Inhibitory effects of furanocoumarins isolated from the roots of Angelica dahurica on prostaglandin E2 production

Article abstract of DOI:10.1055/s-2003-39702

We have isolated five furanocoumarins, byakangelicin, phellopterin, imperatorin, isoimperatorin, and oxypeucedanin methanolate, from the roots of Angelica dahurica (Umbelliferae) and prepared five semi-synthesized compounds by the partial reduction of eac

Full text of DOI:10.1055/s-2003-39702

Hyun Seung Ban1
Soon Sung Lim2
Katsuya Suzuki1
Sang Hoon Jung2
Sanghyun Lee2
Yeon Sil Lee2
Inhibitory Effects of Furanocoumarins
Isolated from the Roots of Angelica dahurica
KukHyun Shin 2
Kazuo Ohuchi1
on Prostaglandin E Production
2
Abstract
potent inhibitory activity on the LPS-induced PGE production.
2
It also inhibited the LPS-induced expressions of cyclooxygenase
We have isolated five furanocoumarins, byakangelicin, phellop- (COX)-2 and microsomal prostaglandin E synthase (mPGES).
terin, imperatorin, isoimperatorin, and oxypeucedanin methano- These findings suggest that the inhibitory effect of furanocou-
late, from the roots of Angelica dahurica (Umbelliferae) and pre- marins on the LPS-induced PGE production is due to the inhibi-
2
pared five semi-synthesized compounds by the partial reduction tion of the expression of COX-2 and mPGES.
of each isolated furanocoumarin, and the effects of these com-
pounds on lipopolysaccharide (LPS)-induced prostaglandin E2 Key words
(
PGE ) production in rat peritoneal macrophages were exam- Angelica dahurica ´ imperatorin ´ prostaglandin E ´ cyclooxygen-
2 2
ined. Among these compounds, imperatorin showed the most ase ´ microsomal prostaglandin E synthase
Introduction
and microsomal PGES (mPGES) are induced by various proin-
flammatory stimulators [5], [6], [7].
Angelica dahurica Benth. et Hook. (Umbelliferae) is one of the
4
08
important medicinal plants in Korea and the roots of which In this study, to find lead compounds for anti-inflammatory
have traditionally been used to reduce headache caused by drugs from natural products, we examined the effects of furano-
cold, toothache, neuralgia and hemorrhage in Chinese medicine coumarins isolated from the roots of Angelica dahurica and semi-
[
1].
synthesized compounds from the isolated furanocoumarins on
lipopolysaccharide (LPS)-induced PGE production in rat perito-
2
The activated macrophages produce prostaglandin (PG) E2 neal macrophages.
which causes pain and fever during the inflammatory process.
PGE synthesis from the released arachidonic acid from mem-
2
brane phospholipids is controlled by two enzymes. One is Materials and Methods
cyclooxygenase (COX) which converts arachidonic acid to
PGH , and the other is PGE synthase (PGES) which converts Isolation of furanocoumarins from the roots of
2
PGH2 to PGE . It is reported that there are two types of COX Angelica dahurica
2
and PGES, respectively [2], [3]. COX-1 and cytosolic PGES The roots of Angelica dahurica Benth. et Hook were purchased
(
sPGES) are the constitutively expressed forms [4], and COX-2 from the local market (Kyung-Dong-Si-Jang) in Seoul, Korea in
Affiliation
1
Laboratory of Pathophysiological Biochemistry, Graduate School of Pharmaceutical Sciences,
Tohoku University, Sendai, Miyagi, Japan
2
Natural Products Research Institute, Seoul National University, Seoul, Korea
Correspondence
Prof. Dr. Kazuo Ohuchi ´ Laboratory of Pathophysiological Biochemistry ´
Graduate School of Pharmaceutical Sciences ´ Tohoku University ´ Aoba Aramaki ´ Aoba-ku ´ Sendai ´
Miyagi 980-8578 ´ Japan ´ Phone: +81-22-217-6860 ´ Fax: +81-22-217-6859 ´
E-mail: ohuchi-k@mail.pharm.tohoku.ac.jp
Received August 9, 2002 ´ Accepted December 15, 2002
Bibliography
Planta Med 2003; 69: 408±412 ´ ꢀ Georg Thieme Verlag Stuttgart ´ New York ´ ISSN 0032-0943

September 2000 and were botanically identified by Prof. H. J. Chi in (6), the side chain of the furanocoumarin, 2,3-dihydroxy-3-me-
at Natural Products Research Institute, Seoul National University, thylbut-2-enyloxy, was removed by the Pd/C catalytic hydrogena-
Korea where the voucher specimen has been deposited (Voucher tion process.
