Synthesis and vasorelaxing evaluation of α-methylidene-γ-butyrolactone bearing quinolin-2(1H)-one and 3,4-dihydroquinolin-2(1H)-one derivatives

Article abstract of DOI:10.1016/S0223-5234(01)01270-3

The main objective of this investigation was to explore the vasorelaxing structure-activity relationships of α-methylidene-γ-butyrolactone bearing quinolin-2(1H)-ones and their 3,4-dihydro derivatives. These target compounds were synthesised in two steps starting from aryl-OH which was treated with a bromomethyl ketone followed by a Reformatsky-type condensation. Quinolin-2(1H)-one α-methylidene-γ-butyrolactones exhibited less vasorelaxing activity than their 3,4-dihydro counterparts. Compounds with a methyl or a phenyl group at the C(γ) of the lactone were more vasorelaxant than the C(γ)-fluorophenyl derivatives in the 3,4-dihydroquinolin-2(1H)-one series. When comparing the positional isomers, α-methylidene-γ-butyrolactone substituted at the 7-position of the 3,4-dihydroquinolin-2(1H)-ones were more active than their 6-substituted counterparts, which in turn were more active than the 8-substituted derivatives. The vasorelaxing effect of these 3,4-dihydroquinolin-2(1H)-ones was proved to be dose dependent. Among them, 7-[(2,3,4,5-tetrahydro-4-methylidene-5-oxo-2-phenyfuran-2-yl)methoxy]- quinolin-2(1H)-one (10b) was the most potent with an IC₅₀ of 9.2 μM on the KCl-induced vasoconstriction of pig coronary arteries.

Full text of DOI:10.1016/S0223-5234(01)01270-3

Eur. J. Med. Chem. 36 (2001) 909–914
´
© 2001 Editions scientifiques et me´dicales Elsevier SAS. All rights reserved
PII: S0223-5234(01)01270-3/SCO
Short communication
Synthesis and vasorelaxing evaluation of a-methylidene-g-butyrolactone bearing
quinolin-2(1H)-one and 3,4-dihydroquinolin-2(1H)-one derivatives
Tai-Chi Wang*, Yue-Ling Zhao, Daih-Huang Kuo
Department of Pharmacy, Tajen Institute of Technology, Pingtung, Taiwan, ROC
Received 28 February 2001; revised 28 June 2001; accepted 7 August 2001
Abstract – The main objective of this investigation was to explore the vasorelaxing structure–activity relationships of a-methylidene-
g-butyrolactone bearing quinolin-2(1H)-ones and their 3,4-dihydro derivatives. These target compounds were synthesised in two
steps starting from aryl-OH which was treated with a bromomethyl ketone followed by a Reformatsky-type condensation.
Quinolin-2(1H)-one a-methylidene-g-butyrolactones exhibited less vasorelaxing activity than their 3,4-dihydro counterparts. Com-
pounds with a methyl or a phenyl group at the C(g) of the lactone were more vasorelaxant than the C(g)-fluorophenyl derivatives
in the 3,4-dihydroquinolin-2(1H)-one series. When comparing the positional isomers, a-methylidene-g-butyrolactone substituted at
the 7-position of the 3,4-dihydroquinolin-2(1H)-ones were more active than their 6-substituted counterparts, which in turn were
more active than the 8-substituted derivatives. The vasorelaxing effect of these 3,4-dihydroquinolin-2(1H)-ones was proved to be
dose dependent. Among them, 7-[(2,3,4,5-tetrahydro-4-methylidene-5-oxo-2-phenyfuran-2-yl)methoxy]-quinolin-2(1H)-one (10b) was
´
the most potent with an IC50 of 9.2 mM on the KCl-induced vasoconstriction of pig coronary arteries. © 2001 Editions scientifiques
et me´dicales Elsevier SAS
vasorelaxant / a-methylidene-g-butyrolactone / quinolin-2(1H)-one / 3,4-dihydroquinolin-2(1H)-one
1. Introduction
one counterparts [3–5]. Searching for novel positive
inotropic quinolin-2(1H)-one drug candidates, Tomi-
Since the discovery of carteolol as an b-adrenergic
naga et al., revealed that a 6-[4-(3,4-dimethoxyben-
blocking agent, a vast number of quinolin-2(1H)-ones
zoyl)-1-piperazinyl]-side chain substituted at the
(carbostyril) and their 3,4-dihydro derivatives had
6-position of the quinolin-2(1H)-one nucleus, exhib-
been synthesised and proved to possess cardiovascu-
ited the most potent positive inotropic effect among
lar activities [1–15]. However, these cardiovascular
