Journal of Ethnopharmacology 112 (2007) 221–227
Selective inducible nitric oxide syntha
bracteanolides from Murdann
b
, W
uei
ine, T
niver
250 W
ry 20
2007
Abstract
Murdanni flam
nor its anti-inflammatory actions are well defined. Nitric oxide (NO), overproduced by activated macrophages via inducible NO synthase (iNOS),
is suggested to be a significant pathogenic factor in various inflammatory tissue injuries. In order to elucidate the anti-inflammatory actions of M.
bracteata, the present study was designed to isolate its active constituents and examine its effects on iNOS in lipopolysaccharide (LPS)-activated
macrophages. Two new hydroxybutenolides, bracteanolide A (1) and B (2), together with (+)-(R)-p-hydroxyphenyllactic acid (3) and isovitexin
(4), were isolated and identified from M. bracteata by the NO production assay. All of the compounds inhibited NO production except 3. Their rank
order of pote
in a concent
rat aorta reve
The selectiv
© 2007 Else
Keywords: M
1. Introdu
Murdan
throughout
been used
monia, nep
and Chang
cient in cl
anti-inflam
its active co
Nitric o
from a var
macrophag
Abbreviat
inducible nitr
∗ Correspon
E-mail ad
0378-8741/$
doi:10.1016/jncy was 1 > 2 > 4. Among these, 1 significantly inhibited NO production, which is associated with its suppression on iNOS induction
ration-dependent manner, with an IC50 of 33.27 ± 0.86M. Nevertheless, isometric tension recordings in isolated endothelium-intact
aled that 1–4 did not affect acetylcholine-induced endothelial NO-dependent relaxation, an index of endothelial NOS (eNOS) activity.
e inhibition on iNOS provides a possible explanation for the anti-inflammatory use of M. bracteata.
vier Ireland Ltd. All rights reserved.
urdannia bracteata; Commelinaceae; Bracteanolide A; Bracteanolide B; Inducible nitric oxide synthase; RAW 264.7 cells
ction
nia bracteata, a perennial herb, is widely distributed
the Indo-China Peninsula (Hong, 1997). It has long
in folk medicine to treat hepatitis, stomatitis, pneu-
hritis and many other inflammatory diseases (Chiu
, 1995). However, the pharmacological data is defi-
early establishing the scientific rationale for the
matory medicinal use of this plant; the search for
nstituents is also limited.
xide (NO) is one of the critical mediators released
iety of cells, such as vascular endothelial cells and
es, that alters cardiovascular homeostasis, leading to
ions: NO, nitric oxide; NOS, nitric oxide synthase; iNOS,
ic oxide synthase
ding author. Tel.: +886 2 2736 1661x6156.
dress: thlee@tmu.edu.tw (T.H. Lee).
changes in the physiological condition, even when large amounts
of NO are released upon external stimulation (Wolf, 1997). Cel-
lular NO, accompanied with l-citrulline, is synthesized from
l-arginine by a family of NO synthase (NOS) (Hibbs et al.,
1987). Among three identified isoforms of NOS, the constitutive
endothelial NOS (eNOS) plays an important role in protection
against the onset and progression of cardiovascular disorders
under physiological conditions. The expression of inducible
NOS (iNOS) is induced by pro-inflammatory stimuli such as
bacterial lipopolysaccharide (LPS) or cytokines (Chartrain et
al., 1994). NO, if produced in large quantities by activated
macrophages overexpressing iNOS, has been implicated in the
pathogenesis of a variety of inflammatory-mediated disorders
including septic shock, stroke, DNA damage, and carcinogene-
sis caused by mutagenesis (Chen et al., 1998). Since cells cannot
sequester and regulate the local concentration of NO, inhibition
of NO synthesis is, therefore, a potential therapeutic approach
for the treatment of these inflammatory diseases. Therefore, the
– see front matter © 2007 Elsevier Ireland Ltd. All rights reserved.
.jep.2007.02.025Guei Jane Wang a, Shih Ming Chen
Yu Min Chang a, Tzong H
a National Research Institute of Chinese Medic
b Department of Clinical Pharmacy, Taipei Medical U
c Graduate Institute of Pharmacognosy, Taipei Medical University,
Received 25 July 2006; received in revised form 15 Februa
Available online 4 March
a bracteata has been used as a Taiwanese folk medicine for its anti-inse suppression by new
ia bracteata
ei Chou Chen a,
Lee c,∗
aipei, Taiwan, ROC
sity, Taipei., Taiwan, ROC
u-Xin Street, Taipei 110, Taiwan, ROC
07; accepted 28 February 2007
matory properties. However, neither its active ingredients

222 G.J. Wang et al. / Journal of Ethnopharmacology 112 (2007) 221–227
screening of bioactive compounds from natural resources, which
modulates the activity and/or expression of iNOS, will prove
invaluable.
