
tical Co. Ltd. PGE2 was a generous from Ono Pharmaceu-
ticals. Anti-PGE2 antibody was obtained from Chemicon
International Inc. [3H]-PGE2 (200 Ci/mmol) and [14C]-AA
(51 mCi/mmol) were from NEN/DuPont. Anti-phospho-
p44/p42 ERK/MAPK antibody and alkaline phosphatase-
conjugated goat anti-rabbit IgG were obtained from New
England Biolabs. COX-1 and -2 were purchased from Cy-
man Chemical. AA was obtained from Sigma Chemical
Co., Ltd. AA and [14C]-AA were diluted by dimethyl sul-
foxide. A final concentration of less than 0.5% dimethyl
sulfoxide was kept. Other chemicals and drugs were of
reagent grade or of the highest quality available.
Phospholipase A2 (PLA2) is subdivided into several
groups based on their structures and enzymatic characteris-
tics [10–12]. Secretory (s)PLA2 is a family of low-molec-
ular-mass (ϳ14 kDa) enzymes that require millimolar con-
centration of Ca2ϩ for enzymatic activity. Cytosolic
(cPLA2), or group IV PLA2, is a ubiquitously distributed 85
kDa enzyme, the activation of which is tightly regulated by
postreceptor transmembrane signaling. cPLA2 is activated
by p42/p44 extracellular signal regulated kinase (ERK)/
mitogen-activated protein kinase (MAPK) under cytosolic
Ca2ϩ concentration of the submicromolar or micromolar
range, and p42/p44 ERK/MAPK was activated by dual
phosphorylation of tyrosine/threonine residues [11,13].
Cyclooxygenase (COX) is the rate-limiting enzyme in
PG synthesis and exists as two isoforms, constitutive
(COX-1) and inducible (COX-2). These isoforms originate
from distinct genes, but are structurally conserved [14,15].
COX-1 is regarded as a constitutive enzyme whose expres-
sion is developmentally regulated. PGs produced by COX-1
primarily function in fluid and electrolyte homeostasis, gas-
tric acid secretion, and platelet aggregation. In contrast,
COX-2 is expressed in response to inflammatory stimuli and
is active in physiological responses to growth factors and
glucocorticoids [16].
2.2. Cell culture
C6 rat glioma cells were grown in F-10 medium con-
taining 15% horse serum and 2.5% fetal bovine serum in a
37° humidified incubator in an atmosphere of 5% CO2 in
air.
2.3. Assay of PGE2
C6 cells were seeded into 12-well plates at the density of
1.0 ϫ 105 cells per well. The experiment was performed 2
days after seeding of cells. The cells were washed twice
with EMEM buffered with 20 mM 2-[4-(2-Hydroxyethyl)-
1-piperazinyl]ethanesulfonic acid (HEPES; EMEM-
HEPES), pH 7.35, and were pre-incubated with or without
␥-mangostin for 10 min. The cells were further incubated
with or without 10 ␮M A23187 for an additional 10 min.
The medium was acidified to pH 4.0 by addition of 1 N HCl,
and PGE2 was extracted twice with ethyl acetate. After ethyl
acetate was evaporated under a stream of N2 gas, the sample
was dissolved in 10 mM Tris-HCl (pH 7.6). PGE2 was
determined by radioimmunoassay, as described previously
[17,18].
In this study, we examined the effect of ␥-mangostin
(Fig. 1), a tetraoxygenated diprenylated xanthone contained
in mangosteen, on AA cascade in C6 rat glioma cells. The
results suggest that ␥-mangostin, a xanthone derivative,
directly inhibits COX activity.
