Biosci. Biotechnol. Biochem.

, 76 (5), 10281031, 2012

Note

Eective Cytochrome P450 (CYP) Inhibitors Isolated

from Tarragon (Artemisia dracunculus)
Zeineb B RAHMI, Tatsuya K ATHO, Rie H ATSUMATA, Asako H IROI, Nami M IYAKAWA,
Emi Y AKOU, Kouichi SUGAYA, Jun-ichi O NOSE, and Naoki A BEy
Department of Nutritional Science, Faculty of Applied Bio-Science, Tokyo University of Agriculture,
1-1-1 Sakuragaoka, Setagaya-ku, Tokyo 156-8502, Japan
Received January 6, 2012; Accepted January 11, 2012; Online Publication, May 7, 2012
[doi:10.1271/bbb.120006]

Two eective cytochrome P450 (CYP) inhibitors were

isolated from tarragon, Artemisia dracunculus. Their
structures were spectroscopically identied as 2E,4Eundeca-2,4-diene-8,10-diynoic acid isobutylamide (1)
and 2E,4E-undeca-2,4-diene-8,10-diynoic acid piperidide (2). Both compounds had dose-dependent inhibitory eects on CYP3A4 activity with IC50 values of
10:0  1:3 M for compound 1 and 3:3  0:2 M for
compound 2, and exhibited mechanism-based inhibition.
This is the rst reported isolation of eective CYP
inhibitors from tarragon (Artemisia dracunculus) purchased from a Japanese market.
Key words:

2E,4E-undeca-2,4-diene-8,10-diynoic acid
isobutylamide;
2E,4E-undeca-2,4-diene8,10-diynoic acid piperidide; cytochrome
P450 inhibitor; tarragon; Artemisia dracunculus

Tarragon (estragon), Artemisia dracunculus, is a

perennial herb with a long history of medicinal and
culinary use.1) In particular, it is a crucial spice for
French cuisine. The ethanolic extract of A. dracunculus
has signicantly decreased the blood glucose level and
improved the insulin level in both genetic and chemical
murine models of hyperglycemia.2,3) It had been
demonstrated in Caco-2 cells that the extract also had
potent inhibitory activity against allergen absorption.4)
Cytochrome P450 monooxygenases (CYPs), comprising a large and ubiquitous enzyme family, are hemecontaining mixed-function oxidases that play a key role
in the hepatic metabolism of lipophilic endogenetic
substrates such as hormones, and of ingested hydrophobic xenobiotics such as medicinal chemicals and
natural products. CYPs generally convert a large number
of exogenous compounds (toxic, carcinogenic, and most
pharmaceutical agents) to less toxic compounds. CYP
enzymes, including CYP3A4, CYP1A2, CYP2C9, and
CYP2D6, are mainly expressed in the human liver and
gastrointestinal tract and metabolize many pharmacologic agents and chemicals.5) The co-administration of
foods that contain CYP inhibitors and drugs often causes
adverse eects (food-drug interaction).68) The extrahepatic expression of CYPs has also been established.
Such CYP families as CYPs 1, 2 and 3 are involved in

the metabolism of carcinogenic, pro-carcinogenic, and

chemotherapeutic substances, and are found in several
extrahepatic tissues. CYPs 17 and 19 are involved in the
metabolism required for respectively producing androgens and estrogens, and are located in the testes, ovaries,
and adrenals.9) Food materials with selective inhibitors
against specic CYPs are therefore potential candidates
for cancer prevention strategies.
To elucidate the food factors in aromatic and
medicinal plants that either inhibit or exploit CYP
enzymes for the treatment of cancer, we evaluated the
dry material of tarragon, A. dracunculus, purchased in a
Japanese market (S&B Foods, Tokyo, Japan) for the
compounds active against CYPs. We report here the
isolation, identication, and inhibitory activities of two
active compounds, 1 and 2, against CYPs 1A2, 2C9,
2D6, and 3A4.
The dry material of A. dracunculus (200 g), which
was purchased from a Tokyo market, was soaked in 80%
acetone (2 L) at room temperature for 24 h. The resulting
ltrate was evaporated in vacuo to an aqueous concentrate and then extracted with three equal volumes of
ethyl acetate (EtOAc) at pH 3.0 to give an EtAOc
extract (21.6 g, IC50 30 mg/mL). This EtOAc extract
was applied to chromatography in a Sephadex LH-20
column (30 mm

