Alcohol absorption inhibitors from bay leaf (Laurus nobilis): Structure-requirements of sesquiterpenes for the activity

Article abstract of DOI:10.1016/S0968-0896(00)00127-9

Through a bioassay-guided separation using inhibitory activity on blood ethanol elevation in oral ethanol-loaded rat, various sesquiterpenes having an α-methylene-γ-butyrolactone moiety, costunolide (1), dehydrocostus lactone (2), zaluzanin D (3), reynosin (4), santamarine (5), 3α-acetoxyeudesma-1,4(15),11(13)-trien-12,6α-olide (6) and 3-oxoeudesma-1,4,11(13)-trien-12,6α-olide (7), were isolated as the active principle from the leaves of Laurus nobilis (bay leaf, laurel). In order to characterize the structure requirement for the activity, several reduction products (2a-2d) and amino acid adducts (2e, 2f) of the α-methylene-γ-butyrolactone moiety were synthesized from 2 and the inhibitory activities of these sesquiterpenes, together with α-methylene-γ-butyrolactone (12) and its related compounds (13-16), were examined. These results indicated that the γ-butyrolactone or γ-butyrolactol moiety having α-methylene or α-methyl group was essential for the inhibitory activity on ethanol absorption. Since 1, 2 and 12 showed no significant effect on glucose absorption, these sesquiterpenes appeared to selectively inhibit ethanol absorption. In addition, the acute toxicities of 1 and 2 in a single oral administration were found to be lower than that of 12. Copyright (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0968-0896(00)00127-9

Bioorganic & Medicinal Chemistry 8 (2000) 2071±2077
Alcohol Absorption Inhibitors from Bay Leaf (Laurus nobilis):
Structure-Requirements of Sesquiterpenes for the Activity
Masayuki Yoshikawa,a,* Hiroshi Shimoda,a Toshiaki Uemura,a Toshio Morikawa,a
Yuzo Kawaharab and Hisashi Matsudaa
aKyoto Pharmaceutical University, Misasagi, Yamashina-ku, Kyoto 607-8412, Japan
bMorishita Jinan Co. Ltd, 1-1-30 Tamatsukuri, Chuou-ku, Osaka 540-8566, Japan
Received 7 March 2000; accepted 12 May 2000
AbstractÐThrough a bioassay-guided separation using inhibitory activity on blood ethanol elevation in oral ethanol-loaded rat,
various sesquiterpenes having an a-methylene-g-butyrolactone moiety, costunolide (1), dehydrocostus lactone (2), zaluzanin D (3),
reynosin (4), santamarine (5), 3a-acetoxyeudesma-1,4(15),11(13)-trien-12,6a-olide (6) and 3-oxoeudesma-1,4,11(13)-trien-12,6a-olide
(7), were isolated as the active principle from the leaves of Laurus nobilis (bay leaf, laurel). In order to characterize the structure
requirement for the activity, several reduction products (2a±2d) and amino acid adducts (2e, 2f) of the a-methylene-g-butyrolactone
moiety were synthesized from 2 and the inhibitory activities of these sesquiterpenes, together with a-methylene-g-butyrolactone (12)
and its related compounds (13±16), were examined. These results indicated that the g-butyrolactone or g-butyrolactol moiety having
a-methylene or a-methyl group was essential for the inhibitory activity on ethanol absorption. Since 1, 2 and 12 showed no signi®cant
e€ect on glucose absorption, these sesquiterpenes appeared to selectively inhibit ethanol absorption. In addition, the acute toxicities
of 1 and 2 in a single oral administration were found to be lower than that of 12. # 2000 Elsevier Science Ltd. All rights reserved.
Introduction
appetite for drinking.2 However, it is known that ben-
zodiazepines lead to abuse, psychological dependence,
and adverse behavioral e€ects. These drugs can produce
physiological dependence when taken chronically, and
result in dose escalation.3 Aldehyde dehydrogenase
inhibitors force alcoholics to quit drinking based on the
fear of unpleasant reaction elicited after ethanol intake,
but these medicines, especially disul®ram, are reported to
show serious toxic symptoms such as fulminant hepatitis,
optic neuritis, peripheral neuropathy, encephalopathy,
or catatonia. Therefore, more safe medications and
treatments are in demand to prevent alcoholic disorders.
Taking large quantities of alcohol in one time is known
to induce an acute alcohol toxicity with acidosis, heart
failure and respiratory depression caused by an auto-
nomic nerve and cerebrum dysfunction. Long-term
consumption of alcohol in large quantities induces a
number of disorders, e.g. hepatopathy, gastrointestinal
disorder, chronic pancreatitis, peripheral nerve disorder,
myocardosis, hypertension and hematopoiesis disorder.
Long-term alcohol consumption not only causes organ
dysfunctions but also can cause phrenopathy and have
threatening social e€ects. Violent crimes and trac acci-
dents caused by alcohol intoxication have increased over
the past decade in developed countries. It is also repor-
ted that there is a correlation between alcohol con-
sumption, the number of alcoholics and the occurrence
of life-threatening assaults.