their positional isomers and both series of quinolin-
activities of quinolin-2(1H)-one skeletons were influ-
2(1H)-ones and its 3,4-dihydro derivatives were active
enced not only by the nature of the peripheral side
[6]. Although the 3,4-dihydroquinolin-2(1H)-ones ex-
chains but also by the anchoring position. For exam-
hibited less potent inotropic effect than that of their
ple, according to Nishi et al., the 6-substituted isomer
3,4-dehydro counterparts, their lack of chronotropic
exhibited the highest antiplatelet potency while the 7-
side effect led to the discovery of a clinically use-
and 8-substituted isomers were much less active when
ful drug, 3,4-dihydro-6-[4-(3,4-dimethoxybenzoyl)-1-
the side chains were maintained as OCH2CH2CH2-
piperazinyl]quinolin-2(1H)-one (Vesnarinone), for the
COOEt, etc. in the 3,4-dihydroquinolin-2(1H)-one se-
treatment of congestive heart failure. Later on, Fu-
ries [2–4]. They have also concluded that
jioka et al. confirmed that the 6-substituted quinolin-
quinolin-2(1H)-ones possessed higher antiplatelet ac-
2(1H)-ones exhibited the most potent positive
tivity than that of their 3,4-dihydroquinolin-2(1H)-
inotropic effect but did not confirm that the 3,4-dihy-
droquinolin-2(1H)-one series was also active [10].
Over the past few years, we were particularly inter-
ested in synthesizing a-methylidene-g-butyrolactones
* Correspondence and reprints:
E-mail address: tcwang@ccsun.tajen.edu.tw (T.-C. Wang).

910
T.-C. Wang et al. / European Journal of Medicinal Chemistry 36 (2001) 909–914
5-oxofuran-2-yl)methoxy]-quinolin-2(1H)-one (12a)
in 53% overall yield. The same synthetic procedure
was applied for the synthesis of 12d. Synthesis of
compounds 9a–d, 10a–d, 11a–d, and 12b, c were
previously reported [12–15].
that were evaluated for their cardiovascular activities
[12–20]. Our results demonstrated that quinolin-
2(1H)-one a-methylidene-g-butyrolactones 12a–d
were less active as antiplatelet agents than their 3,4-di-
hydro counterparts 10a–d. Moreover, their inhibitory
activities against arachidonic acid (AA)-induced
platelet aggregation decrease in the order 7-substi-
tuted 10b–d>6-substituted 9b–d>8-substituted 11b–
d as shown in table I [14, 15]. We describe herein the
preparation and vasorelaxing evaluation of certain
a-methylidene-g-butyrolactones bearing quinolin-
2(1H)-ones and their 3,4-dihydro counterparts to
classify further the structure–activity relationships in
that series.
3. Biological investigation and discussion
The vasorelaxing effects of these compounds were
represented as percentages of the reduction of the
contraction height induced by 30 mM KCl (table II).
Data are expressed as means S.E.M. The IC50 values
represent the concentration at which 50% reduction in
KCl-induced tone was observed. 6-[(2,3,4,5-Tetrahy-
dro-2-methyl-4-methylene-5-oxofuran-2-yl)methoxy]-
3,4-dihydroquinolin-2(1H)-one (9a) with an aliphatic
Me substituent at the C(g) lactone, and its PhꢀC(g)
analogue (9b) possessed IC50 of 24.5 and 15.8 mM,
respectively. The C(g)-fluorophenyl substituted com-
pound (9c) is less active while the C(g)-biphenyl sub-
stituted one (9d) is completely inactive. The same
pharmacological trend was observed for 12a–d in
which 12a (IC50 =51.8) and 12b (IC50 =38.0) were
more active than 12c, while 12d was inactive. Results
in table II also showed that C(g)-biphenyl derivatives
(10d and 11d) were inactive whereas 10a, 10b, and 11a
2. Chemistry
The preparation of the a-methylidene-g-butyrolac-
tones is illustrated in figure 1. Alkylation of 7-hydroxy-
quinolin-2(1H)-one (4) with bromoacetone under ba-
sic conditions provided 7-(2-oxopropoxy)quinolin-
2(1H)-one (8a) which was then reacted with ethyl
2-(bromomethyl)acrylate and zinc powder in dry te-
trahydrofuran (THF) to afford the Reformatsky
product 7-[(2,3,4,5-tetrahydro-2-methyl-4-methylene-
Table I. IC50 Values (mM) of a-methylidene-g-butyrolactone derivatives on the platelet aggregation induced by AA and PAF.