2.4. Immunoblotting
C6 cells were seeded into 6-well plates at the density of
2.0 ϫ 105 cells per well. Two days after seeding, the cells
were washed twice with EMEM-HEPES and pre-incubated
with or without ␥-mangostin for 10 min at 37°. After the
cells were incubated with or without 10 ␮M A23187 for an
additional 2 min, the medium was aspirated. The cells were
solubilized by the addition of Laemmli sample buffer [19],
the composition of which was Tris-HCl 187.5 mM, sodium
dodecyl sulfate 6%, glycerol 30%, 2-mercaptoethanol 15%,
pH 6.8. The sample was boiled at 95° for 5 min. Electro-
phoresis was performed on 11% acrylamide gels. Proteins
were transferred electrically from the gel onto Immobilon
polyvinylidene difluoride membranes (Millipore) by the
semi-dry blotting method [20]. The immunoblots were
blocked for 2 h with 2% bovine serum albumin (BSA) in
Tris-buffered saline containing 0.05% Tween 20 (TBST) at
2. Materials and methods
2.1. Materials
Fetal bovine serum was obtained from Cell Culture Lab-
oratory (Cleveland, OH, USA); horse serum was purchased
from Dainippon Pharmaceutical Co. Ltd.; F-10 (Nutrient
Mixture: Ham) was obtained from GIBCO BRL; ␥-Man-
gostin was purified from the fruit hull of G. mangostana L.,
with its purity of more than 90% determined by an high-
performance liquid chromatography analysis. It was dis-
solved in dimethyl sufoxide to make a concentration of 20
mM and used after its dilution. Eagle’s minimum essential

K. Nakatani et al. / Biochemical Pharmacology 63 (2002) 73–79
75
(30:4:1, v/v). The ether layer was dried by a stream of
nitrogen gas and applied to a TLC plate (LK5D, Whatman).
The developer used was the upper phase of ethyl acetate/
water/isooctane/acetic acid (110:100:50:20, v/v). [14C]-AA
metabolites were visualized as radioluminogram with a Mo-
lecular Imager (GS363, Bio-Rad).
2.7. Assay of the enzyme activities of COX-1 and -2
Fig. 1. Chemical structure of ␥-mangostin.
Activities of COX-1 and -2 were determined according
to the procedure described [22]. COX-1 or -2 enzyme pro-
tein (each 0.5–1.0 unit) was dissolved in 150 ␮L of 50 mM
Tris-buffer containing 2 ␮M hematin and 5 mM tryptophan
(reaction mixture). The reaction mixture was pre-incubated
with or without drugs for 10 min at 37° and further incu-
bated with 50 ␮L AA for 10 min at 37°. Then, the reaction
was terminated by the addition of 20 ␮L of 1N HCl. AA
metabolites were extracted twice with ethyl acetate. After
ethyl acetate was evaporated under a stream of nitrogen gas,
the sample was dissolved in 10 mM Tris-HCl (pH 7.6).
PGE2 was determined by radioimmunoassay, as described
previously [17].
25° and incubated with anti-phospho-p42/p44 ERK/MAPK
antibody (rabbit) at 1 ␮g/mL for 2 h at 25°. The immuno-
blots were washed several times and incubated with a
1:3000 dilution of alkaline phosphatase-conjugated goat an-
ti-rabbit IgG in TBST containing 2% BSA overnight at 4°.
Blots were developed by using a chemiluminescence assay
kit (Bio Rad) and visualized by exposing the membrane to
Hyper-film ECL (Amersham).
2.5. Analysis of AA liberation
C6 cells were seeded into 12-well plates at the density of
1.0 ϫ 105 cells per well. Two days after seeding, the
medium was changed to Dulbecco’s modified Eagle’s me-
dium containing 0.3 ␮Ci/ml of [14C]-AA, and the cell were
incubated for 18 h. The cells were washed twice with
EMEM-HEPES-3 mg/mL albumin solution (pH 7.35) and
pre-incubated with or without ␥-mangostin for 10 min at
37°C. After the cells were incubated with or without 10 ␮M
A23187 for 5 min, the reaction was terminated by removing
the medium to tubes [21]. After acidifying the medium (0.6
mL) to pH 4.0 by 1N HCl, it was mixed with chloroform
(1.2 mL) and water (0.6 mL). The lower phase was dried by
the stream of nitrogen gas and applied to a thin layer
chromatography (TLC) plate (LK5D, Whatman). The de-
veloper used was the upper phase of benzene-isooctane-
acetic acid (60:30:3, v/v) [21]. [14C]-AA metabolites were
visualized as radioluminogram with a Molecular Imager
(GS363, Bio-Rad).
2.8. Data analysis
IC values were calculated from non-linear regression
50
analysis of the data. The statistical differences (P Ͻ 0.05) of
values was determined with ANOVA.