Contrary to synthetic medicine treatments, many Chinese
traditional medicines have been known to have pre-
ventive and therapeutic e€ects for ethanol intoxication,
but their active components have not been identi®ed so
far. Recently, we isolated a number of saponins with
inhibitory e€ect on ethanol absorption from natural
medicines. Some examples include elatosides from the
bark of Aralia elata, sapindosides from the pericarps of
Sapindus mukurossi, escins from the seeds of Aesculus
hippocastanum, camelliasaponins from the seeds of
Camellia japonica, senegins and senegasaponins from the
root of Polygala senega var. latifolia, and momordins
from the fruit of Kochia scoparia.4 9 Furthermore, we
Benzodiazepines and vitamins are medicated for alcoholic
patients in the elimination phase of the cure process to
suppress withdrawal.1 Cyanamide and disul®ram, aldehyde
dehydrogenase inhibitors, are also used to decrease the
*Corresponding author. Tel.: +81-75-595-4633; fax: +81-75-595-
4768; e-mail: shoyaku@mb.kyoto-phu.ac.jp
0968-0896/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(00)00127-9

2072
M. Yoshikawa et al. / Bioorg. Med. Chem. 8 (2000) 2071±2077
have characterized dihydro¯avonols named hovenitins
that have an inhibitory e€ect on alcohol-induced muscular
relaxation and hepatoprotective activity from the fruit
of Hovenia dulcis.10
In the course of our studies on anti-alcohol substances
from natural resources, we found that the methanolic
extract of the leaves of Lauraceae plant, Laurus nobilis
(bay leaf, laurel) especially showed potent inhibitory
activity on ethanol absorption in rats. This plant is
widely cultivated in Mediterranean countries and the
leaves are used as a herb in cooking for soup or stew as
well as herbal medicine for excitant, stomachic, and
digestant purposes in Europe. The pharmacological
properties of L. nobilis include inhibitory e€ects on NO
production (leaves), anti-ulcer (seeds) and anti-diabetic
(leaves) e€ects, and the enhancement of liver glutathione
S-transferase (GST) activity.11 14 In this report, we
describe the isolation of seven sesquiterpenes (1±7) from
the leaves of L. nobilis, its inhibitory e€ect on ethanol
absorption, and the elucidation of the structure require-
ments for its activity.15 In addition, we describe the acute
toxicity level of the principal inhibitors, costunolide (1)
and dehydrocostus lactone (2), comparing to the active
compound, a-methylene-g-butyrolactone (12).
Chart 1. Sesquiterpenes (1±10) isolated from the leaves of Laurus nobilis.
fraction and isolated four types of sesquiterpenes, that is,
germacrane-type (costunolide16 (1, 0.18%)), guiane-type
(dehydrocostus lactone17 (2, 0.008%) and zaluzanin D18
(3, 0.006%)), eudesmane-type (reynosin19 (4, 0.018%),
santamarine20 (5, 0.041%), 3a-acetoxyeudesma-1,4(15),11
(13)-trien-12,6a-olide21 (6, 0.004%), 3-oxoeudesma-1,4,
11(13)-trien-12,6a-olide22 (7, 0.009%) and b-eudesmol23
(8, 0.0025%)), and caryophyllene-type (b-caryophyllene24
(9, 0.0008%) and caryophyllene oxide25 (10, 0.0008%))
(Chart 1). Fig. 1 shows the e€ects of the sesquiterpenes
on blood ethanol elevation. Compounds 1±7 having an
a-methylene-g-butyrolactone moiety inhibited blood
ethanol elevation at 50 mg/kg. In particular, 1, 6 and 7
showed potent suppression at 25 mg/kg. On the other
hand, 8±10 lacking the lactone moiety had no e€ect at
50 mg/kg. These results suggest that an a-methylene-g-
butyrolactone moiety is needed to inhibit blood ethanol
elevation.
Results and Discussion
E€ects of extract and sesquiterpenes isolated from
L. nobilis on blood ethanol elevation in ethanol-loaded rats
Table 1 shows the e€ects of the extract and fractions
obtained from the leaves of L. nobilis on blood ethanol
elevation in ethanol-loaded (20%, 5 mL/kg) rats. Metha-
nolic extract (yield: 20.0% from dried leaves) inhibited the
elevation of blood ethanol level from 125 to 500 mg/kg
in a dose-dependent manner. The ethyl acetate soluble
fraction (7.3%) also strongly inhibited blood ethanol
elevation, but no suppressive e€ect was observed in the
water soluble fraction (12.7%). Then, we performed
bioassay-guided separation of the ethyl acetate soluble
Synthesis of reduction products (2a±2d) and amino acid
adducts (2e, 2f) from dehydrocostus lactone (2)
Table 1. E€ects of methanolic extract and its fractions from the leaves
of Laurus nobilis on blood ethanol elevation in ethanol-loaded ratsa
To study the structure±activity relationships in the a-
methylene-g-butyrolactone moiety of 2, we prepared four
reductants (2a±2d) and two amino acid adducts (2e,
2f)26,27 linked with l-proline and l-cysteine at the 13-
position in 2 (Chart 2). The reductants (2a±2d) were pre-
pared by the following procedure: namely, the treatment
of 2 with lithium aluminum hydride (LiAlH4) gave a
11a-methyl-6,12-diol derivative28 (2a, 46%), and treat-
ment with sodium borohydride (NaBH4) yielded mokko
lactone28 (2b, 18%), the 11a-methyl lactol mixture28 (2c,
70% ca. 1:1 ratio), and the 11a-methyl-12-methylene deri-
vative (2d, 9%). These known constituents (2a, 2b and
2c) were identi®ed by comparison of their physical data
with reported values.28
Blood ethanol (mg/mL)
Treatment
Dose
(mg/kg)
N
0.5 h
1 h
2 h
Control
Methanolic
extract
Ð
5
5
5
5
0.64Æ0.04
0.59Æ0.03
0.28Æ0.04
125
250
500
0.15Æ0.04**b 0.34Æ0.10** 0.32Æ0.03
0.05Æ0.02** 0.05Æ0.02** 0.11Æ0.03**
0.01Æ0.01** 0.01Æ0.01** 0.03Æ0.02**
Control
AcOEt-soluble
fraction
Water-soluble
fraction
Ð
5
5
5
5
5
0.88Æ0.03
0.64Æ0.02
0.32Æ0.02
125
250
125
250
0.12Æ0.02** 0.24Æ0.06** 0.10Æ0.05**
0.05Æ0.01** 0.08Æ0.01** 0.11Æ0.04**
0.90Æ0.04
0.87Æ0.03
0.65Æ0.03
0.64Æ0.04
0.34Æ0.02
0.36Æ0.04
aEach compound was given orally to fasted (20±22 h) rats and ethanol
(20%, 5 mL/kg) was loaded 30 min thereafter. Blood samples were
collected from the infraorbital venous plexus at 0.5, 1 and 2 h after
administration of ethanol. Values are meansÆS.E.M.