NO prod
RAW 264.7
degree of i
ing the effe
et al., 2006
acetate lay
the ability
cells stimu
some bioac
study, a ser
tural elucid
and identifi
two known
tial as anti
were tested
LPS-stimu
assessed. T
significantl
macrophag
bracteanoli
which may
diseases ca
further stud
additional r
disorders.
2. Materia
2.1. Gener
Optical
ital spectro
IR 300 sp
on a Shim
tra were re
DMX-500
residual un
erences. T
pulse sequ
were obtai
m-nitroben
(Pharmacia
raphy. HPL
(BDS Hyp
Keystone,
performed
Germany).
2.2. Plant
The wh
K. Morton
the suburbs
tified by D
Center for Biodiversity, Academia Sinica, Taipei, Taiwan. The
voucher specimens (no. 10112005) were deposited in the Grad-
nstitu
, Taiw
xtrac
wh
ed th
s. T
s so
ns s
uen
ated
ceta
olut
the s
stud
ivity,
hano
ep w
epha
nol w
ed fr
by T
elop
n-su
imil
ilar
, 0.65
porti
s el
tR, 1
in)
ctea
◦ (c
.7),
160
ble
S (
09.0
ctea
◦ (c
.8),
160
ble
S (
23.0
ell c
W 2
ne)
ch
eek
M) s
llin–uction induced by LPS through iNOS expression in
cells, a mouse macrophage cell line, may reflect the
nflammation and may provide a measure for assess-
cts of test drugs on the inflammatory process (Jiang
). Through bioassay-guided fractionation, the ethyl
er of the whole plant extracts of M. bracteata has
to inhibit NO production in cultured RAW 264.7
lated with bacterial LPS. M. bracteata may contain
tive components worthy of being investigated. In this
ies of extraction, separation, purification, and struc-
ation was investigated, which has led to the isolation
cation of two new hydroxybutenolides as well as
phenolics. In an attempt to evaluate their poten-
-inflammatory agents, the isolated pure compounds
for their activities on iNOS in unstimulated- and
lated RAW 264.7 cells. Their cytotoxicity was also
he results indicated that constituents of M. bracteata
y inhibited NO production in LPS-stimulated murine
es. Of all the compounds tested, the new extract
de A (1) is the most potent and selective for iNOS,
have potential in the prevention and treatment of
used by an increased expression of iNOS, although
y is still warranted. These results also serve as an
ationale for the use of M. bracteata in inflammatory
ls and methods
al experimental procedures
rotations were measured using a JASCO P-1020 dig-
polarimeter. IR spectra were recorded on a Thermo
ectrometer. UV spectra were measured in MeOH
azu UV 1601 spectrophotometer. The NMR spec-
corded in CD3OD at room temperature on a Bruker
SB spectrometer, and the solvent resonances of the
deuterated solvent were used as internal shift ref-
he 2D NMR spectra were recorded using standard
ences. Positive ion FAB-MS and HR-FAB-MS data
ned on a JOEL SX-102A mass spectrometer using
zyl alcohol (NBA) as the matrix. Sephadex LH-20
Biotech) was used for gel permeation chromatog-
C was performed using a semi-preparative column
ersil C18, 10 mm i.d. × 250 mm, Thermo Hypersil-
Runcorn, UK; detector, refractive index). TLC was
using silica gel 60 F254 plates (200m, Merck,
material
ole plant of Murdannia bracteata (C. B. Clarke) J.
ex D. Y. Hong (Commelinaceae) were collected in
of Taipei, Taiwan on Oct. 11, 2005, and were iden-
r. Ching-I Peng, a research fellow in the Research
uate I
Taipei
2.3. E
The
extract
2 week
aqueou
fractio
Subseq
evapor
ethyl a
water s
Then,
in this
bioact
in met
tion st
on a S
metha
collect
sitions
for dev
vanilli
with s
ing sim
18–21
of this
(1:4) a
(8 mg;
11.00 m
Bra
+ 21.4
245 (3
1726,
see Ta
FAB-M
m/z = 2
Bra
− 13.4
246 (3
1729,
see Ta
FAB-M
m/z = 2
2.4. C
RA
cell li
Resear
once-w
(DME
penicite of Pharmacognosy, Taipei Medical University,
an for future reference.
tion and isolation of 1–4
ole plant of M. bracteata (6 kg) was successively
ree times with 12 l of MeOH at room temperature for
he methanolic extract (530 g) was adjusted to 85% in
lution for an n-hexane partition, which generated two
oluble in aqueous methanol and n-hexane (12.5 g).
tly, the aqueous methanol-soluble fraction was then
to dryness (103 g) and further partitioned between
te (400 ml × 2) and water (400 ml). The remaining
ion was extracted two times with n-butanol (400 ml).
ubsequent separation was guided by bioassays used
y. The ethyl acetate layer (15.5 g), with significant
was evaporated to a brown residue and re-dissolved
l for chromatographic purification. The first separa-
as carried out using gel permeation chromatography
dex LH-20 column (3 cm i.d. × 65 cm) and eluted by
ith a flow rate of 2.5 ml/min. Each fraction (15 ml)
om the ethyl acetate layer was checked for its compo-
LC using ethyl acetate/acetic acid/H2O (85:10:10)
ment, and observation under UV 254 nm. Dipping in
lfuric acid were used in the detection of compounds
ar chromophores. Subsequently, fractions contain-
compounds and exhibiting stronger bioactivity (#fr.