No.: NPRI-2013). The coarsely powdered plant material (3 kg)
was extracted three times with methanol under reflux over a to- Macrophage culture
tal 4 h period. The resultant extracts were combined and concen- Peritoneal cells were harvested from Sprague-Dawley rats ac-
trated under reduced pressure to afford 520 g of residue. The me- cording to the procedure described previously [10]. The perito-
thanol extract was partitioned between n-hexane: methanol:- neal cells were suspended in Eagle's minimal essential medium
water (10:1:9), and the residue was then partitioned between (EMEM, Nissui, Tokyo, Japan) containing 10% calf serum (Flow
ethyl acetate and water. Column chromatography of the n-hex- Laboratories, North Rydge, Australia), penicillin G potassium
ane fraction (30.4 g) over silica gel (70±230 mesh, Art, 7734 (18 mg/mL) and streptomycin sulfate (50 mg/mL) (Meiji Seika, To-
5
Merck) using n-hexane-diethyl ether as a stepwise gradient elu- kyo, Japan), then seeded at a density of 7.5”10 cells/0.5 mL/well
tion (0±100% of diethyl ether) gave phellopterin (4, 0.98 g), im- in 24-well plastic tissue culture dishes (Corning Glass Works,
6
peratorin (6, 2.88 g) and isoimperatorin (8, 1.68 g). The ethyl Corning, NY, USA) or 3.0”10 cells/2 mL/well in 6-well plastic tis-
acetate fraction (75.3 g) obtained by removal of the solvent was sue culture dishes (Corning Glass Works), and incubated at 378C
column chromatographed on silica gel (70±230 mesh, Art, 7734 for 2 h. The wells were then washed three times with medium to
Merck) using the methylene chloride-methanol as a stepwise remove non-adherent cells, and the adherent cells were further
gradient elution (0±20% of methanol) to give byakangelicin (1, incubated at 378C for 20 h. After three washes, the adherent cells
2
0
3
.69 g), [a] : + 17.58 (c 0.10, C H N) and oxypeucedanin metha- were used for subsequent experiments.
D 5 5
20
nolate (9, 0.11 g), [a] : + 18.08 (c 0.20, CHCl ). The five com-
D
3
pounds isolated from the dried roots of Angelica dahurica (Um- Incubation of macrophages with drugs
belliferae) (1, 4, 6, 8 and 9 in Table 2) were identified by compar- The adherent cells were incubated at 378C for 8 h in 0.5 mL of
ison of the spectral data (NMR, IR, UV) with authentic com- medium containing 10% calf serum and the indicated concen-
pounds [8], [9].
trations of each furanocoumarin, indomethacin (Sigma Chemi-
cal Co., St. Louis, MO, USA) or dexamethasone (Sigma). Each fur-
anocoumarin was dissolved in dimethyl sulfoxide (DMSO) and
added to the medium. Indomethacin and dexamethasone were
Semi-synthesization of derivatives from the isolated
furanocoumarins
By the semi-synthetic methods, 2¢,3¢-dihydrobyakangelicin (2), dissolved in ethanol and added to the medium. The final con-
2
0
[
(
a] : + 18.58 (c 0.30, C H N), 2¢,3¢,3¢,4-tetrahydrobyakangelicin centration of DMSO and ethanol was adjusted to 0.1% (v/v),
D 5 5
3), [a]20: + 17.58 (c 0.30, C H N), 2¢,3¢-dihydro-8-hydroxybergap- respectively. Control medium contained the same amount of
D
5
5
ten (5), 2¢,3¢-dihydro-8-hydroxypsoralen (7) and 2¢,3¢-dihydrooxy- DMSO and ethanol.
2
0
peucedanin methanolate (10), [a] : + 19.08 (c 0.25, CHCl ) were
D
3
synthesized. Furanocoumarins have one double bond in each fur- Measurement of PGE concentrations
2
409
an ring and pyrano-2-one ring, and a carbon-carbon double bond After incubation, the conditioned medium was collected, and
of furanocoumarins is selectively hydrogenated under the action centrifuged at 1500”g and 48C for 5 min. The concentration of
of a molar ratio of 10% palladium/charcoal (Pd/C). For the selective PGE2 in the supernatant was then radioimmunoassayed [10].
hydrogenation of the double bond in the furan ring, Pd/C and fur- PGE antiserum was purchased from Assay Designs (Ann Arbor,
2
anocoumarin at 1:1 molar ratio were dissolved in ethanol (1 g/50 MI, USA).