The 11a-methyl-12-methylene derivative (2d) showed a
molecular ion (M+) peak at m/z 218 in the electron
bAsterisks denote signi®cant di€erences from control, **: P<0.01.

M. Yoshikawa et al. / Bioorg. Med. Chem. 8 (2000) 2071±2077
2073
Figure 1. E€ects of the sesquiterpenes from the leaves of L. nobilis on blood ethanol elevation in ethanol-loaded rats. Each compound was given
orally to fasted (20±22 h) rats, and ethanol (20%, 5 mL/kg) was loaded 30 min thereafter. Blood samples were collected from the infraorbital venous
plexus at 0.5, 1 and 2 h after administration of ethanol. Symbols are represented by the following , control: *, 25 mg/kg: *, and 50 mg/kg: ~.
Values are means with S.E.M. (N=3±5), and asterisks denote signi®cant di€erences from control, *: P<0.05, **: P<0.01.
Chart 2. Reduction products (2a±2d) and amino acid adducts (2e, 2f) from dehydrocostus lactone (2).
impact (EI)-MS measurement, and its molecular formula
C15H22O was determined by high-resolution MS mea-
surement. The IR spectrum of 2d showed no absorption
bands due to the g-lactone function, which was observed
in that of 2. The 1H NMR (CDCl3) and 13C NMR (Table
2) spectra of 2d were closely similar to those of 2b,
except for the signals due to the 12-methylene (d 3.48 (1H,
dd, J=6.1, 11.0 Hz, 12a-H), 3.66 (1H, dd, J=5.2, 11.0Hz,
12b-H)) function. Furthermore, the stereostructure of the
11-position of 2d was determined by the nuclear Over-
hauser e€ect spectroscopy (NOESY) experiment, in which
the NOE correlation was observed between the 6-proton
(d 3.23 (1H, dd, J=8.5, 8.6 Hz)) and the 11-proton (d 2.09
(1H, dddq, J=5.2, 6.1, 10.3, 7.0 Hz)). Thus, the stereo-
structure of 2d was characterized to be the hydrofuran
structure having a 11a-methyl group.
Amino acid adducts (2e, 2f) were prepared in a similar
manner as described in previous papers.26,27 Saussur-
eamine B (2e) and 2f were obtained by treatment of 2 with
l-proline or l-cysteine in the presence of triethylamine in
ethanol or 70% aqueous ethanol.
E€ects of synthetic derivatives from 2 on blood ethanol
elevation in ethanol-loaded rats
Table 3 shows the inhibitory activity of the derivatives
(2a±2f) from 2 on ethanol absorption. The diol derivative
(2a) completely lacked the activity which shows that the
lactone ring is necessory for the inhibitory activity.
Mokko lactone (2b) (a reductant of the 11-exo-methylene
group) and 2c (a reductant of the 12-carbonyl group)
maintained the activity. On the other hand, compound

2074
M. Yoshikawa et al. / Bioorg. Med. Chem. 8 (2000) 2071±2077
2d lacking the 12-oxygen function, and amino acid
adducts (2e, 2f) lacked the activity. These results indi-
cate that a lactone or a lactol ring is essential for the
activity. It is not yet clari®ed that whether the exo-
methylene or the methyl group participate in the
expression of the activity, because both 2 and 2b showed
the activity. However, since amino acid adducts (2e, 2f)
lacked the activity, a large hydrophilic group such as
amino acid groups conjugated to the methyl group at
the 13-position seems to reduce the activity.
Chart 3. Chemical structures of g-butyrolactones and a related com-
pound.
E€ects of the ꢀ-lactone derivatives (12±15) and tetrahydro-
furan (16) on blood ethanol elevation in ethanol-loaded rats
Table 4 shows the activities of the g-lactone derivatives
(12±15) and tetrahydrofuran (16) which are suggested to
be the active site of the sesquiterpenes (1±7) from L.
nobilis. a-Methylene-g-butyrolactone (12) dramatically
suppressed blood ethanol elevation at doses of 12.5 to
50 mg/kg. 3-Methyl-2(5H)-furanone (13) also suppressed
blood ethanol elevation at doses of 25 and 50 mg/kg.