g) were combined to give one major portion. HPLC
on on a reversed-phase column with acetonitrile/H2O
uent, 2 ml/min, afforded 1 (25 mg; tR, 8.96 min), 2
4.90 min), 3 (11 mg; tR, 10.13 min) and 4 (47 mg; tR,
.
nolide A (1). amorphous white powder; [α]25D =
0.1, MeOH); UV (MeOH): λmax (log ε) = 214 (3.9),
303 (3.8), 329 (3.9) nm; IR (KBr): vmax = 3324,
3, 1299 cm−1; 1H NMR (CD3OD, 500 MHz):
1; 13C NMR (CD3OD, 125 MHz): see Table 1;
NBA): m/z = 209 [M + H]+; HR-FAB-MS (NBA):
453 [M + H]+ calcd. for C10H8O5 + H+: 209.0450.
nolide B (2). amorphous white powder; [α]25D =
0.1, MeOH); UV (MeOH): λmax (log ε) = 215 (3.9),
305 (3.8), 333 (3.9) nm; IR (KBr): vmax = 3369,
5, 1298 cm−1; 1H NMR (CD3OD, 500 MHz):
1; 13C NMR (CD3OD, 125 MHz): see Table 1;
NBA): m/z = 223 [M + H]+; HR-FAB-MS (NBA):
602 [M + H]+ calcd. for C11H10O5 + H+: 223.0607.
ulture
64.7 cells (a transformed murine macrophage
obtained from the Bioresource Collection and
Center (Hsinchu, Taiwan) were maintained by
ly passage in Dulbecco’s modified Eagle medium
upplemented with 10% fetal calf serum (FCS) and
streptomycin.

G.J. Wang et al. / Journal of Ethnopharmacology 112 (2007) 221–227 223
Table 1
1H and 13C NMR spectroscopic data (CD3OD, 500 MHz) for compounds 1 and 2 [δ in ppm, mult. (J in Hz)]
Position
2
3
4
5
1′
2′
3′
4′
5′
6′
OCH3
a Multiplici
2.5. NO m
Cell aliq
24-well pla
DMEM fo
To assess
pounds 1–4
a non-selec
tive iNOS
0.1%) were
to the cells
sequently c
production
was mixed
adding 1 pa
1 part 1% s
at room tem
measured b
ment, Inc.,
blank. The
a standard c
ture mediu
calculated
2.6. Cytoto
A redox
cytotoxicit
After the c
ment descr
was added
were incub
lowing inc
spectropho
2.7. Weste
The exp
cells was a
grown to co
rum
d to
end o
d ly
min
erna
rd (
rote
S-p
(Bio
hen s
7 mM
20, a
as pe
iNO
fin
adish
zed
ence
nima
lt m
nal
The1
13Ca 1H HMBC (H→C)
174.1 s
111.8 d 6.23 s 2, 4, 5, 1′
165.9 s
99.9 d 6.45 s 2, 3
122.6 s
116.0 d 7.20 s 4, 3′, 4′, 6′
146.7 s
150.5 s
116.5 d 6.82 d (8.2) 1′, 3′, 4′
122.3 d 7.18 d (8.2) 4, 2′, 4′
ties were obtained from DEPT experiments.
easurement
uots (5 × 105 cells/ml) were grown to confluence on
tes for 24 h. The medium was changed to serum-free
r another 4 h to render the attached cells quiescent.
the effects on LPS-induced NO production, com-
, two positive control Nω-nitro-l-arginine (l-NNA,
tive NOS inhibitor) and aminoguanidine (a selec-
inhibitor; 100M) or vehicle (dimethyl sulfoxide;
added in the absence or presence of LPS (50 ng/ml)
for another 24 h. The culture supernatant was sub-
ollected for the nitrite assay as a reflection of NO
(Green et al., 1982). Briefly, an aliquot of supernatant
with an equal volume of Griess reagent (prepared by
rt 0.1% napthylethylenediamine dihydrochloride to
ulfanilamide in 5% phosphoric acid) and incubated
perature for 10 min. The absorbance at 550 nm was
y a microplate spectrophotometer (Bio-Tek Instru-
Winooski, VT, USA). Fresh medium was used as the
nitrite concentration was determined by reference to
urve by using sodium nitrite diluted in the stock cul-
m. Results are expressed as percentage of inhibition
versus vehicle plus LPS-treated cells.
xicity assay
in a se
expose
At the
ice-col
for 20
the sup
Bradfo
Total p
8% SD
branes
were t
ing 13
Tween
tion w
mouse
CA). A
horser
visuali
Biosci
2.8. A
Adu
(Natio
tested.indicator, alamarBlue, was used to measure the
y as shown previously (Kwack and Lynch, 2000).