9a
9b
0.57
2.33
9c
0.60
1.83
9d
3.29
6.23
10a
2.31
51.4
10b
0.23
6.13
10c
0.28
3.54
10d
1.91
11.4
AA
PAF
1.64
13.3
11a
11b
4.39
21.7
11c
3.75
15.6
11d
7.73
78.0
12a
12b
0.50
3.30
12c
0.40
5.00
12d
AA
PAF
35.7
3.90
2.40
\100
53.3
14.7
Figure 1.

T.-C. Wang et al. / European Journal of Medicinal Chemistry 36 (2001) 909–914
911
Table II. Inhibitory effects of a-methylidene-g-butyrolactone derivatives on KCl induced contraction of pig coronary arteries.
3 mM
10 mM
30 mM
100 mM
IC50
Control
9a
101.595.2
103.898.9
94.197.6
70.097.1
0.090.0 a
0.090.0 a
56.393.1
69.098.1
0.090.0 a
0.090.0 a
0.090.0 a
79.596.4
0.090.0 a
69.899.6
20.591.7 a
72.599.0
0.090.0 a
0.090.0 a
41.7910.2 a
86.898.3
0.090.0 a
–
91.09.5.0
65.099.0 a
27.0912.0 a
24.596.8
15.893.9
9b
1100.092.0
54.098.0 a
7.096.0 a
–
9c
9d
–
–
–
–
–
–
–
10a
10b
10c
10d
11a
11b
11c
11d
12a
12b
12c
12d
Nifedipine
78.096.6 b
58.0910.9 a
30.3912.4 a
23.299.8
85.097.8 c
28.596.9 a
0.090.0 a
9.291.2
86.7911.3
52.4911.6 a
12.7913.0 a
16.897.2
–
–
–
–
119.0910.1 c
84.395.6 b
53.8912.3 a
44.997.0
–
–
–
–
–
–
–
–
–
–
–
–
102.095.2
92.095.7
82.095.0 b
55.096.5 a
51.899.0
38.097.1
78.093.5 a
43.099.9 a
–
–
–
–
–
–
–
–
–
–
–
–
a Significantly different from control value at PB0.001.
b Significantly different from control value at PB0.01.
c Significantly different from control value at PB0.05.
were active vasorelaxants with IC50 of 23.2, 9.2, and
44.9 mM, respectively. These results are interesting,
since earlier studies indicated that these compounds
also have antiplatelet and cytotoxic activities, with the
g-fluorophenyl lactones being better antiplatelet
agents than their corresponding g-methyl counter-
parts [12–20] while the g-biphenyl lactones were the
most cytotoxic [21–23].
them, 7-[(2,3,4,5-Tetrahydro-4-methylidene-5-oxo-2-
phenyfuran-2-yl)methoxy]-quinolin-2(1H)-one (10b)
was the most potent, being able to completely inhibit
pig coronary arterial contraction induced by 30 mM
KCl at a concentration of 30 mM. Preparation of its
pure enantiomers for biological evaluation is cur-
rently ongoing.
According to previous results [6, 10], 3,4-dihy-
droquinolin-2(1H)-one containing compounds were
weaker inotropic agents than their unsaturated coun-
terparts. However, we found here the opposite results
for vasorelaxing and antiplatelet activities. Quinolin-
2(1H)-ones 12a–c exhibited less vasorelaxing activity
than their 3,4-dihydro counterparts 10a–c. Previous
reports have also indicated that the peripheral
aminoalcohol or ester type of side chain at the 6-posi-
tion of the 3,4-dihydroquinolin-2(1H)-one moiety re-
sulted in the most potent inotropic and antiplatelet
activities [1–11]. However, according to our results,
a-methylidene-g-butyrolactones substituted at the 7-
position of the 3,4-dihydroquinolin-2(1H)-ones, e.g.
10a–c, were more active than their 6-substituted
counterparts 9a–c, which in turn were more active
than the 8-substituted derivatives 11a–c.
4. Conclusions
Some
a-methylidene-g-butyrolactones
bearing
quinolin-2(1H)-ones and their 3,4-dihydro derivatives
10a, 10b, 11a, 12a, and 12b exhibited a complete
inhibition of the contraction height induced by 30
mM KCl at a concentration of 100 mM while others
9d–12d at the same concentration, were inactive.