These results correlate with results of 2 and 2b. g-
Butyrolactone (15) and tetrahydrofuran (16) had no
activity. In addition, g-methylene-g-butyrolactone (14),
an isomer of 12, lacked the activity at 50 mg/kg. These
results indicate that the lactone ring with an a-methyl-
ene or a-methyl group in the ring is essential for the
activity (Chart 3).
Table 2. 13C NMR data for 2c and 2da
2c
2d
12a-OH
12b-OH
C-1
C-2
C-3
C-4
C-5
C-6
C-7
C-8
47.6
30.0
47.6
29.9
32.1*
153.2
52.7
83.7
53.3
32.2
37.6
151.1**
49.2
104.0
15.8
110.8
107.7***
46.2
29.0
31.2
154.2
54.1
68.4
47.1
26.9
34.3
151.6
36.3
66.5
13.5
109.9
110.8
32.4*b
153.8
54.6
86.7
47.8
31.6
38.0
C-9
C-10
C-11
C-12
C-13
C-14
C-15
151.4**
45.3
99.3
11.4
110.8
108.1***
E€ects of the costunolide (1), dehydrocostus lactone (2),
and ꢁ-methylene-ꢀ-butyrolactone (12) on serum glucose
elevation in glucose-loaded rats
aMeasured in CDCl3 (d) at 125 MHz.
b*, **, ***Assignments may be interchangeable in same row.
In order to study the selectivity of the active sesqui-
terpenes on blood ethanol elevation, we investigated the
e€ects of 1, 2 and 12 on glucose absorption in rats. Figure
Table 3. E€ects of derivatives (2a±2f) of dehydrocostus lactone (2)
on blood ethanol elevation in ethanol-loaded ratsa
Table 4. E€ects of g-butyrolactone derivatives (12±16) on blood
ethanol elevation in ethanol-loaded ratsa
Blood ethanol (mg/mL)
Blood ethanol (mg/mL)
Dose
(mg/kg)
N
0.5 h
1 h
2 h
Dose
(mg/kg)
N
0.5 h
1 h
2 h
Control
2
Ð
25
50
5
5
5
0.74Æ0.02
0.55Æ0.03
0.54Æ0.01
0.40Æ0.05
0.21Æ0.01
0.17Æ0.02
Control
a-Methylene-g-
butyrolactone (12)
Ð
12.5
25
5
5
5
5
0.68Æ0.02
0.61Æ0.05 0.14Æ0.04
0.32Æ0.04**b 0.71Æ0.05 0.38Æ0.01
0.07Æ0.04** 0.18Æ0.10** 0.03Æ0.03
0.04Æ0.02** 0.12Æ0.07** 0.04Æ0.03
0.20Æ0.09**b 0.29Æ0.09* 0.17Æ0.02
50
Control
2a
2c
Ð
50
25
50
5
5
5
5
0.78Æ0.04
0.80Æ0.04
0.68Æ0.07
0.75Æ0.07
0.66Æ0.06
0.60Æ0.06
0.25Æ0.02
0.28Æ0.02
0.17Æ0.05
Control
3-Methyl-2(5H)-
furanone (13)
Ð
25
50
5
4
4
0.82Æ0.04
0.64Æ0.04 0.37Æ0.04
0.26Æ0.07** 0.28Æ0.10* 0.30Æ0.07
0.24Æ0.04** 0.31Æ0.10** 0.31Æ0.07
0.10Æ0.04** 0.34Æ0.15 0.16Æ0.11
Control
Mokko
lactone (2b)
Ð
25
50
5
5
5
0.67Æ0.04
0.65Æ0.07
0.22Æ0.02
0.17Æ0.05
0.21Æ0.08
Control
g-Methylene-g-
butyrolactone (14)
Ð
25
50
4
4
4
0.54Æ0.07
0.59Æ0.06
0.67Æ0.02
0.51Æ0.05 0.18Æ0.04
0.60Æ0.06 0.17Æ0.05
0.53Æ0.01 0.12Æ0.07
0.14Æ0.04** 0.33Æ0.09
0.04Æ0.01** 0.41Æ0.15
Control
2d
Ð
50
4
4
0.61Æ0.06
0.45Æ0.04
0.17Æ0.04
Control
g-Butyrolactone (15)
Ð
25
50
5
5
5
0.42Æ0.03
0.42Æ0.05
0.51Æ0.04
0.80Æ0.05
0.82Æ0.03
0.74Æ0.07
0.42Æ0.03 0.17Æ0.03
0.34Æ0.06 0.16Æ0.03
0.35Æ0.03 0.12Æ0.04
0.72Æ0.03 0.48Æ0.15
0.73Æ0.03 0.33Æ0.02
0.67Æ0.02 0.34Æ0.01
0.47Æ0.09
0.40Æ0.09
0.18Æ0.03
Control
Saussureamine
B (2e)
Ð
50
5
4
0.82Æ0.04
0.76Æ0.07
0.64Æ0.04
0.59Æ0.03
0.37Æ0.04
0.36Æ0.04
Control Ð
Tetrahydrofuran (16) 25
50
6
5
4
2f
50
4
0.85Æ0.02
0.60Æ0.03
0.32Æ0.10
aValues are meansÆS.E.M.
aValues are meansÆS.E.M.
bAsterisks denote signi®cant di€erences from control, *: P<0.05, **:
P<0.01.
bAsterisks denote signi®cant di€erences from control, *: P<0.05, **:
P<0.01.