ulture supernatant was removed for NO measure-
ibed above, a solution of 10% al amarBlue in DMEM
to each well containing RAW 264.7 cells. The plates
ated at 37 ◦C in humidified 5% CO2 for 3 h. Fol-
ubation, the absorbance of the alamarBlue was read
tometrically at dual wavelengths of 570 and 600 nm.
rn blot assay
ressions of iNOS in LPS-stimulated RAW 264.7
nalyzed. RAW 264.7 cells (1.5 × 107 cells/ml) were
nfluence on culture dishes for 24 h. After starvation
controlled
relative hum
oratory cho
water were
approved b
of Nationa
conducted
Animal Ca
2.9. Vascu
The me
previously
2006). In b
fixed in org2
13Ca 1H HMBC (H→C)
173.5 s
112.6 d 6.30 s 2, 4, 5, 1′
163.9 s
104.8 d 6.30 s 2, 3
122.3 s
115.7 d 7.14 s 4, 3′, 4′, 6′
146.9 s
150.8 s
116.6 d 6.82 d (8.1) 1′, 3′, 4′
122.1 d 7.13 d (8.1) 4, 2′, 4′
55.6 q 3.52 s 5
-free medium for 4 h, the cells were simultaneously
LPS (50 ng/ml) and compounds 1–4 for another 24 h.
f the experiments, the cells were washed and lysed in
sis buffer. The lysates were centrifuged at 3000 × g
at 4 ◦C. The cytoplasmic protein concentration in
tants was determined by the protein-dye method of
1976), using bovine serum albumin as a standard.
in (50g per lane) from each sample was run on
olyacrylamide gels and transferred to PVD mem-
-Rad Laboratories Inc., Hercules, CA). Membranes
erially incubated, first with blocking buffer contain-
NaCl, 20 mM Tris–HCl (pH 7.5), 0.2% (vol/vol)
nd 5% (wt/vol) non-fat milk for 1 h. The next incuba-
rformed with a mouse monoclonal antibody against
S (1:1000, Transduction Laboratories, San Diego,
al incubation was carried out with anti-rabbit IgG
peroxidase (1:5000). Immunoreactive bands were
with a chemiluminescence detection kit (Amersham
s, Piscataway, NJ).
ls
ale Sprague-Dawley rats, weighing 250–280 g
Laboratory Animal Center, Taipei, Taiwan), were
rats were allowed to acclimate to environmentally
quarters with a constant temperature of 20–22 ◦C,
idity 55% and a light cycle of 12:12 h. Standard lab-
w (Purina Mills, Richmond, IN, USA) and drinking
provided ad libitum. All animal experiments were
y the Institutional Animal Care and Use Committees
l Research Institute of Chinese Medicine and were
in accordance with the National Institutes of Health
re standards.
lar tension experiment
thods employed were essentially the same as our
published methods (Wang et al., 1996; Ko et al.,
rief, aortic rings from Sprague-Dawley rats were
an chambers isometrically under passive tension of

224 G.J. Wang et al. / Journal of Ethnopharmacology 112 (2007) 221–227
1.8 g for 60 min. Functional integrity of the endothelium was
confirmed by an observation of more than 95% relaxation in
response to
phenylephr
1–4 on eN
choline (10
(considered
traction in
and recove
in the prese
vehicle (dim
of concent
on the perc
tion. A com
pre-contrac
2.10. Reag
The foll
dine, bovi
l-NNA, L
ethylenedia
amide, fro
USA); alam
illo, CA, U
Austria); p
bad, CA, U
iNOS from
Compound
stock solut
concentrati
0.1% and h
vascular te
2.11. Stati
For ea
mean ± S.E
performed
compatible
Ver. 10.0).
treatments
(ANOVA)
interaction
using the K
ANOVA. P
according t
tistically si
3. Results
A metha
fractionate
in n-hexan
Among the
significant
0.26%. Thi
by Sephadex LH-20 column chromatography and HPLC, and
each step was guided by a bioassay to give compounds 1–4. Of
1 and 2 were identified to be new based on their spectral
is.
molecular formula for 1, C10H8O5, was determined
NMR and HR-FAB-MS data. The IR spectrum of 1
ted the presence of a hydroxyl (3324 cm−1), a carbonyl
cm−1) and aromatic (1603 cm−1) groups. The 1H NMR
m of 1 (Table 1) displayed an ABX coupled aromatic
at δH 7.20 (1H, s), 7.18 (1H, d, J = 8.2 Hz) and 6.82 (1H,
8.2 Hz), ascribable to H-2′, -6′ and -5′, respectively, as
s one olefinic proton at δH 6.23 (1H, s), and one hemiac-
oton at δH 6.45 (1H, s) in association with its 13C NMR
was characteristic of an aromatic ring annexed to a con-
d γ-butyrolactone skeleton (Rasmussen et al., 1995; Li
2003). Thus, the structure of 1 (Fig. 1) was established
2,5-dihydro-5-hydroxy-4-[3′,4′-dihydroxyphenyl]-furan-
and named bracteanolide A.
pound 2 possessed spectroscopic data closely com-
e to that of 1 except that its OH-5 functionality in
replaced with a methoxy group in 2. Its 1H NMR
1) exhibited one signal for an additional methoxy sin-
δH 3.52 when compared with that of 1. Therefore, 2 was
ed as 2,5-dihydro-5-methoxy-4-[3′,4′-dihydroxyphenyl]-
2-one, and named bracteanolide B.
ctroscopic data of (+)-(R)-p-hydroxyphenyllactic acid (3)
been isolated from the metabolites of Ceratocystis spp.