3. Results
3.1. Effect of ␥-mangostin on PGE2 synthesis stimulated
by A23187 in C6 cells
A23187, a Ca2ϩ ionophore, is known to stimulate PG
synthesis mediated through an activation of cPLA2 follow-
ing by AA liberation in glial cells [23]. In C6 rat glioma
cells, A23187 potently stimulated PGE2 release (2.3 ng/well
from 0.05 ng/well), and ␥-mangostin inhibited the release in
a concentration-dependent manner, with the IC50 value of
about 5 ␮M (Fig. 2).
2.6. Assay of AA conversion by microsomes of C6 cells
C6 cells were seeded into 150-mm dishes at the density
of 5.0 ϫ 105 cells per dish. Three days after seeding, C6
cells were homogenized in 50 mM Tris-buffer (pH 7.4)
containing 1 mM ethylenediaminetetraacetic acid. Micro-
somes were prepared by centrifugation at 3,000 ϫ g for 5
min at 4° followed by centrifugation of the supernatant at
100,000 ϫ g for 60 min at 4°. Microsomes (90 ␮g of
protein/tube) were pre-incubated in 50 mM Tris-buffer con-
taining 2 ␮M hematin and 5 mM tryptophan with or without
␥-mangostin for 10 min at 37°, and they were incubated
with 0.1 ␮Ci/mL of [14C]-AA for 40 min at 37°. After the
cells were incubated, the reaction was terminated by the
addition of ice-cold diethyl ether/methanol/1 M citric acid
3.2. Effect of ␥-mangostin on p42/p44 ERK/MAPK
phosphorylation and AA liberation
To determine the site of action of ␥-mangostin, we first
examined upstream of AA liberation. The p42/p44 ERK/
MAPK is known to be involved in a wide range of cellular
functions, including cPLA2 activation [24,25]. ␥-Mangostin
(3–30 ␮M) slightly augmented, but do not inhibited
A23187-induced phosphorylation of p42/p44 ERK/MAPK
(data not shown), indicating that the acting site of ␥-man-
gostin for inhibition of PGE2 release is downstream of
p42/p44 ERK/MAPK.
Because ␥-mangostin did not cause the inhibition of

76
K. Nakatani et al. / Biochemical Pharmacology 63 (2002) 73–79
Fig. 2. Effect of ␥-mangostin on PGE2 release from C6 cells. The cells
were prei-ncubated with indicated concentrations of ␥-mangostin for 10
min, and they were stimulated by 10 ␮M A23187 (F) or vehicle (E) for 10
min. The released PGE2 in the medium was determined by radioimmuno-
assay. Each point represents the mean with S.E. of three determinations.
*Significant difference from A23187 alone.
Fig. 4. Effect of ␥-mangostin on COX activity of microsomes of C6 cells.
Microsomes of C6 cells were pre-incubated with indicated concentrations
of ␥-mangostin for 10 min and were incubated with [14C]-AA for 40 min.
[
14C]-AA metabolites were analyzed in radioluminogram with a molecular
imager (GS363, Bio-Rad) after separation by TLC (upper figure). Densi-
tometric analysis of radioluminogram (lower figure). The results are shown
as the -fold increase as without ␥-mangostin of 1. Each point represents the
mean with S.E. of three determinations. *Significant difference from with-
out ␥-mangostin.
phospholyration of p42/p44 ERK/MAPK, we examined the
effect of ␥-mangostin on AA liberation from [14C]-AA-labeled
cells (Fig. 3). ␥-Mangostin slightly augmented, but did not
inhibited A23187-induced AA liberation, suggesting that
␥-mangostin inhibits PGE2 synthesis at the downstream of AA
liberation. Interestingly, ␥-mangostin alone at a concentration
of 30 ␮M did activate rather than inhibit the liberation of AA.
examined the conversion of AA to PGE2 in the microsomal
fraction. Western blotting analysis revealed that COX-1, but
not COX-2, was expressed in C6 cells under this experi-
mental condition (data not shown). Incubation of the mi-
crosomes with [14C]-AA resulted in the productions of
[14C]-eicosanoids. Among them, [14C]-PGE2 was a main
metabolite. ␥-Mangostin inhibited [14C]-PGE2 production
from [14C]-AA in a concentration-dependent manner, with
the IC50 value of about 17 ␮M (Fig. 4).