M. Yoshikawa et al. / Bioorg. Med. Chem. 8 (2000) 2071±2077
2075
Figure 2. E€ects of costunolide (1), dehydrocostus lactone (2), and a-methylene-g-butyrolactone (12) on increase of serum glucose level in glucose-
loaded rats. Each compound was given orally to fasted (20±22 h) rats and glucose (1 g/kg) was loaded 30 min thereafter. Blood samples were col-
lected from the infraorbital venous plexus at 0.5, 1, and 2 h after administration of glucose. Symbols are represented by the following, control: *,
25 mg/kg: *, 50 mg/kg: ~, and glucose untreated group ~. Values are means with S.E.M. (N=5), and asterisks denote signi®cant di€erences from
control *: P<0.05, **: P<0.01.
2 shows the e€ects of the 1, 2 and 12 on serum glucose
elevation in glucose-loaded (1 g/kg) rats. Costunolide
(1) and dehydrocostus lactone (2) have no in¯uence on
serum glucose level at 50 mg/kg. Compound 12 slightly
suppressed serum glucose elevation at 50 mg/kg, but the
e€ect was not signi®cant. These results suggest that ses-
quiterpenes from L. nobilis selectively inhibit the alcohol
absorption rather than the glucose absorption.
Table 5. Single dose oral toxicity of costunolide (1), dehydrocostus
lactone (2), and a-methylene-g-butyrolactone (12) in micea
Mortality
Dose (mg/kg)
N
1
3
6
24 (h)
7 (day)
Total
1
250
500
250
500
50
125
250
500
6
6
6
6
6
6
6
6
0
0
0
0
0
0
1
6
0
0
0
0
0
0
2
Ð
0
0
0
0
0
0
0
Ð
0
0
0
0
0
0
3
Ð
0
0
0
0
0
0
Ð
Ð
0
0
0
0
0
0
6
6
2
12
Single dose oral toxicity of costunolide (1), dehydrocostus
lactone (2) and ꢁ-methylene-ꢀ-butyrolactone (12) in mice
Table 5 shows the result of single dose oral toxicities of
1, 2 and 12 in mice. No death was observed in the
groups given 1 or 2 at doses of 250 and 500 mg/kg. On
the other hand, all mice given 12 were dead within 24 h
(250 mg/kg) and 1 h (500 mg/kg). The mice in the group
given 12 at 500 mg/kg died of breathing cessation and
generalized convulsion. g-Butyrolactone (15) is reported
to induce an increase in l-dihydroxyphenylalanine
(DOPA) synthesis and its accumulation in the brain.29,30
Its metabolite, g-hydroxybutyric acid, easily passes
through the brain barrier and causes DOPA accumula-
tion in the striatum.31 DOPA accumulation leads to a
decrease in dopamine synthesis and causes reduction of
dopamine neuronal impulse ¯ow.32,33 Since the chemical
structure of 12 is similar to 15, 12 may act on the brain
and show toxicity by the damage of neurotransmission
through dopamine.
aEach compound was given orally to fasted (22 h) mice. Mice were con-
tinuously fasted for 24h and housed in standard conditions for 7 days.
Experimental
The following instruments were used to measure physical
data of dehydrocostus lactone (2) derivatives: speci®c
rotations, Horiba SEPA-300 digital polarimeter (l=5cm);
IR spectra, Shimadzu FTIR-8100 spectrometer; EI-MS
and high-resolution MS, JEOL JMS-GCMATE mass
spectrometer; FAB MS and high-resolution MS, JEOL
JMS-SX 102A mass spectrometer; 1H NMR spectra,
JNM-LA500 (500MHz) spectrometer; 13C NMR spectra,
JNM-LA500 (125MHz) spectrometer with tetramethyl-
silane as an internal standard.
In conclusion, we isolated seven active sesquiterpenes
(1±7) from the leaves of L. nobilis as alcohol absorption
inhibitors. The active moiety in these sesquiterpenes was
found to be the a-methylene-g-butyrolactone moiety.
Furthermore, a lactone or a lactol ring with an a-methyl
or a-methylene group was essential to show suppression of
ethanol absorption. Alternative medical treatments such as
these can be used to regulate alcohol ingestion or absorp-
tion and reduce the risk of alcoholic disorders. These
sesquiterpenes not only have a potent preventive e€ect
for acute alcohol toxicity but also can be available to
help patients recover from addiction to alcohol abuse.
The following experimental conditions were used for
chromatography: ordinary-phase silica gel column chro-
matography, Silica-gel BW-200 (Fuji Silysia Chemical,
Ltd, 150±350 mesh); reversed-phase silica gel column
chromatography, Chromatorex ODS DM1020T (Fuji
Silysia Chemical, Ltd, 100±200 mesh); TLC, pre-coated
TLC plates with Silica gel 60F254 (Merck, 0.25 mm)
(ordinary phase) and Silica gel RP-18 60F254 (Merck,
0.25 mm) (reversed-phase); reversed-phase HPTLC, pre-
coated TLC plates with Silica gel RP-18 60WF254S
(Merck, 0.25 mm); detection was achieved by spraying
with 1% Ce(SO4)2±10% aqueous H2SO4 and heating.

2076
M. Yoshikawa et al. / Bioorg. Med. Chem. 8 (2000) 2071±2077
Animals and materials
Chemical transformation of dehydrocostus lactone (2)
Male Wistar rats aged 4±6 weeks and ddY mice aged 5
weeks were purchased from Kiwa Laboratory Animals
Co., Ltd (Wakayama, Japan), and housed in an air-con-
ditioned room at 23Æ2 ꢀC. Standard laboratory chow
(MF, Oriental Yeast Co., Ltd, Tokyo, Japan) and tap
water were given ad libitum. They were fasted for 20±
22 h prior to experiments, but were supplied with water
ad libitum. Test samples were suspended in 5% acacia
solution and given orally at 5 mL/kg in each experiment.