mL ethanol) and vigorously mixed for 30 min at room temperature
under H gas (1 atm) (Fig.1, process a). For the hydrogenation of Western blotting analysis of COX-1, COX-2, cPGES and mPGES
2
the two double bonds, Pd/C and furanocoumarin were dissolved The peritoneal cells (3”106 cells) collected as described above
in ethanol at 2:1 molar ratio (Fig.1, process b), and hydrogenated were incubated at 378C for 8 h in 2 mL medium containing drugs.
as described above. In the cases of phellopterin (4) and imperator- Cytosolic fractions were then prepared, and were applied on
SDS-polyacrylamide gel (8% for COX and 14% for PGES). Immu-
noblotting was carried out as described previously [11] using
COX-1 antibody (Santa Cruz Biotechnology, CA, USA), COX-2 an-
tibody (Santa Cruz Biotechnology), cPGES antibody (Cayman
Chemical Co., Ann Arbor, MI, USA) and mPGES antibody (Cayman
Table 1 Effects of the fractions from the roots of Angelica dahurica on
LPS-induced PGE production in rat peritoneal macrophages
2
Treatment
PGE
2
(ng/ml)
Chemical Co.).
None
2.07  0.15***
8.40  0.15
Statistical analysis
LPS (0.1 mg/ml)
The statistical significance of the results was analyzed by Dun-
nett's test for multiple comparisons and Student's t-test for un-
paired observations.
LPS + methanol extract (30 mg/ml)
LPS + n-hexane fraction (30 mg/ml)
LPS + methylene chloride fraction (30 mg/ml)
LPS + ethyl acetate fraction (30 mg/ml)
LPS + water fraction (30 mg/ml)
LPS + indomethacin (0.1 mM)
5.61 0.21***
2.53  0.11***
1.96  0.13***
6.54  0.28**
8.40  0.06
Results and Discussion
1.89  0.17***
Incubation with LPS (0.1 mg/mL) for 8 h prominently increased
Values are the means from four samples with s.e.m.
Statistical significance; **P < 0.01, ***P < 0.001 vs. LPS control.
PGE production in rat peritoneal macrophages (Table 1). From
2
Ban HS et al. Inhibitory Effects of¼ Planta Med 2003; 69: 408±412

Table 2 Chemical structures of five furanocoumar-
ins isolated from the roots of Angelica dahurica
the methanol extract of the roots of Angelica dahurica, n-hexane, pendent manner, and at 3 mM, PGE production was suppressed
2
methylene chloride, ethyl acetate and water fractions were ob- to 49% of the LPS control. The release of radioactivity to the
3
tained, and the effects of these fractions at 30 mg/mL on the LPS- medium from [ H]arachidonic acid-labeled macrophages [12]
induced PGE production were examined. The methanol extract was increased by LPS (0.1 mM) treatment from 1 to 4 h, but im-
2
and all the fractions except for the water fraction significantly in- peratorin (6) at 3 to 30 mM did not inhibit the LPS-induced re-
hibited the LPS-induced PGE production (Table 1). Among these lease of radioactivity (data not shown). These findings suggest
2
fractions, the methylene chloride fraction showed the most po- that the inhibition of the LPS-induced PGE production by im-
2
tent inhibitory activity, and followed by the n-hexane fraction.
The inhibitory effect of the methylene chloride fraction at 30 mg/ that catalyzes the release of arachidonic acid from membrane
mL was almost the same as that of indomethacin at 0.1 mM (Ta- phospholipids.
peratorin (6) is not due to the inhibition of phospholipase A2
4
10
ble 1).
Treatment with LPS (0.1 mg/mL) for 8 h prominently induced the
From the methylene chloride fraction and the n-hexane fraction, expression of COX-2 and mPGES (Fig. 3A). In the presence of dex-
we isolated five furanocoumarins (Table 2) and prepared five amethasone (0.1 mM), both the LPS-induced increases in the lev-
semi-synthetic compounds from the isolated furanocoumarins els of COX-2 and mPGES were suppressed. Imperatorin (6) at 3 to
(
Fig.1). Among the isolated furanocoumarins and semi-synthe- 30 mM also suppressed the LPS-induced expression of COX-2 and
sized compounds, phellopterin (4), imperatorin (6) and isoim- mPGES in a concentration-dependent manner (Fig. 3A), at which
peratorin (8) inhibited the LPS-induced PGE2 production more concentrations the LPS-induced PGE production was also inhib-
2
effectively than the other compounds at 30 mM. The inhibitory ited concentration-dependently (Fig. 3B). The levels of COX-1
effects of these compounds at 30 mM were almost the same as and cPGES were not affected by treatment with LPS alone or
that of indomethacin at 0.1 mM (Fig. 2). The inhibitory activity of with LPS and imperatorin (6) (Fig. 3A). These findings suggest
the semi-synthesized compounds (2, 3, 5, 7 and 10) did not ex- that the inhibitory effect of imperatorin (6) on the LPS-induced
ceed over that of the isolated furanocoumarins. For example, PGE2 production is due to the inhibition of the expression of
phellopterin (4) showed more potent inhibitory activity than both COX-2 and mPGES.