. Cheacetylcholine (1M) in tissues pre-contracted with
ine (0.3M). To investigate the effects of compounds
OS function, cumulative concentrations of acetyl-
nM–10M) were applied during the sustained phase
as 100%) of phenylephrine (0.3M)-induced con-
endothelium-intact aortic rings. Following washing
ry for 30 min, similar experiments were carried out
nce of 1–4, l-NNA or aminoguanidine at 10M or
ethyl sulfoxide; 0.1%) for 20 min. The construction
ration–response curves for acetylcholine was based
entage of relaxation of the agonist-induced contrac-
plete relaxation was considered attained when the
ted rings returned to the base line position.
ents
owing drugs were used: acetylcholine, aminoguani-
ne serum albumin, dimethyl sulfoxide, DMEM,
PS (Escherichia coli Serotype 055:B5), napthyl-
mine dihydrochloride, phenylephrine and sulfanil-
m Sigma–Aldrich Chemical Co. (St. Louis, MO,
arBlue, from Biosource International Co. (Camar-
SA); FCS, from PAA Laboratories GmbH (Pasching,
enicillin–streptomycin from Invitrogen Co. (Carls-
SA); mouse monoclonal antibody against mouse
Transduction Laboratories (San Diego, CA, USA).
s 1–4 were dissolved in dimethyl sulfoxide to make
ions, respectively, and stored at −30 ◦C. The final
on of the vehicle in the solutions never exceeded
ad no effects on NO production, iNOS expression,
nsion or cell viability assay.
stic analysis
ch experimental series, data are presented as
. and n represents the number of independently
experiments. All data were analyzed by an IBM-
statistical software package (SPSS for Windows,
The significance of the concentrations and sample
was determined by two-way analysis of variance
with repeated measures. If there were significant
s, the simple main effect of each factor was assessed
ruskal-Wallis nonparametric post hoc analysis for
ost hoc comparisons were carried out between means,
o suitability. P-values less than 0.05 indicated a sta-
gnificant difference.
and discussion
nolic extract of the whole plant of M. bracteata was
d by liquid–liquid partitioning into fractions soluble
e, ethyl acetate, n-butanol, and H2O, successively.
m, the ethyl acetate-soluble fraction has the most
inhibitory activity on iNOS with an overall yield of
s fraction was further chromatographed sequentially
these,
analys
The
by 13C
indica
(1726
spectru
system
d, J =
well a
etal pr
data, 1
jugate
et al.,
to be
2-one,
Com
parabl
1 was
(Table
glet at
assign
furan-
Spe
having
Fig. 1 mical structure of compounds 1–4 from Murdannia bracteata.

G.J. Wang et al. / Journal of Ethnopharmacology 112 (2007) 221–227 225
Fig. 2. Inhibi
RAW 264.7 c
2–4 (100M
100M) or v
amount of NO
compared wit
to 30.60 ± 0.0
were interp
ature (Ayer
obtained as
with that o
catappa (L
To acce
264.7 cells
viability te
aBlue assa
revealed n
concentrati
ments. NO
in RAW26
and may p
on the infla
dine, a sel
NOS inhib
the effect
stimulated
LPS (50 ng
tectable lev
From these
264.7 cells
shown), bu
264.7 cells
icantly elic
which is a
compounds
with 91.60
inhibitors a
tions, the in
85.21 ± 0.87% and 46.16 ± 3.07%, respectively. Fig. 2 shows
that the new bracteanolide A (1, 12.5–100M) strongly inhib-
S-induced NO production in a concentration-dependent
r, with an IC50 value of 33.27 ± 0.86M. The observed
seems not to be related to the cytotoxicity of 1, since
ed no impairment of cell viability (Fig. 3). In order
ermine if the observed inhibitory effect of 1 on the
matory mediators was directly related to the modulation
S induction, we examined its protein expression lev-
Western blot analysis. As shown in Fig. 4A, 130 kDa
S protein expression was undetectable in unstimulated
264.7 cells. In response to LPS, the amount of iNOS was
dly up-regulated. At the indicated concentrations, 1 had an
nt suppressive effect on iNOS protein induction without
ng housekeeping protein expression. This effect was also
tration-dependent (Fig. 4B). The inhibitory effect of 1
S protein expression in activated RAW 264.7 cells could
t, at least in part, for the suppression of NO production.
r experiments were conducted for aminoguanidine and l-
as illustrated in Fig. 4A. Clearly, 1 is more potent than
of these two positive controls.
cteanolide B (2), an analogue of bracteanolide A, has no
ce on cell viability at concentrations up to 100M in
sence of LPS (Fig. 3). Different from bracteanolide A,
s a moderate inhibitory activity on NO production at the
al test concentration (Fig. 2). The nitrite level of 2 was
1.12% of cells treated with LPS plus vehicle group.