Reduction of 2 with LiAlH4. A solution of 2 (150 mg,
0.65 mmol) in tetrahydrofuran (THF, 5.0 mL) was trea-
ted with LiAlH4 (62 mg) and the mixture was stirred at
0 ꢀC for 30 min. The reaction mixture was then quenched
with aq sat. ether and subsequently 4N aq KOH, and the
whole was extracted with AcOEt. The AcOEt extract was
washed with brine, then dried over MgSO4 powder and
®ltered. Removal of the solvent from the ®ltrate under
reduced pressure furnished a residue, which was puri®ed
by HPLC (YMC-Pack ODS-A 250Â20mm i.d. (MeOH:
H2O 70:30 v/v)) to give 2a27 (66 mg, 46%).
g-Butyrolactones (12, 14, 15) were purchased from
Sigma-Aldrich Chemie Gmbh (Steinheim, Germany). 3-
Methyl-2(5H)-furanone (13) and tetrahydrofuran (16)
were obtained from Tokyo Kasei Kogyo Co. Ltd
(Tokyo, Japan).
Reduction of 2 with NaBH4. A solution of 2 (600 mg,
2.6 mmol) in MeOH (10.0 mL) was treated with sodium
borohydride (NaBH4, 100 mg) and the mixture was stir-
red at 0 ꢀC for 1 h. The reaction mixture was quenched at
acetone, and extracted with AcOEt. The AcOEt extract
was washed with brine, then dried over MgSO4 powder
and ®ltered. Removal of the solvent from the ®ltrate
under reduced pressure furnished a residue, which was
puri®ed by HPLC (YMC-Pack ODS-A 250Â20 mm i.d.
(MeOH:H2O 75:25 v/v)) to give mokko lactone28 (2b,
106 mg, 18%), two lactol mixture28 (2c, 489 mg, 70%),
and 2d (53 mg, 9%).
Isolation of chemical constituents from the leaves of
L. nobilis
The dried leaves of L. nobilis (5.0kg, harvested in Turkey)
were extracted with methanol at room temperature and
evaporated in vacuo under 40ꢀC to obtain the methanolic
extract (999g, 20.0% from the leaves). The methanolic
extract (366g) was partitioned into an ethyl acetate±water
mixture to obtain the ethyl acetate-soluble fraction (134 g,
7.3%) and the water-soluble fraction (232 g, 12.7%).
The ethyl acetate-soluble fraction (130 g) was subjected
to ordinary-phase silica-gel chromatography (2.6 kg,
hexane:AcOEt (30:1)!AcOEt!MeOH) to give eleven
fractions (Fr.1 (747mg), Fr.2 (7.55 g), Fr.3 (7.33g), Fr.4
(23.3 g), Fr.5 (3.48g), Fr.6 (13.0 g), Fr.7 (3.76 g), Fr.8
(10.5 g), Fr.9 (11.7 g), Fr.10 (10.2 g), Fr.11 (35.2 g)). Frac-
tion 1 (747mg) was separated by ordinary-phase silica-gel
[35 g, hexane:AcOEt (50:1)!(3:1)) to give b-caryophyllene
(9, 14mg, 0.0008%) and caryophyllene oxide (10, 14mg,
0.0008%). A part of each fraction 4±8 was subjected to
reversed-phase silica-gel (60% aq MeOH!100% MeOH)
column chromatography and ®nally HPLC (Cosmosil
5C18, 250Â20mm i.d., 50%!80% aq MeOH) to give
costunolide (1, 767 mg, 0.18%), dehydrocostus lactone (2,
34mg, 0.008%), zaluzanin D (3, 96mg, 0.006%), reynosin
(4, 299 mg, 0.018%), santamarine (5, 447 mg, 0.041%),
3a-acetoxyeudesma-1,4(15),11(13)-trien-12,6a-olide (6,
70 mg, 0.004%), 3-oxoeudesma-1,4,11(13)-trien-12,6a-
olide (7, 155 mg, 0.009%), and b-eudesmol (8, 42 mg,
0.0025%). These sesquiterpenes (1±10) were identi®ed by
comparison of their physical data with those of authentic
samples (1, 2, 5) or with reported values.18,19,21 25
2c. 1H NMR (CDCl3) d: (the signals due to the 12a-OH
derivative34) 1.03 (3H, d, J=6.7 Hz, 13-H3), 1.10 (1H,
m, 8b-H), 1.73 (1H, m, 11-H), 1.82, 1.92 (1H each, both
m, 2-H2), 1.90 (1H, m, 7-H), 2.03 (1H, m, 8a-H), 2.06,
2.45 (1H each, both m, 9-H2), 2.48, 2.50 (1H each, both
m, 3-H2), 2.74 (1H, dd, J=8.8, 9.2 Hz, 5-H), 2.83 (1H,
m, 1-H), 3.54 (1H, dd, J=9.2, 9.5 Hz, 6-H), 4.70, 4.81
(1H each, both s, 14-H2), 4.99, 5.15 (1H each, both br s,
15-H2), 5.25 (1H, dd, J=4.0, 4.3 Hz, 12b-H) (the signals
due to 12b-OH derivative34) 1.10 (3H, d, J=7.0 Hz, 13-
H3), 1.23 (1H, m, 8b-H), 1.56 (1H, m, 7-H), 1.69 (1H,
m, 11-H), 1.82, 1.92 (1H each, both m, 2-H2), 1.97, 2.43
(1H each, both m, 9-H2), 2.00 (1H, m, 8a-H), 2.48, 2.50
(1H each, both m, 3-H2), 2.67 (1H, dd, J=8.5, 9.4 Hz,
5-H), 2.81 (1H, m, 1-H), 3.74 (1H, dd, J=9.4, 9.5 Hz, 6-
H), 4.68, 4.81 (1H each, both s, 14-H2), 4.99, 5.15 (1H
each, both br s, 15-H2), 5.06 (1H, dd, J=4.3, 4.8 Hz,
12a-H). 13C NMR (CDCl3): given in Table 2.