2¢,3¢-dihydro-8-hydroxybergapten (5), a partially reduced form
of phellopterin. These findings suggest that double bonds in fur- COX-2 gene is regulated by transcription factors such as NF-kB,
an ring and pyran-2-one ring, and R and R substitutions seem AP-2 and C/EBP [13], however, the regulation mechanism of
1
2
to be important for the expression of the inhibitory activity on mPGES gene expression remains to be clarified. From our find-
the LPS-induced PGE production.
ings, it is suggested that imperatorin (6) inhibits the activation
of transcription factors. It is reported that imperatorin (6) inhi-
2
To clarify the mechanism for the inhibition of PGE production bits LPS-induced nitric oxide (NO) production through the sup-
2
by furanocoumarins, we examined the effects of imperatorin (6) pression of inducible NO synthase (iNOS) expression in murine
on the LPS-induced increase in the levels of COX-2 and mPGES. macrophage cell line RAW 264.7 [14]. Because inductions of
As shown in Fig. 3B, imperatorin (6) at 3, 10 and 30 mM inhibited COX-2 and iNOS are regulated by similar transcription factors,
the LPS-induced PGE production at 8 h in a concentration-de- the mechanism for the suppression of COX-2 expression by im-
2
Ban HS et al. Inhibitory Effects of¼ Planta Med 2003; 69: 408±412

Fig. 1 Preparation of semi-synthetic compounds
from the isolated furanocoumarins. a. H , Pd/C 1
2
eq., 30 min, b. H , Pd/C 2 eq., 1h.
2
411
sion of COX-2 and mPGES. Our preliminary experiments revealed
that imperatorin (6) has no direct inhibitory activity on isolated
COX-1 and COX-2 as Liu et al. [15] reported that isoimperatorin
(8) showed no effect on COX-1 activity. Therefore, the inhibition
of the LPS-induced PGE production by these furanocoumarins
2
might not be due to the direct inhibition of COX-1 and COX-2.
In conclusion, among the five isolated furanocoumarins from the
roots of Angelica dahurica and five semi-synthesized compounds,
phellopterin (4), imperatorin (6) and isoimperatorin (8) showed
the potent inhibitory activity on LPS-induced PGE production.
2
Furthermore, the inhibition of the LPS-induced PGE production
2
by imperatorin (6) was suggested to be due to the inhibition of
Fig. 2 Effects of the isolated and semi-synthesized furanocoumarins the induction of COX-2 and mPGES. These furanocoumarins
on LPS-induced PGE production in rat peritoneal macrophages. Rat
peritoneal macrophages (7.5”10 cells) were incubated at 378C for
2
might be the lead compounds that inhibit PGE production by
suppressing the induction of COX-2 and mPGES.
2
5
8
h in 0.5 mL of medium containing each compound (30 mM) or indo-
methacin (IM, 0.1 mM) in the presence of LPS (0.1 mg/mL). PGE con-
2
centrations in the conditioned medium were radioimmunoassayed.
Values are the means from four samples with s.e.m. shown by vertical
bars. Statistical significance; **P < 0.01, ***P < 0.001 vs. LPS control.
Acknowledgements
This work was supported in part by the Joint Research Program
peratorin (6) might be the same as that for the suppression of under the Japan-Korea Basic Scientific Cooperation Program
iNOS expression.
(JSPS 147 to K.O., KOSEF 996-0700-003-2 to K.H.S.).