700 mm), using methanol as the
eluent. Fifteen 20-mL fractions were collected to give
2.3 g of active fractions 710. These active fractions
were chromatographed by preparative HPLC (Capcell
Pak C18 UG120 column, 15 mm

250 mm, Shiseido,
Tokyo, Japan), using 0.05% TFA-CH3 CN (6:4) at a ow
rate of 8.8 mL/min with detection at 254 nm, to gave
compounds 1 (tR 27:4 min) and 2 (tR 32:9 min).
Further purication was performed by thin-layer chromatography (PLC silica gel 60 F254 Art. no. 13794,
Merck, Darmstadt, Germany), developed by using a
solvent system of chloroform-methanol (9:1), to aord
compounds 1 (Rf 0:66, 18.2 mg) and 2 (Rf 0:75,
17.3 mg).
Active compound 1 was obtained as a colorless
amorphous powder with the molecular formula
C15 H19 NO, based on high-resolution electrospray ionization mass spectra (HRESI-MS, pos.), m=z 252.13546
M Na (calcd. as 252.13643 for C15 H19 NONa). The
infrared spectrum of 1 showed absorption bands, max

y
To whom correspondence should be addressed. Fax: +81-3-5477-2674; E-mail: abe@nodai.ac.jp
Abbreviation: CYP, cytochrome P450 monooxygenase

Eective CYP Inhibitors Isolated from Tarragon

5

7
9
11

1'
1

N
H

2'

3'

4'

10

1
O
N

2
Fig. 1. Chemical Structures of Active Compounds 1 and 2.

(ATR) cm1 : 3422, 3294, 2936, 2855, 2223, 1621, 1590,

1440, 1257, and 997. The 1 H-nuclear magnetic resonance (NMR, CDCl3 , 400 MHz) spectra of 1 exhibited
19 protons, H : 0.93 (6H, d, J 6:6 Hz, 30 -H3 , 40 -H3 ),
1.81 (1H, m 20 -H), 2.00 (1H, s, 11-H), 2.39 (2H, br.s,
6-H2 ), 2.39 (2H, br.s, 7-H2 ), 3.16 (2H, t, J 6:6 Hz,
10 -H), 5.70 (1H, br.t, NH), 5.83 (1H, d, J 14:8 Hz,
2-H), 6.05 (1H, m, 5-H), 6.20 (1H, dd, J 10:7 and
15.1 Hz, 4-H), 7.18 (1H, dd, J 10:7 and 15.1 Hz, 3-H),
and the 13 C-NMR data (CDCl3 , 100 MHz) gave 14
peaks, C : 18.8 (C-7), 20.1 (C-30 and C-40 ), 28.6 (C-20 ),
31.3 (C-6), 40.7 (C-10 ), 65.1 (C-9 or C-10), 65.5 (C-9 or
C-10), 68.2 (C-11), 76.9 (C-8), 123.2 (C-2), 129.8 (C-4),
139.0 (C-5), 140.4 (C-3), 166.2 (C-1). The structure was
further elucidated by interpreting the heteronuclear
multiple-quantum coherence (HMQC) and heteronuclear multiple-bond coherence (HMBC). The data for
active compound 1 obtained with these ndings were
in agreement with 2E,4E-undeca-2,4-diene-8,10-diynoic
acid isobutylamide, consistent with the literature
(Fig. 1).10)
Compound 2 was identied as a colorless amorphous
powder with the molecular formula C16 H19 NO, HRESIMS (pos.), m=z 264.13722 M Na (calcd. as
262.13643 for C16 H19 NONa). The structure was elucidated by interpreting the 1D- and 2D-NMR spectra.
Based on these ndings, compound 2 was identied as
2E,4E-undeca-2,4-diene-8,10-diynoic acid piperidide
(Fig. 1).10) Compounds 1 and/or 2 had previously been
respectively isolated from Achillea and Echinacea
species.1115) Compound 1 isolated from Echinacea has
recently been reported as a CYP3A4 inhibitor.16)
The inhibitory eects of compounds 1 and 2 on
isoforms of CYP1A2, CYP2C9, CYP2D6, and CYP3A4
were investigated. The determined inhibitory eects of
compounds 1 and 2 were based on the formation of
the uorescent metabolite (3-cyano-7-hydroxycoumarin,
CHC) from the Vivid CYP Blue substrate (7-ethyloxymethyloxy-3-cyanocoumarin, EOMCC or 7-benzyloxymethyloxy-3-cyanocoumarin, BOMCC) by each
CYP isoform in black 96-well microtiter plates
(Sumitomo Bakelite, Tokyo, Japan). EOMCC was the
substrate used for testing the inhibitory activity of
CYP1A2 or CYP2D6, and BOMCC was used for testing
that of CYP3A4 or CYP2C9. The positive controls were
safrole for CYP1A2, dicoumarol for CYP2C9, cimetidine for CYP2D6, and erythromycin for CYP3A4. An
adequate concentration of compound 1 or 2, or a positive
control was dissolved in methanol. Briey, 40 mL of the