25
2d. colorless oil, ‰ꢀŠ +44.8ꢀ (c=0.1, CHCl3). High-
d
resolution EI±MS: Calcd for C15H22O (M+): 218.1671.
1
Found : 218.1668. IR (®lm): 1028, 887 cm 1. H NMR
(CDCl3) d: 0.91 (3H, d, J=7.0 Hz, 13-H3), 1.39 (1H,
dddd, J=4.5, 8.9, 12.5, 17.7 Hz, 8b-H), 1.66 (1H, dddd,
J=1.8, 4.5, 8.6, 10.3 Hz, 7-H), 1.81 (1H, dddd, J=3.7,
8.9, 12.9, 17.5Hz, 2-H), 1.90 (1H, dddd, J=1.8, 4.0, 8.5,
17.7Hz, 8a-H), 1.99 (1H, dddd, J=5.5, 7.9, 13.4, 17.5Hz,
2-H), 2.09 (1H, dddq, J=5.2, 6.1, 10.3, 7.0 Hz, 11-H),
2.09, 2.35 (1H each, both m, 9-H2), 2.33, 2.54 (1H each,
both m, 3-H2), 2.76 (1H, dd, J=8.0, 8.5 Hz, 5-H), 2.90
(1H, ddd, J=7.9, 8.0, 8.9 Hz, 1-H), 3.23 (1H, dd, J=8.5,
8.6 Hz, 6-H), 3.48 (1H, dd, J=6.1, 11.0 Hz, 12a-H), 3.66
(1H, dd, J=5.2, 11.0 Hz, 12b-H), 4.72 (1H, dd, J=1.8,
1.9 Hz, 14-H), 4.84 (1H, d, J=1.2 Hz, 14-H), 5.03, 5.13
(1H each, both br s, 15-H2). 13C NMR (CDCl3): given
in Table 2. EI±MS m/z (%): 218 (M+, 25), 159 (100).
Measurement of blood ethanol elevation in ethanol-loaded
rats
Test samples were administered orally to fasted (20±22h)
rats. Thirty min thereafter, ethanol (20 v/v%, 5 mL/kg) was
given orally. Blood samples were collected from the infra-
orbital venous plexus at 0.5, 1 and 2 h after the adminis-
tration of ethanol. Then blood was immediately mixed
with 10-fold volume of ice-cold 0.33N perchloric acid and
centrifuged (4 ꢀC, 3000 r.p.m., 10 min). Blood ethanol level
in the supernatant was assessed by the enzymatic method
(F-kit1 ethanol, Boehringer, Mannheim, Germany).

M. Yoshikawa et al. / Bioorg. Med. Chem. 8 (2000) 2071±2077
2077
Measurement of serum glucose elevation in glucose-loaded
rats
11. Matsuda, H.; Kageura, T.; Toguchida, I.; Ueda, H.;
Morikawa, T.; Yoshikawa, M. Life Sci. 2000, 66, 2151.
12. A®®, F. U.; Khalil, E.; Tamimi, S. O.; Disi, A. J. Ethno-
pharmacol. 1997, 58, 9.
13. Gurman, E. G.; Bagirova, E. A.; Storchilo, O. V. Fiziol.
Zh SSSR Im. I. M. Sechenova 1992, 78, 109.
14. Wada, K.; Ueda, N.; Sawada, H.; Amemiya, N.; Haga,
M. Natural Medicines 1997, 51, 283.
15. Preliminary communication on this subject: Matsuda, H.;
Shimoda, H.; Uemura, T.; Yoshikawa, M. Bioorg. Med.
Chem. Lett. 1999, 9, 2647.
16. Ming, C. W.; Mayer, R.; Zimmermann, H.; Rucker, G.
Phytochemistry 1989, 28, 3233.
Test samples were administered orally to fasted rats.
Thirty min thereafter, glucose (1g/kg) was given orally.
Blood samples were collected from the infraorbital venous
plexus at 0.5, 1, and 2 h after the administration of glucose.
Then, blood was centrifuged (4ꢀC, 3000r.p.m., 10 min).
Serum glucose level was assessed by enzymatic method
(Glucose C-II test Wako1, Wako Pure Chemical,
Osaka, Japan).
17. Mathur, S. B.; Hiremath, S. V.; Kulkarni, G. H.; Kelkar,
G. R.; Bhatta, C.; Simonovic, D.; Rao, A. S. Tetrahedron
1965, 21, 3575.
18. Ando, M.; Kusaka, H.; Ohara, H.; Takase, K.; Yamaoka,
H.; Yanagi, Y. J. Org. Chem. 1989, 54, 1952.