3.4. Effect of ␥-mangostin on the conversion of AA to
PGE2 in microsomal fraction
COX, existing in microsomes, is the rate-limiting en-
zyme in the conversion of AA to prostanoids. Therefore, we
3.5. Effect of ␥-mangostin on enzyme activities of COX-1
and -2 in vitro
Because ␥-mangostin inhibited the conversion of AA to
PGE2 in microsomes, we examined the direct effect of
␥-mangostin on COX-1 or -2 enzyme in vitro, comparing
the effects of indomethacin, a non-selective inhibitor of
COX, and NS398, a selective inhibitor of COX-2. Indo-
methacin strongly inhibited both the COX-1 and -2 activi-
ties in a concentration-dependent manner (Figs. 5C and D).
On the other hand, NS-398 inhibited the COX-2 activity in
a concentration-dependent manner, but not COX-1 activity
(Figs. 5E and F). ␥-Mangostin inhibited both the COX-1
and -2 activities in a concentration-dependent manner, with
the IC50 values of about 0.8 ␮M and 2 ␮M, respectively
(Figs. 5A and B).
Fig. 3. Effect of ␥-mangostin on liberation of AA. The cells were labeled
with 0.3 ␮Ci/mL of [14C]-AA for 18 h. After washing out, the cells were
pre-incubated with indicated concentrations of ␥-mangostin for 10 min,
and they were stimulated by 10 ␮M A23187 for 5 min. [14C]-AA was
analyzed in radioluminogram with a molecular imager (GS363, Bio-Rad)
after separation by TLC (upper figure). Densitometric analysis of radiolu-
minogram (lower figure). Open column: basal liberation, hatched column:
A23187-induced liberation. The results are shown as the -fold increase
from control (without A23187). Each column represents the mean with
S.E. of three determinations.
To clarify the characteristics of the inhibition of COX-1
or -2, we analyzed the kinetics of the inhibitory effect of
␥-mangostin. Figs. 6 and 7 show typical saturation curves
(Figs. 6A and 7A) and Lineweaver-Burk plots (Figs. 6B and

K. Nakatani et al. / Biochemical Pharmacology 63 (2002) 73–79
77
Fig. 5. Effects of ␥-mangostin, indomethacin, and NS-398 on enzyme activities of COX-1 and -2. The reaction mixture with COX-1 (A, C, E) or -2 (B, D,
F) were pre-incubated with indicated concentrations of ␥-mangostin (A, B), indomethacin (C, D), or NS-398 (E, F) for 10 min and were incubated with AA
for 10 min. The released PGE2 in the medium was determined by radioimmunoassay. Each point represents the mean with S.E. of three determinations.
*Significant difference from without ␥-mangostin.
7B) for COX-1 and -2 activities in the presence or ab-
sence of 0.5 ␮M ␥-mangostin. In the COX-1 activity
(Fig. 6), the Km value was increased from 0.75 to 1.7 ␮M
in the presence of ␥-mangostin, whereas the Vmax value
was unaffected (6.24 nmol/l⅐min for control and 6.53
nmol/l⅐min with ␥-mangostin). In the COX-2 activity
(Fig. 7), the Km value was increased from 0.25 to 0.75
␮M in the presence of ␥-mangostin, whereas the Vmax
value was unaffected (6.24 nmol/l⅐min for control and
6.53 nmol/l⅐min with ␥-mangostin). Thus, ␥-mangostin
has an ability to inhibit the COX-1 and -2 activities
competitively.
Fig. 6. (A) Saturation curve and (B) Lineweaver-Burk plot of COX-1
activity of ␥-mangostin in a cell-free system. The reaction mixture with
COX-1 were pre-incubated with ␥-mangostin for 10 min and were incu-
bated with indicated concentrations of AA for 10 min. The released PGE2
in the medium was determined by radioimmunoassay. Each point repre-
sents the mean with S.E. of three determinations.
Fig. 7. (A) Saturation curve and (B) Lineweaver-Burk plot of COX-2
activity of ␥-mangostin in a cell-free system. The reaction mixture with
COX-2 were pre-incubated with ␥-mangostin for 10 min and were incu-
bated with indicated concentrations of AA for 10 min. The released PGE2
in the medium was determined by radioimmunoassay. Each point repre-
sents the mean with S.E. of three determinations.