1029

obtained solution was added to 50 mL of an enzyme/

NADPH-CYP reductase/glucose 6-phosphate/glucose
6-phosphate dehydrogenase mixture in a 96-well microtiter plate, and pre-incubated for 20 min at 25  C. The
reaction was initiated by adding 10 mL of the substrate/
NADP (starter) mixture for 15 min at 25  C. The
reaction was terminated by adding 10 mL of each stop
solution (30 mM of -naphthoavone for CYP1A2,
100 mM of sulfaphenazole for CYP2C9, 10 mM of
quinidine for CYP2D6, and 30 mM of ketoconazole for
CYP3A4). The CYP1A2 and CYP2D6 activities were
assayed by monitoring the metabolism of EOMCC to
CHC. CYP2C9 and CYP3A4 activity was assayed by
monitoring the metabolism of BOMCC to CHC.
Fluorescence was monitored with a uorescence plate
reader at an excitation wavelength of 409 nm and an
emission wavelength of 460 nm. Activity was measured
as the rate of uorescent metabolite production over the
course of the reaction. The IC50 values were calculated
by linear interpolation.17) Compounds 1 and 2 inhibited
the CYP3A4 enzyme activity in vitro. Both compounds
showed a dose-dependent inhibitory eect on the
CYP3A4 activity with IC50 values of 10:0  1:3 mM
for compound 1 and 3:3  0:2 mM for compound 2
(Table 1). Compound 2 had no more inhibitory strength
than the positive control, erythromycin (IC50 2:1 
0:2 mM). We next investigated the inhibitory eects
of compounds 1 and 2 on CYP1A2, CYP2C9, and
CYP2D6. The IC50 values for these compounds toward
CYP1A2, CYP2C9, and CYP2D6 are shown in Table 1.
Their inhibitory proles towards CYP1A2 were similar
to those towards CYP3A4, revealing compound 2 to also
have the more potent inhibitory eect. Three positive
controls, cimetidine, dicoumarol, and safrole, also had
inhibitory eects on the respective CYP enzymes
(Table 1).
A metabolite of terminal acetylene formed by CYP
covalently binds to a nitrogen atom of the CYP heme
group, which is irreversibly inactivated through the
formation of a heme adduct; for example, oxidative
metabolites of arylacetylenes by CYP form irreversible
heme alkylation products with iron porphyrine.18) Compounds 1 and 2 possess terminal acetylene structures in
their diynoic moieties. These compounds may produce
irreversible heme alkylation products, because no type II
spectra, which appeared in association with the formation of a coordination linkage to the iron atom in iron
porphyrin, of the mixture of CYP3A4 and compound 1
or 2 were apparent.19) This type of inhibition is
considered as being mechanism-based. 17-Ethynylestradiol, an orally active semi-synthetic steroidal
estrogen with a terminal acetylene moiety, inactivates
CYP3A4 through its metabolites with mechanism-based
inhibition by the CYP enzyme.20) Time-dependent CYP
inhibition with NADPH-CYP reductase and a coenzyme
(NADPH) is one of the key factors involved in
mechanism-based inhibition. After pre-incubating for
20 min, using CYP3A4 and the reductase with the
coenzyme (an NADPH-CYP reductase/glucose 6-phosphate/glucose 6-phosphate dehydrogenase mixture and
NADP ), compounds 1 and 2 showed respective IC50
values of 10:0  1:3 and 3:3  0:2 mM with respect to
the reference activity. In the absence of the starter
(NADP ), the inhibitory activities of all the tested