, at t
d ce
). Th
ctiva
rabletory effect of compounds 1–4 on LPS-induced NO production.
ells were co-incubated with LPS (50 ng/ml) and 1 (12.5–100M),
), Nω-nitro-l-arginine (l-NNA, 100M), aminoguanidine (AG,
ehicle for 24 h. NO production was determined by measuring the
metabolites in the medium. n = 6–8 in each group. *P < 0.05 when
h vehicle-treated cells. Vehicle group represented as 100% is equal
5M of NO produced in the medium per well of cells.
reted by comparison with those reported in the liter-
et al., 1986). Compound 4, a major component, was
a yellow powder whose spectral data was consistent
f isovitexin, having been isolated from Terminalia
in et al., 2000).
ss the possible cytotoxicity of 1–4 against RAW
, a mouse macrophage cell line, we conducted a cell
ited LP
manne
effect
it show
to det
inflam
of iNO
els by
of iNO
RAW
marke
appare
affecti
concen
on iNO
accoun
Simila
NNA
either
Bra
influen
the pre
2 exert
maxim
50.11 ±
Only 4
reduce
(Fig. 3
LPS-a
compast determined in the absence of LPS using the alam-
y. Up to a concentration of 100M, all compounds
o significant cytotoxicity (data not shown). Thus,
ons of 100M were chosen for subsequent experi-
production induced by LPS through iNOS expression
4.7 cells may reflect the degree of inflammation
rovide a measure for assessing the effects of drugs
mmatory process. In this experiment, aminoguani-
ective iNOS inhibitor, and l-NNA, a non-selective
itor, were used as positive inhibitors. To investigate
of 1–4 on NO production, RAW 264.7 cells were
with 100M of 1–4 in the absence or presence of
/ml). Without LPS, RAW 264.7 cells released unde-
els of NO after 24-h incubation (data not shown).
results, all the compounds failed to stimulate RAW
to produce detectable amounts of NO (data not
t inhibited NO production by LPS-activated RAW
to varying extents. As shown in Fig. 2, LPS signif-
ited the accumulation of 30.60 ± 0.05M of nitrite,
stable metabolite in the medium. Among the tested
, 1, at 100M, is the most potent in the assay,
± 1.51% inhibition, even better than the positive
minoguanidine and l-NNA. Under the same condi-
hibitory effects of aminoguanidine and l-NNA were
Fig. 3. Effec
264.7 cells. R
1 (12.5–100
aminoguanidi
by alamarBlu
vehicle-treatehe concentration used for NO inhibition, has weakly
ll viability by 9.25 ± 1.78% in the presence of LPS
is compound appeared to inhibit NO production in
ted RAW 264.7 cells (Fig. 2), but this effect was
with its cytotoxicity upon cells during culture in
t of compounds 1–4 on cell viability in LPS-activated RAW
AW 264.7 cells were co-incubated with LPS (50 ng/ml) and
M), 2–4 (100M), Nω-nitro-l-arginine (l-NNA, 100M),
ne (AG, 100M) or vehicle for 24 h. Cell viability was detected
e assay. n = 6–8 in each group. *P < 0.05 when compared with
d cells.

226 G.J. Wang et al. / Journal of Ethnopharmacology 112 (2007) 221–227
Fig. 4. The i
protein expre
from three s
24 h-LPS (50
(12.5–100M
100M) or v
by densitome
experiments.
the presenc
the effect o
tested conc
(Figs. 2 an
In addit
of NO ch
contributes
ations are
endothelium
derived h
Fleming et
from activ
whereas th
tance arteri
reports (Ku
dependent
wild-type e
inhibition o
is the majo
order to ex
of eNOS f
1–4 on ace
intact aorti
compound,
sion of the aortic rings (data not shown). Fig. 5 shows the
vasorelaxation effects of acetylcholine (10 nM 10M) in the
ce of 1, reference compounds or vehicle in aortic rings
ntracted with phenylephrine (0.3M). When the concen-
of acetylcholine reached 1.0M, a marked relaxation
ed (97.46 ± 0.81%) in the vehicle-treated group. The non-
ve NOS inhibitor l-NNA (10M) completely inhibited
holine-evoked vasorelaxation whereas aminoguanidie, a
ve iNOS inhibitor, did not change the effect induced
tylcholine. In the presence of 1 (10M) for 20 min,
orelaxation induced by acetylcholine was not obviously
d. The maximal relaxation of this compound was 100%,
was similar to aminoguanidine (10M) yielding maxi-
laxation of 98.08 ± 1.92%. These findings suggest that 1,
l-NNA, did not modify the activity of eNOS. This com-
exhibits significant and comparable selectivity for the
ion of iNOS and may be developed as a potential therapeu-
tegy for anti-inflammation, without changes in vascular
and systemic blood pressure. Similar to 1, compounds
d not change the endothelial NO-dependent relaxation
d by acetylcholine, an index of eNOS acitivity (data not
).