78
K. Nakatani et al. / Biochemical Pharmacology 63 (2002) 73–79
4. Discussion
tin in purified COX enzymes are similar to the IC50 values in
intact cells. Then, the microsomal preparation was excep-
tional, and ␥-mangostin might be degraded with P450 en-
zymes in the microsomal preparation. Because ␥-mangostin
inhibited both the COX-1 and -2, the drug seems to be
similar to indomethacin in its selectivity to COX. Indometh-
acin is known as a competitive, time-dependent, reversible
COX inhibitor. It forms an enzyme-inhibitor (EI) complex
but then secondarily changes the structure of the protein to
produce an intermediate state of enzyme-inhibitor (EI*)
complex without covalently modifying the protein [34]. EI*
complex formation is relatively slow, occurring in seconds
to minutes, and EI* slowly reverts to EI [14]. Aspirin is a
competitive, irreversible COX inhibitor. It converts EI to an
EI* complex by covalent modification (acylation) of the
protein. Once an EI* complex is formed with aspirin, it is
not possible for the protein to revert to EI [14]. Like indo-
methacin, ␥-mangostin does not have a site of covalent
modification (e.g. acetoxyl group). Then, ␥-mangostin may
be a reversible COX inhibitor. Furthermore, the kinetic
analysis in this study indicated that ␥-mangostin inhibited
the COX-1 and -2 activities competitively. Thus, ␥-man-
gostin competes with arachidonate for binding to the
COX-1 and -2 active site.
The fruit hull of G. mangostana L., mangosteen, has
been widely used as an anti-inflammatory agent for the
treatment of skin infections, wounds, and diarrhea for many
years in Southeast Asia [1]. The crude extract of the fruit
hull has been reported to possess several pharmacological
activities, such as inhibitory activity against HIV-1 protease
[26] and antimicrobial activity [27]. Several constituents
contained in the fruit hull of mangosteen, such as ␣-man-
gostin, ␥-mangostin, or another substances, have diverse
pharmacological activities, such as anti-histamine activity
[28], inhibition of the sarcoplasmic reticulum Ca2ϩ-pump-
ing adenosine 5Ј-triphosphate (ATP)ase [29], and anti-se-
rotonin activity [30]. Among then, we selected ␥-mangostin
to examine its activity of PGE2 synthesis. In the analysis
using C6 rat glioma cells, we show that ␥-mangostin has a
potent inhibitory effect on PGE2 synthesis. A synthetic
xanthone derivative, 3-[2-(cyclopropylamino)ethoxy] xan-
thone inhibits AA-induced platelet aggregation and throm-
boxane formation [31], showing its possible inhibition of
COX activity. In the present study, we showed for the first
time that ␥-mangostin, a xanthone derivative derived from
natural products, directly binds to COX and inhibits its
activity. This effect of ␥-mangostin may contribute to its
anti-inflammatory activity.
C6 cells, derived from rat glial tumors induced by N-
nitrosomethyl urea [32], are commonly used as an estab-
lished cell line for a model of glial cells. Glial cells are
known as an important source of PGs in the CNS [33].
Because a high PGE2 level is observed in some diseases,
such as multiple sclerosis and AIDS-associated dimension,
drugs reducing PGE2 synthesis in glial cells have a possi-
bility to improve the diseases with inflammation in the
brain. Thus, ␥-mangostin is one of the candidates of the
drugs for treatment brain diseases accompanied with in-
flammation, although further detailed analysis is necessary,
such as metabolism and distribution of the drug in vivo.
Because ␥-mangostin has an inhibitory effect of PGE2
synthesis, we examined its site of action. However, ␥-man-
gostin slightly augmented, but not inhibited, the phosphor-
ylation of p42/p44 ERK/MAPK. Furthermore, the A23187-
induced AA liberation was slightly augmented, but not
inhibited, by ␥-mangostin. Interestingly, ␥-mangostin alone
at a concentration of 30 ␮M did activate rather than inhibit
the liberation of AA, although it inhibited PGE2 synthesis.
Because the acting site of ␥-mangostin may be downstream
of AA liberation, we examined the effect of ␥-mangostin on
the conversion of AA to PGE2 in microsomal preparations.