1030

Z. B RAHMI et al.
Table 1. Inhibitory Eects of Compounds 1 and 2, 17-Ethynylestradiol and the Positive Controls on CYP Activities
IC50 (mM)
Compound

CYP

1
2
17-Ethynylestradiol
Safrolea
Dicoumarolb
Cimetidinec
Erythromycind

1A2

2C9

2D6

3A4

12:9  0:3
2:5  2:2
125:7  32:2
3:9  0:7

>1000
>1000
42:9  21:5

3:8  0:2

266:0  101:6
470:7  97:4
45:3  14:9

25:2  3:0

10:0  1:3
3:3  0:2
0:7  0:2

2:1  0:2

Mean  SD (n 3), a Positive control for CYP1A2, b Positive control for CYP2C9, c Positive control for CYP2D6, d Positive control for CYP3A4.

NADP+ (+)

NADP+ (-)
100

100

90

90

80

80

Inhibition ratio (%)

Inhibition ratio (%)

70
60
50
40
30

70
60
50
40
30

20

20

10

10

0
0

10

20

30

10

Pre-incubation time (min)

20

30

Pre-incubation time (min)

Fig. 2. Time-Dependent Inhibitory Eects of the Tested Compounds on CYP3A4 Activity.

Each value is presented as the mean  standard deviation of triplicate determinations. Compound 1 ( , 10.0 mM), compound 2 ( , 3.3 mM),
17-ethynylestradiol ( , 0.7 mM), erythromycin ( , 2.1 mM) and ketoconazole ( , 0.002 mM). The time-dependent inhibitory eects of
compounds 1 and 2 were determined by using black 96-well microtiter plates, based on the formation of uorescent metabolites by the CYP
enzyme. Briey, 90 mL of an incubation mixture containing compound 1, 2 or a positive control at a concentration around the IC50 value of each
were pre-incubated at room temperature in the absence (A) or the presence (B) of the starter (NADP ) for various times (0, 10, 20, and 30 min).

compounds, including compounds 1 and 2 with preincubation, were steady at around their IC50 values
(Fig. 2A). In the presence of the starter, however, the
inhibitory eect of compound 1 increased from
19:6  3:9% (without pre-incubation) to 63:5  3:7%
(with pre-incubation for 30 min) at 10.0 mM, whereas that
of compound 2 increased from 29:8  3:6% (without
pre-incubation) to 83:9  3:5% (with pre-incubation for
30 min) at 3.3 mM. The known mechanism-based inhibitors, erythromycin and 17-ethynylestradiol, used as
the positive controls showed similar behavior; in
contrast, a known competitive inhibitor (ketoconazole)
showed no time-dependent inhibition (Fig. 2B). The
dose and pre-incubation time-dependent manner of
compounds 1 and 2 suggest that they were potent
mechanism-based inhibitors.
Food-drug interactions can be critically important,
and interactions between the bioactive components in
herbs and CYPs are generally recognized. The inhibition
of CYP3A4, CYP1A2, CYP2C9, and CYP2D6 by
bioactive components in foods containing herbs changes
the bioavailability of clinically used drugs. CYP3A4 is
the most abundant drug-metabolizing enzyme in the
human liver and gastrointestinal tract, and the major
enzyme responsible for the metabolism of pharmacologic agents.21) The inhibitory activator against CYP3A4
with mechanism-based inhibition causes irreversible
deactivation of the enzyme and can completely inactivate the metabolism of several pharmacologic agents.