overall experimental results of the present study sug-
at the constituents, two hydroxybutenolides as well as
xin, isolated from M. bracteata, substantially inhibit NO
tion in LPS-stimulated RAW 264.7 murine macrophages.nhibitory effect of bracteanolide A (1) on LPS-induced iNOS
presen
pre-co
tration
occurr
selecti
acetylc
selecti
by ace
the vas
affecte
which
mal re
unlike
pound
inhibit
tic stra
tension
2–4 di
induce
shown
The
gest th
isovite
producssion in RAW 264.7 cells. (A) A representative immunoblot
eparate experiments is shown. Lysates were prepared from
ng/ml)-stimulated cells in combination with bracteanolide A
), Nω-nitro-l-arginine (l-NNA, 100M), aminoguanidine (AG,
ehicle. (B) Results were generated as integrated intensity units
try and expressed as percentage of vehicle from three separate
*P < 0.05 when compared with vehicle-treated cells.
e of LPS. In contrast to the other three compounds,
f 3 seems not to be linked to NO production at the
entrations, although it did not affect cell viability
d 3).
ion to the inflammatory iNOS pathway, basal release
iefly derived from constitutively activated eNOS
to vasorelaxation. Endothelium-dependent relax-
achieved by a combination of nitric oxide (NO),
-derived prostacyclin (PGI2) and endothelium-
yperpolarizing factor (EDHF) (Rubanyi, 1993;
al., 1996). The tonical contribution of NO, derived
ation of eNOS, is most prominent in the aorta,
at of EDHF is most prominent in the distal resis-
es (Shimokawa et al., 1996). According to previous
et al., 1996; Chataigneau et al., 1999), endothelium-
relaxation induced by acetylcholine in the aorta from
NOS(+/+) mice was completely abolished by acute
f NOS indicating that, in large-conduit vessels, NO
r endogenous endothelium-derived vasodilator. In
plore whether 1–4 has an effect on modification
unction, besides iNOS, we examined the effects of
tylcholine-evoked relaxation in isolated endothelium
c rings. In the vasorelaxation bioassay, each tested
given individually did not alter the baseline ten-
The new compound bracteanolide A (1) is the most potent. Its
dramatic in
decreased e
iNOS is se
endothelial
activity. Al
Fig. 5. The v
presence of
aminoguanidi
endothelium p
*P < 0.05 whehibitory effect on NO synthesis is associated with its
xpression of iNOS protein. The modulating effect on
lective, since it did not effect acetylcholine-evoked
NO-dependent vasorelaxation an index of eNOS
l of these findings seem to provide a rationale for the
asorelaxation effects of acetylcholine (10 nM–10M) in the
compound 1 (10М), Nω-nitro-l-arginine (l-NNA, 10M),
ne (AG, 10M) or vehicle for 20 min in rat aortic rings with intact
recontracted with phenylephrine (0.3M). n = 6–8 in each group.
n compared with vehicle-treated cells.

G.J. Wang et al. / Journal of Ethnopharmacology 112 (2007) 221–227 227
anti-inflammatory use of M. bracteata, which might be linked
with its ability to reduce NO production and immunoregulatory
function.
4. Conclusion
To our knowledge, this is the first report on the scientific
rationale of M. bracteata for anti-inflammatory medicinal use.
The specific iNOS inhibitory effect of 1 could be potentially
developed as a selective inhibitor of iNOS for future therapeutic
use.
Acknowledgements
This research was supported, in part, by grants to G.J.W.
(NSC 95-2320-B-077-001, 95-DBCMR-04 and CCMP 95-RD-
201) and T.H.L. (NSC 94-2113-M-038-003). The authors would
like to thank Dr. Loren Wold for proofreading and editing the
manuscript, and Ms. Shwu-Huey Wang and Ms. Shou-Ling
Huang for t
ter of Taipe
the College
Reference
Ayer, W.A., B
The chem
cystis spec
Canadian
Bradford, M.
microgram
ing. Analy
Chartrain, N.A
E.P., Billia
structure,
synthase g
Chataigneau,
houtte, P.M
endothelia
cology 12
Chen, Y.C., L
species an
Cellular P
Chiu, N.Y., Chang, K.H. (Eds.), 1995. The Illustrated Medicinal Plants of Taiwan
(VI). SMC Publishing Inc., Taipei, p. 256.
Fleming, I., Bauersachs, J., Busse, R., 1996. Paracrine functions of the coronary
vascular endothelium. Molecular and Cellular Biochemistry 157, 137–145.