Their vasorelaxing activities were influenced not only
by the nature of the peripheral lactones but also by
the position of the lactone chains attached to it. The
quinolin-2(1H)-one or its fragment 3,4-dihydro equiv-
alent may be considered as the key determinant of the
pharmacophore, while the lactones characterized both
by the g-substituents and their anchoring positions,
are believed to play a modulatory role with a prefer-
ence for a methyl or a phenyl group. In contrary to
previous reports, 3,4-dihydroquinolin-2(1H)-one con-
The vasorelaxing effects of these 3,4-dihydroquino-
lin-2(1H)-ones proved to be dose dependent. Among

912
T.-C. Wang et al. / European Journal of Medicinal Chemistry 36 (2001) 909–914
1
taining compounds were more potent vasorelaxing
agents than that of their unsaturated counterparts.
These vasorelaxing activities decreased in the order
7-substituted>6-substituted>8-substituted.
194–195 °C; H-NMR (DMSO): l 5.69 (s, CH2O), 6.31
(d, J=9.4 Hz, HꢀC(3)), 6.78–8.14 (m, 12 arom. H),
7.81 (d, J=9.4 Hz, HꢀC(4)), 11.50 (br s, NH); 13C-
NMR (DMSO): 70.35 (CH2O), 99.34, 110.60, 113.68,
118.80, 127.06, 128.56, 128.64, 129.15, 129.25, 133.07,
138.80, 139.96, 140.48, 145.29 (arom. C), 159.72 (C-7),
162.18 (C-2), 193.72 (C-2%); Anal. Calc. for
C23H17NO3·0.25H2O: C, 76.76; H, 4.90; N, 3.89; Found:
C, 76.43; H, 4.93; N, 3.88%.
5. Experimental protocols
5.1. Chemistry
Melting points (m.p.) were determined on an Yanaco
micromelting-point apparatus and are uncorrected. The
ultraviolet (UV) (umax (log m) in nm) absorption spectra
were obtained in a Beckman UV–vis spectrophotome-
ter. Nuclear magnetic resonance (NMR) (1H and 13C)
spectra were obtained with a Varian Gemini-200 spec-
trometer. Chemical shifts were expressed in parts per
million (l) with TMS as an internal standard. Thin-
layer chromatography (TLC) was run on precoated (0.2
mm) silica gel 60 F-254 plates manufactured by EM
Laboratories, Inc., and short wave UV light (254 nm)
was used to detect the UV-absorbing spots. Elemental
analyses (C, H, N) were carried out in a Heraeus
CHN-O-Rapid elemental analyser and the results were
within 0.4% of theoretical values.
5.1.3. Synthesis of 7-[(2,3,4,5-tetrahydro-2-methyl-4-
methylidene-5-oxofuran-2-yl)methoxy]quinolin-2(1H)-
one (12a)
To a solution of 8a (0.65 g, 3 mmol) in dry THF (60
mL) were added activated Zn powder (0.26 g, 3.9
mmol), hydroquinone (6 mg), and ethyl 2-(bro-
momethyl)acrylate (0.78 g, 4 mmol). The mixture was
refluxed under N2 for 6 h (TLC monitoring). After
cooling, it was poured into an ice-cold 5% HCl solution
(300 mL) and extracted with CH2Cl2 (3×60 mL). The
CH2Cl2 extracts were combined and washed with H2O,
dried (Na2SO4), and then evaporated to give a residual
solid, which was crystallised from a 1:10 mixture
CH2Cl2–Et2O to afford 12a (0.66 g, 77% yield); m.p.:
147–148 °C; UV: 248 (sh, 3.98), 333 (4.11) (0.1 N
HCl–MeOH), 281 (3.67), 324 (4.04), 338 (3.93)
(MeOH), 233 (sh, 4.51), 325 (3.94) (0.1 N NaOH–
5.1.1. Synthesis of 7-(2-oxopropoxy)quinolin-2(1H)-one
(8a)
1
7-Hydroxyquinolin-2(1H)-one (4; 1.61 g, 10 mmol),
K2CO3 (1.38 g, 10 mmol) and dry DMF (50 mL) were
stirred at room temperature (r.t.) for 30 min. To this
solution was added bromoacetone (1.37 g, 10 mmol) in
dry DMF (10 mL) in one portion. The resulting mixture
was stirred continuously at r.t. for 24 h. (TLC monitor-
ing) and then poured into ice water (100 mL). The white
solid thus obtained was collected and crystallised from
CH2Cl2 to afford 8a (1.50 g, 69%); m.p.: 160–161 °C;
1H-NMR (DMSO): l 2.18 (s, Me), 4.87 (s, CH2O), 6.31
(d, J=9.6 Hz, HꢀC(3)), 6.71–7.59 (m, 3 arom. H), 7.81
(d, J=9.6 Hz, HꢀC(4)), 11.55 (br s, NH); 13C-NMR
(DMSO): l 26.20 (Me), 72.17 (CH2O), 99.18, 110.30,
113.66, 118.83, 129.26, 139.91, 140.44 (arom. C), 159.41
(C-7), 162.14 (C-2), 203.51 (C-2%); Anal. Calc. for
C12H11NO3·0.125H2O: C, 65.67; H, 5.17; N, 6.38;
Found: C, 65.82; H, 5.25; N, 6.45%.