The conversions of AA to PGE2 was potently inhibited by
␥-mangostin, suggesting that the site of ␥-mangostin was
COX. In the analysis of enzyme activity in vitro, we found
that ␥-mangostin directly interacted with COX-1 and -2 and
inhibited their activities. There is a certain difference in the
In conclusion, we showed for the first time that ␥-man-
gostin, a tetraoxygenated diprenylated xanthone from man-
gosteen, reduces PG generation through its direct inhibition
of COX. ␥-Mangostin is an attractive drug because its
analogs of tetraoxygenated diprenylated xanthones are con-
tained in many plants, vegetables, and fruits.
References
[1] Mahabusarakam W, Iriyachitra P, Taylor WC. Chemical constituents
of garcinia mangostana. J Nat Prod 1987;50:474–8.
[2] Likhitwitayawuid K, Chanmahasasthien W, Ruangrungsi N, Krungk-
rai J. Xanthones with antimalarial activity from Garcinia dulcis.
Planta Medica 1998;64:281–2.
[3] Dharmaratne HRW, Wijesinghe WMNM, Thevanasem V. Antimi-
crobial activity of xanthones from Calophyllum species, against me-
thicillin-resistant Staphylococcus aureus (MRSA). J Ethnopharm
1999;66:339–42.
[4] Ohishi N, Suzuki T, Ogasawara T, Yagi K. Xanthone derivative as
inhibitors for monoamine oxidase. J Mol Cat 2000;10:291–4.
[5] Fretland DJ. Potential role of prostaglandins and leukotrienes in
multiple sclerosis and experimental allergic encephalomyelitis. Pros-
taglandins Leukot Essent Fatty Acids 1992;45:249–57.
[6] Griffin DE, Wesselingh SL, McArthur JC. Elevated central nervous
system prostaglandins in human immunodeficiency virus-associated
dementia. Ann Neurol 1994;35:592–7.
[7] Weissmann G. Prostaglandins as modulators rather than mediators of
inflammation. J Lipid Mediat 1993;6:275–86.
[8] Minghetti L, Polazzi E, Nicolini A, Levi G. Opposite regulation of
prostaglandin E2 synthesis by transforming growth factor-␤1 and
interleukin 10 in activated microglial cultures. J Neuroimmunol 1998;
82:31–9.
IC values of ␥-mangostin between intact cells and micro-
somal preparations. However, the IC values of ␥-mangos-
50
[9] Goldstein IM. Agents that interfere with arachidonic acid metabo-
lism. In: Gallin JI, Goldstein IM, Snyderman R, editors. Inflamma-
50

K. Nakatani et al. / Biochemical Pharmacology 63 (2002) 73–79
79
tion, basic principles and clinical correlates. 2nd ed. New York:
Raven Press, 1992. p 1127–37.
inhibition by nonsteroidal anti-inflammatory drugs containing a car-
boxylic acid moiety. J Biol Chem 1995;270:29372–7.
[10] Dennis EA. The growing phospholipase A2 superfamily of signal
transduction enzymes. Trends Biochem Sci 1997;22:1–2.
[11] Leslie CC. Properties and regulation of cytosolic phospholipase A2.
J Biol Chem 1997;272:16709–12.
[23] Nakahata N, Kutsuwa M, Kyo R, Kubo M, Hayashi K, Ohizumi Y.
Analysis of inhibitory effects of scutellariae radix and baicalein on
prostaglandin E2 production in rat C6 glioma cells. Am J Chin Med
1998;26:311–23.
[12] Tischfield JA. A reassessment of the low molecular weight phospho-
lipase A2 gene family in mammals. J Biol Chem 1997;272:17247–50.
[13] Lin LL, Wartmann M, Lin AY, Knopf JV, Hyslop PA, Jakubowski
JA. cPLA2 is phosphorylated, and activated by MAP kinase. Cell
1993;72:269–78.
[24] Nishida E, Gotoh Y. The MAP kinase cascade is essential for diverse
signal transduction pathways. Trends Biochem Sci 1993;18:128–31.
[25] Pumiglia KM, Decker SJ. Cell cycle arrest mediated by the MEK/
mitogen-activated protein kinase pathway. Proc Natl Acad Sci USA
1997;94:448–52.