Green, L.C., Wagner, D.A., Glogowski, J., Wishnik, I.S., Tannenbaum, S.R.,
1982. Analysis of nitrate, nitrite and [15N]nitrate in biologic fluids. Analyt-
ical Biochemistry 126, 131–138.
Hibbs Jr, J.B., Taintor, R.R., Vavrin, Z., 1987. Macrophage cytotoxicity: role for
l-arginine deiminase and imino nitrogen oxidation to nitrite. Science 235,
473–476.
Hong, D.Y., 1997. Angiospermae: monocotyledoneae. In: Wu, K.F. (Ed.), Flora,
Reipublicae Popularis Sinicae. Delectis Florae Reipublicae Popularis Sini-
cae Agendae Academiae Sinicae Edita, Tomus, 13. Science Press, Beijing,
pp. 109–110.
Jiang, J.S., Shih, C.M., Wang, S.H., Chen, T.T., Lin, C.N., Ko, W.C., 2006. Mech-
anisms of suppression of nitric oxide production by 3-O-methylquercetin in
RAW 264.7 cells. Journal of Ethnopharmacology 103, 281–287.
Ko, H.C., Chen, K.T., Chen, C.F., Su, J.P., Chen, C.M., Wang, G.J., 2006.
Chemical and biological comparisons on Evodia with two related species
of different locations and conditions. Journal of Ethnopharmacology 108,
257–263.
Ku, D., Guo, L., Dai, J., Acuff, C.G., Steinhelper, M.E., 1996. Coronary vascular
and endothelial reactivity changes in trans-genic mice overexpressing atrial
uretic
K., Ly
ecular
, Den
tituen
681.
., Kuo
from
53–2
sen, S
tituen
i, G.M
diseas
awa, H
akay
. The
el size
lation
.J., S
vasor
lar ca
. The
–102
., 199
lizatiohe NMR data acquisition in the Instrumentation Cen-
i Medical University and Instrumentation Center of
of Science, National Taiwan University, respectively.
s
rowne, L.M., Feng, M.C., Orszanska, H., Saeedi-Ghomi, H., 1986.
istry of the blue stain fungi. Part 1. Some metabolites of Cerato-
ies associated with mountain pine beetle infected lodgepole pine.
Journal of Chemistry 64, 904–909.
M., 1976. A rapid and sensitive method for the quantitation of
quantities of protein utilizing the principle of protein–dye bind-
tical Biochemistry 72, 248–254.
., Geller, D.A., Koty, P.P., Sitrin, N.F., Nussler, A.K., Hoffman,
r, T.R., Hutchinson, N.I., Mudgett, J.S., 1994. Molecular cloning,
and chromosomal localization of the human inducible nitric oxide
ene. Journal of Biological Chemistry 269, 6765–6772.
T., Feletou, M., Huang, P.L., Fishman, M.C., Duhault, J., Van-
., 1999. Acetylcholine-induced relaxation in blood vessels from
l nitric oxide synthase knockout mice. British Journal of Pharma-
6, 219–226.
in-Shiau, S.Y., Lin, J.K., 1998. Involvement of reactive oxygen
d caspase 3 activation in arsenite induce apoptosis. Journal of
hysiology 177, 323–324.
natri
Kwack,
Mol
Li, L.Y.
cons
680–
Lin, Y.L
side
47, 2
Rasmus
cons
Rubany
and
Shimok
Y., T
1996
vess
circu
Wang, G
The
cellu
cells
1016
Wolf, G
locafactor. American Journal of Physiology 271, H2368–H2376.
nch, R.G., 2000. A new non-radioactive method for IL-2 bioassay.
Cell 10, 575–578.
g, Z.W., Fu, H.Z., Li, J., Proksch, P., Lin, W.H., 2003. Chemical
ts from the marine sponge Iotrochoto birotulata. Pharmazie 58,
, Y.H., Shiao, M.S., Chen, C.C., Ou, J.C., 2000. Flavonoid glyco-
Terminalia catappa L. Journal of the Chinese Chemical Society
56.
., Wolff, C., Rudolph, H., 1995. Compartmentalization of phenolic
ts in Sphagnum. Phytochemistry 38, 35–39.
., 1993. The role of endothelium in cardiovascular homeostasis
es. Journal of Cardiovascular Pharmacology 22 (Suppl 4), S1–S14.
., Yasutake, H., Fujii, K., Owada, M.K., Nakaike, R., Fukumoto,
anagi, T., Nagao, T., Egashira, K., Fujishima, M., Takeshita, A.,
importance of the hyperpolarizing mechanism increases as the
decreases in endothelium-dependent relaxations in rat mesenteric
. Journal of Cardiovascular Pharmacology 28, 703–711.
han, J., Pang, P.K., Yang, M.C., Chou, C.J., Chen, C.F., 1996.
elaxing action of rutaecarpine: direct paradoxical effects on intra-
lcium concentration of vascular smooth muscle and endothelial
Journal of Pharmacology and Experimental Therapeutics 276,
1.
7. Nitric oxide and nitric oxide synthase: biology, pathology,
n. Histology and Histopathology 12, 251–261.