MeOH); H-NMR (DMSO): l 1.48 (s, Me), 2.80 (dt,
J=17.4 Hz, 2.8, HꢀC(3%)), 3.08 (dt, J=17.4 Hz, 2.6,
HꢀC(3%)), 4.11 (s, CH2O), 5.75 (t, J=2.6 Hz, 1H,
CH2=C(4%)), 6.07 (t, J=2.6 Hz, 1H, CH2ꢁC(4%)), 6.31
(d, J=9.4 Hz, HꢀC(3)), 6.74–7.59 (m, 3 arom. H), 7.81
(d, J=9.4 Hz, HꢀC(4)), 11.59 (br s, NH); 13C-NMR
(DMSO): l 23.34 (Me), 35.97 (C-3%), 72.78 (CH2O),
81.70 (C-2%), 99.27, 110.45, 113.82, 118.92, 121.26,
129.46, 136.02, 140.06, 140.51 (arom. C), 159.84 (C-7),
162.30 (C-2), 169.29 (C-5%); Anal. Calc. for C16H15NO4:
C, 67.36; H, 5.30; N, 4.91; Found: C, 67.12; H, 5.28; N,
4.92%.
The same procedure was used to convert 8d to 12d.
5.1.4. 7-[(2,3,4,5-Tetrahydro-4-methylidene-5-oxo-2-
(1,1%-biphenyl-4-yl)furan-2-yl)methoxy]quinolin-2(1H)-
one (12d)
5.1.2. 7-[2-Oxo-2-(1,1%-biphenyl-4-yl)ethoxy]-
quinolin-2(1H)-one (8d)
Prepared from 2-bromo-4%-phenylacetophenone by the
Yield: 76%; m.p.: 215–216 °C; UV: 250 (sh, 4.60),
333 (4.29) (0.1 N HCl–MeOH), 251 (4.54), 324 (4.24),
338 (4.13) (MeOH), 236 (sh, 4.81), 325 (4.11) (0.1 N
NaOH–MeOH); 1H-NMR (DMSO): l 3.22 (dt, J=
same procedure as described for 8a in 60% yield; m.p.:

T.-C. Wang et al. / European Journal of Medicinal Chemistry 36 (2001) 909–914
913
17.2 Hz, 2.8, HꢀC(3%)), 3.65 (dt, J=17.2 Hz, 2.8,
HꢀC(3%)), 4.34, 4.46 (AB type, J=10.4 Hz, CH2O), 5.82
(t, J=2.6 Hz, 1H, CH2ꢁC(4%)), 6.14 (t, J=2.6 Hz, 1H,
CH2ꢁC(4%)), 6.31 (d, J=9.4 Hz, HꢀC(3)), 6.77–7.77 (m,
12 arom. H), 7.80 (d, J=9.4 Hz, HꢀC(4)), 11.53 (br s,
NH); 13C-NMR (DMSO): l 37.11 (C-3%), 73.32 (CH2O),
84.09 (C-2%), 99.44, 110.40, 113.84, 118.95, 121.63,
125.77, 126.73, 126.90, 127.69, 128.97, 129.38, 134.92,
139.42, 139.94, 140.14, 140.42 (arom. C), 159.60 (C-7),
162.16 (C-2), 168.90 (C-5%); Anal. Calc. for
C27H21NO4·0.125H2O: C, 76.18; H, 5.03; N, 3.29;
Found: C, 75.99; H, 5.08; N, 3.34%.