The consequence is that patients taking medicines must

be aware of possible interactions between the drug being
taken and the foods they are consuming, including foods
containing CYP enzyme inhibitors.22)
The ndings of the present study prompt us to report
for the rst time that compounds 1 and 2 exhibited
strong CYP inhibitory eects in vitro. These ndings
suggest that routine heavy consumption of tarragon has
potential negative eects in numerous food-drug interactions. According to the CYP isozymes that can
selectively inhibit depending on the dierence in
structure of alkylamide,23) tarragon could also become
a signicant food material for preventing cancer through
its inhibition of the metabolic activation of carcinogens.

Acknowledgments
We appreciate the helpful suggestions and advice
contributed by Dr. H. Koshino (RIKEN). We thank Dr.
Y. Q. Ye for measuring the high-resolution mass
spectra. This work was performed as a part of the
Advanced Research Project of Tokyo University of
Agriculture.

References
1)
2)

Obolskiy D, Pischel I, Feistel B, Glotov N, and Heinrich M,

Agric. Food Chem., 59, 1136711384 (2011).
Logendra S, Ribnicky DM, Yang H, Poulev A, Ma J, Kennelly
EJ, and Raskin I, Phytochemistry, 67, 15391546 (2006).

Eective CYP Inhibitors Isolated from Tarragon

3)
4)

5)

6)
7)
8)
9)
10)
11)
12)
13)

Wang ZQ, Ribnicky D, Zhang XH, Raskin I, Yu Y, and Cefalu

WT, Metab. Clin. Exp., 57, S58S64 (2008).
Kobayashi S, Watanabe J, Fukushi E, Kawabata J, Nakajima M,
and Watanabe M, Biosci. Biotechnol. Biochem., 67, 12501257
(2003).
Ortiz PR, Cytochrome P450: Structure, Mechanism, and
Biochemistry second ed., Plenum, New York, pp. 473551
(1995).
Bailey DG, Arnold JM, and Spence JD, Clin. Pharmacokinet.,
26, 9198 (1994).
Bailey DG, Malcolm J, Arnold O, and Spence JD, Br. J. Clin.
Pharmacol., 46, 101110 (1998).
Bailey DG, Spence JD, Munoz C, and Arnold JMO, Lancet,
337, 268269 (1991).
Bruno RD and Njar VCO, Bioorg. Med. Chem., 15, 50475060
(2007).
Hofer O, Greger H, Robien W, and Wemer A, Tetrahedron, 42,
27072716 (1986).
Yang M, Li J-X, Li X, and Jin Z-J, Pharmazie, 60, 554558
(2005).
Greger H and Werner A, Planta Med., 90, 482486 (1990).
Kuropka G and Glombtzs KW, Planta Med., 53, 440442
(1987).

14)

15)
16)
17)
18)
19)
20)
21)

22)

23)

1031

Hormann J, Redel D, Forgo P, Szabo P, Freund TF, Haller J,

Bojnik E, and Benyhe S, Phytochemistry, 72, 18481853
(2011).
Mulla-Jatec B, Breu W, Probstle A, Redl K, Greger H, and
Bauer R, Planta Med., 60, 3740 (1994).
Modarai M, Yang M, Suter A, Kortenkamp A, and Heinrich M,
Planta Med., 76, 378385 (2010).
Huang Y-T, Onose J, Abe N, and Yoshikawa K, Biosci.
Biotechnol. Biochem., 73, 855860 (2009).
Montellano PRO and Komives EA, J. Biol. Chem., 260, 3330
3336 (1985).
Dawson JH, Andersson LA, and Sono M, J. Biol. Chem., 257,
36063617 (1982).
Lin HL, Kent UM, and Hollenberg PF, J. Pharmacol. Exp.
Ther., 301, 160167 (2002).
Guengerich FP, Cytochrome P450: Structure, Mechanism and
Biochemistry Third ed., ed. de Montellano PRO, Plenum, New
York, pp. 377530 (2005).
Correia MA and de Montellano PRO, Cytochrome P450:
Structure, Mechanism and Biochemistry Third ed., ed. de
Montellano PRO, Plenum, New York, pp. 247322 (2005).
Modarai M, Gertsch J, Suter A, Heinrich M, and Kortenkamp A,
J. Pharm. Pharmacol., 59, 567573 (2007).