[26] Chen S-X, Wan M, Loh B-N. Active constituents against HIV-1
protease from garcinia mangostana. Planta Med 1996;62:381–2.
[27] Sundaram BM, Gopalakrishnan C, Subramanian S, Shankaranaray-
anan D, Kameswaran L. Antimicrobial activities of Garcinia man-
gostana. Planta Medica 1983;48:59–60.
[14] Smith WL, DeWitt DL. Prostaglandin endoperoxide H synthesis-1
and -2. Adv Immunol 1996;62:167–215.
[15] Williams CS, DuBios RN. Prostaglandin endoperoxide synthesis:
why two isoforms? Am J Physiol 1996;270:G393–400.
[16] DeWitt DL, Meade EA. Serum and glucocorticoid regulation of gene
transcription and expression of prostaglandin H synthase-1 and pros-
taglandin H synthase-2 isozymes. Arch Biochem Biophys 1993;306:
94–102.
[17] Nakahata N, Imata K, Ohkawa T, Watanabe Y, Ishimoto H, Ono T,
Ohizumi Y, Nakanishi H. Mastoparan elicits prostaglandin E2 gen-
eration and inhibits inositol phosphate accumulation via different
mechanisms in rabbit astrocytes. Biochem Biophys Acta 1996;1310:
60–6.
[18] Nakatani K, Nakahata N, Hamada Y, Tsurufuji S, Ohizumi Y. Me-
dium change amplifies mitogen-activated protein kinase-mediated
prostaglandin E2 synthesis in Swiss 3T3 fibroblasts. Eur J Pharm
1998;356:91–100.
[19] Laemmli UK. Cleavage of structural proteins during the assembly of
the head of bacteriophage T4. Nature 1970;227:680–5.
[20] Ohkubo S, Nakahata N, Ohizumi Y. Thromboxane A2 stimulates
mitogen-activated protein kinase and arachidonic acid liberation in
rabbit platelets. Prostaglandins 1996;52:403–13.
[28] Chairungsrilerd N, Furukawa K-I, Ohta T, Nozoe S, Ohizumi Y.
pharmacological properties of ␣-mangostin, a novel histamine H1
receptor antagonist. Eur J Pharmacol 1996;314:351–6.
[29] Furukawa K-I, Shibusawa K, Chairungsrilerd N, Ohta T, Nozoe S,
Ohizumi Y. The mode of inhibitory action of ␣-mangostin, a novel
inhibitor, on the sarcoplasmic reticulum Ca2ϩ-pumping ATPase from
rabbit skeletal muscle. Jpn J Pharmacol 1996;71:337–40.
[30] Chairungsrilerd N, Furukawa K-I, Ohta T, Nozoe S, Ohizumi Y.
␥-mangostin, a novel type of 5-hydroxytryptamine 2A receptor an-
tagonist. Naunyn-Schmiedeberg’s Arch Pharmacol 1997;357:25–31.
[31] Chen WY, Ko FN, Lin CN, Teng CM. The effect of 3-[2-(cyclopro-
pylamino)ethoxy]xanthone on platelet thromboxane formation.
Thromb Res 1994;75:81–90.
[32] Benda P, Lightbody J, Sato G, Levine L, Sweet W. Differentiated rat
glial cell strain in tissue culture. Science 1968;161:370–1.
[33] Seregi A, Keller M, Hertting G. Are cerebral prostanoids of astroglial
origin? Studies on the prostanoid forming system in developing rat
brain and primary cultures of rat astrocytes. Brain Res 1987;404:113–
20.
[21] Rho MC, Nakahata N, Nakamura H, Murai A, Ohizumi Y. Involve-
ment of phospholipase C-␥2 in activation of mitogen-activated pro-
tein kinase and phospholipase A2 by zooxanthellatoxin-A in rabbit
platelets. J Pharmacol Exp Ther 1997;282:496–504.
[22] Mancini JA, Riendeau D, Falgueyret JP, Vickers PJ, O’neill GP.
Arginine 120 of prostaglandin G/H synthase-1 is required for the
[34] Stanford N, Roth GJ, Shen TY, Majerus TY. Lack of covalent
modification of prostaglandin synthase (cyclooxygenase) by indo-
methacin. Prostaglandins 1977;13:669–75.