Further experiment is necessary to clarify the action mechanism
of furanocoumarins for the inhibition of the LPS-induced expres-
Ban HS et al. Inhibitory Effects of¼ Planta Med 2003; 69: 408±412

References
1
Encyclopedia of Traditional Chinese Materia Medica. Shanghai, China:
The people's press, 1977; Vol. IV: 4534
Smith WL, Marnett LJ. Prostaglandin endoperoxide synthase: struc-
2
ture and catalysis. Biochimica et Biophysica Acta 1991; 1083: 1±17
His LC, Hoganson CW, Babcock CT, Smith WL. Characterization of a
3
tyrosyl radical in prostaglandin endoperoxide synthase-2. Biochem-
ical and Biophysical Research Communications 1994; 202: 1592±8
O'Banion MK, Winn VD, Young DA. cDNA cloning and functional activ-
4
ity of a glucocorticoid-regulated inflammatory cyclooxygenase. Pro-
ceedings of the National Academy of Sciences of the United States of
America 1992; 89: 4888±92
5
O'Sullivan MG, Chilton FH, Huggins EM Jr, McCall CE. Lipopolysacchar-
ide priming of alveolar macrophages for enhanced synthesis of prosta-
noids involves induction of a novel prostaglandin H synthase. Journal
of Biological Chemistry 1992; 267: 14547±50
6
Nishizuka Y. Intracellular signaling by hydrolysis of phospholipids and
activation of protein kinase C. Science 1992; 258: 607±14
Appleby SB, Ristimaki A, Neilson K, Narko K, Hla T. Structure of the hu-
7
man cyclo-oxygenase-2 gene. Biochemical Journal 1994; 302: 723±7
Saiki Y, Morinaga K, Okegawa O, Sakai S, Amaya Y. On the coumarins of
8
the roots of Angelica dahurica Benth. et Hook. Yakugaku Zasshi 1971;
91: 1313±7
9
Baek NI, Ahn EM, Kim HY, Park YD. Furanocoumarins from the root of
Angelica dahurica. Archives of Pharmacal Research 2000; 23: 467±70
Ohuchi K, Watanabe M, Yoshizawa K, Tsurufuji S, Fujiki H, Suganuma
M, Sugimura T, Levine L. Stimulation of prostaglandin E2 production
by 12-O-tetradecanoylphorbol 13-acetate (TPA)-type and non-TPA-
type tumor promoters in macrophages and its inhibition by cyclohex-
imide. Biochimica et Biophysica Acta 1985; 834: 42±7
1
0
Fig. 3 Effects of various concentrations of imperatorin (6) on LPS-in-
duced expressions of COX-1, COX-2, cPGES and mPGES, and PGE pro-
11
12
2
Kim YP, Yamada M, Lim SS, Lee SH, Ryu N, Shin KH, Ohuchi K. Inhibi-
6
duction. (A) Rat peritoneal macrophages (3”10 cells) were incubated
tion by tectorigenin and tectoridin of prostaglandin E production and
2
at 378C for 8 h in 2 mL of medium containing the indicated concentra-
tions of imperatorin or dexamethasone (DEX, 0.1 mM) in the presence
of LPS (0.1 mg/mL). The protein levels of each COX and PGES were de-
cyclooxygenase-2 induction in rat peritoneal macrophages. Biochimi-
ca et Biophysica Acta 1999; 1438: 399±407
Ohuchi K, Sugawara T, Watanabe M, Hirasawa N, Tsurufuji S, Fujiki H,
Christensen SB, Sugimura T. Analysis of the stimulative effect of thap-
sigargin, a non-TPA-type tumour promoter, on arachidonic acid me-
tabolism in rat peritoneal macrophages. British Journal of Pharmacol-
ogy 1988; 94: 917±23
Devaux Y, Seguin C, Grosjean S, de Talance N, Camaeti V, Burlet A, Zan-
nad F, Meistelman C, Mertes PM, Longrois D. Lipopolysaccharide-in-
duced increase of prostaglandin E2 is mediated by inducible nitric
oxide synthase activation of the constitutive cyclooxygenase and in-
duction of membrane-associated prostaglandin E synthase. Journal of
Immunology 2001; 167: 3962±71
Wang CC, Chen LG, Yang LL: Inducible nitric oxide synthase inhibitor
of the Chinese herb I. Saposhnikovia divaricata (Turcz.) Schischk. Can-
cer Letters 1999; 145: 151±7
Liu JH, Zschocke S, Reininger E, Bauer R. Inhibitory effects of Angelica
pubescens f. biserrata on 5-lipoxygenase and cyclooxygenase. Planta
Medica 2000; 64: 525±9
termined by Western blotting analysis. (B) Rat peritoneal macrophages
5
(
7.5”10 cells) were incubated at 378C for 8 h in 0.5 mL of medium
containing the indicated concentrations of imperatorin or DEX in the
presence of LPS. PGE2 concentrations in the conditioned medium
were radioimmunoassayed. Values are the means from four samples
with s.e.m. shown by vertical bars. Statistical significance; ***P <
13
4
12
0
.001vs. LPS control.
14
15
Ban HS et al. Inhibitory Effects of¼ Planta Med 2003; 69: 408±412