19. Farouk, S.E-F.; Dantel, A. B. J. Chem. Soc. Perkin Trans.
1 1983, 355.
20. Romo de Vivar, A. Tetrahedron 1965, 21, 1741.
21. Zdero, C.; Bohlmann, F. Phytochemistry 1990, 29, 183.
22. Rossi, C.; Ambrogi, V.; Grandolini, G.; Scarcia, V.; Fur-
lanai, A. J. Pharm. Sci. 1986, 75, 784.
23. Van Beek, T. A.; Kleis, R.; Posthumus, M. A.; Van Veld-
huizen, A. Phytochemistry 1989, 28, 1909.
Single dose oral toxicity in mice
Various doses of costunolide (1), dehydrocostus lactone
(2), or a-methylene-g-butyrolactone (12) were adminis-
tered orally to 22 h-fasted mice. Mice were continuously
fasted for 24 h and housed in standard conditions for 7
days. General symptoms and death in mice were
observed during the experimental period.
Statistics
Values are expressed as meansÆS.E.M. Signi®cant di€er-
ences were calculated by Dunnett's multiple comparison
test. Probability (P) values less than 0.05 were considered
signi®cant.
24. Barrero, A. F.; Sanchez, J. F.; Ferrol, N. Tetrahedron
Lett. 1989, 30, 247.
25. Warnho€, E. W.; Srinivasan, V. Can. J. Chem. 1973, 51,
3955.
26. Yoshikawa, M.; Hatakeyama, S.; Inoue, Y.; Yamahara, J.
Chem. Pharm. Bull. 1993, 41, 214.
References and Notes
27. Matsuda, H.; Kageura, T.; Inoue, Y.; Morikawa, T.;
Yamahara, J.; Masayuki, Y., to be published.
28. Hikino, H.; Meguro, K.; Kusano, G.; Takemoto, T.
Yakugaku Zasshi 1967, 87, 74.
1. Lejoyeux, M.; Solomon, J.; Ades, J. Alcohol Alcohol 1998,
33, 563.
29. Oshiro, Y.; Sato, S.; Kurahashi, N.; Tanaka, T.; Kikuchi,
T.; Tottori, K.; Uwahodo, Y.; Nishi, T. J. Med. Chem. 1998,
41, 658.
30. Spampinato, U.; Esposito, E.; Samanin, R. J. Neurochem.
1985, 45, 980.
2. Peachey, J. E.; Naranjo, C. A. Drugs 1984, 27, 171.
3. Dupuy, O.; Flocard, F.; Vial, C.; Rode, G.; Charles, N.;
Boisson, D.; Flechaire, A. Rev. Med. Interne. 1995, 16, 67.
4. Yoshikawa, M.; Yamahara, J. In Saponins Used in Tradi-
tional and Modern Medicine; Waller, G. R.; Yamasaki, M.
Ed., Plenum: New York, 1996, pp 207±218.
5. Yoshikawa, M.; Murakami, T.; Harada, E.; Murakami, N.;
Yamahara, J.; Matsuda, H. Chem. Pharm. Bull. 1996, 44, 1915.
6. Yoshikawa, M.; Murakami, T.; Matsuda, H.; Yamahara,
J.; Murakami, N.; Kitagawa, I. Chem. Pharm. Bull. 1996, 44,
1454.
7. Yoshikawa, M.;Murakami, T.; Yoshizumi, S.; Murakami, N.;
Yamahara, J.; Matsuda, H. Chem. Pharm. Bull. 1996, 44, 1899.
8. Yoshikawa, M.; Murakami, T.; Matsuda, H.; Ueno, T.;
Kadoya, M.; Yamahara, J.; Murakami, N. Chem. Pharm.
Bull. 1996, 44, 1305.
9. Yoshikawa, M.; Shimada, H.; Morikawa, T.; Yoshizumi,
S.; Murakami, T.; Matsuda, H.; Hori, K.; Yamahara, J.
Chem. Pharm. Bull. 1997, 45, 1300.
31. Watanabe, K.; Watanabe, H.; Goto, Y.; Shimizu, M.;
Maeda-Hagiwara, M. Jpn. J. Pharmacol. 1983, 33, 1163.
32. Argiolas, A.; Fadda, F.; Melis, M. R.; Marcou, M.; Por-
ceddu, M. L.; Gessa, G. L. Eur. J. Pharmacol. 1982, 85, 23.
33. Dougherty, G. G.; Ellinwood, E. H. Jr. Physiol. Behav.
1983, 30, 607.
34. The 1H NMR and 13C NMR data of 2c (the 12a and 12b-
hydroxyl mixture) were assigned with the aid of homo- and
1
hetero-correlation spectroscopy (1H±1H, H±13C COSY), dis-
tortionless enhancement by polarization transfer (DEPT), and
heteronuclear multiple band correlation (HMBC) experi-
ments. Particularly, the stereostructure of the 12-hydroxyl
1
group in 2c was determined by H NMR nuclear Overhauser
e€ect spectroscopy (NOESY) experiment, which showed NOE
correlations between the 12b-proton and the 6-proton in the
12a-hydroxyl derivative, and between the 12a-proton and the
7-proton in the 12b-hydroxyl derivative.
10. Yoshikawa, M.; Murakami, T.; Ueda, T.; Yoshizumi, S.;
Ninomiya, K.; Murakami, N.; Matsuda, H.; Saito, M.; Fujii,
W.; Tanaka, T.; Yamahara, J. Yakugaku Zasshi 1997, 117, 108.