5.2.2. In vitro vasorelaxing evaluation [26]
Porcine hearts were obtained from a local abattoir
within 30 min of slaughter and were transported to the
laboratory in Krebs-Henseleit (KH) buffer solution with
the following composition (mM): NaCl (120), KCl (5.2),
KH2PO4 (1), MgSO4 (1.3), CaCl2 (2.5), NaHCO3 (15.5),
glucose (11.3), and pyruvate (1). The pH value of the
KH buffer solution was adjusted to 7.4. The right
coronary and anterior descending branch of the left
coronary arteries were dissected from each heart and
stored overnight at 4 °C in an oxygenated KH buffer
solution. On one of the two following days, these arter-
ies were cleaned of any remaining connective tissue and
cut into 3-mm rings.
5.2. Biology
Coronary arterial rings were suspended in organ bath
filled with 20 mL KH solution. The bath solution was
gassed with a mixture of 95% O2 and 5% CO2 and the
temperature was maintained at 37 °C throughout the
experiment. Each ring was suspended by two fine stain-
less-steel wire clips; one clip was anchored inside the
organ bath and the other was connected to a force
transducer (model FT03, Grass Instrument, USA). Iso-
metric tension was measured by Cyber 380 and Digidata
1320A (Axon Instrument, USA) and recorded in a
computer.
5.2.1. In vitro antiplatelet evaluation
5.2.1.1. Reagents
Arachidonic acid (AA), EDTA, and bovine serum
albumin were purchased from Sigma Chem. Co.
Platelet-activating factor (PAF) was purchased from
Calbiochem-Behring Co.
5.2.1.2. Platelet aggregation
Blood was collected from the rabbit marginal ear
vein, anticoagulated with EDTA (6 mM) and cen-
trifuged for 10 min at 90×g and at room temperature.
Platelet suspensions were prepared from the plasma
according to the washing procedures previously de-
scribed [24]. Platelet numbers were determined with a
Coulter counter (Model ZM) and adjusted to 4.5×108
platelets mL−1. The platelet pellets were suspended in
Tyrode’s solution of the following composition (mM):
NaCl (136.8), KCl (2.8), NaHCO3 (11.9), MgCl2 (2.1),
NaH2PO4 (0.33), CaCl2 (1.0) and glucose (11.2), con-
taining bovine serum albumin (0.35%). The platelet
suspension was stirred at 1200 rpm and the aggregation
was measured at 37 °C by the turbidimetric method as
described by O’Brien [25] using a Chrono-Log Lumi-ag-
gregometer. To eliminate solvent effects on aggregation,
the final concentration of dimethylsulfoxide (DMSO)
was fixed at 0.5%. The percentage of aggregation was
calculated using the absorbance of a platelet suspension
as 0% aggregation and the absorbance of Tyrode’s
solution as 100% aggregation.
Tissues were allowed to equilibrate for a minimum of
1 h before testing had begun. KCl (30 mM) was then
added into the organ chamber to contract these rings.
When the contraction reached a stable plateau (usually
15 min), compounds (100 mM) were added in organ bath
to screen the activity of relaxation. Concentration–re-
sponse curves for some more potent compounds were
established by cumulative addition at doses of 3, 10, 30
and 100 mM.
Contractile responses were calculated as the difference
between resting tension and maximum tension devel-
oped in response to KCl stimulation. Data are expressed
as mean S.E.M. from a number (n=4–6) of experi-
ments. Statistical analyses were performed using Student
test. Pꢀ0.05 was considered significant.
Acknowledgements
Financial support of this work by the National Sci-
ence Council of the Republic of China is gratefully ac-
knowledged. We also thank the National Center for
High-Performance Computing for providing computer
resources and chemical database services.
The IC50 values (mM) for 12a and 12d were as follows:
3.90 and 2.40, respectively, on the platelet aggregation
induced by AA; 53.3 and 14.7, respectively, that of
induced by PAF.

914
T.-C. Wang et al. / European Journal of Medicinal Chemistry 36 (2001) 909–914
[13] Wang T.C., Chen Y.L., Tzeng C.C., Liou S.S., Tzeng W.F.,
References
Chang Y.L., Teng C.M., Helv. Chim. Acta 81 (1998) 1038–
1047.
[1] Tominaga M., Tone H., Nakagawa K., Takada K., Hoshino
Y., Watanabe K., Chem. Pharm. Bull. 29 (1981) 2166–2181.
[14] Chen Y.L., Wang T.C., Fang K.C., Chang N.C., Tzeng C.C.,
Heterocycles 50 (1999) 453–462.
[2] Nishi T., Yamamoto K., Shimizu T., Kanbe T., Kimura Y.,
Nakagawa K., Chem. Pharm. Bull. 31 (1983) 798–810.
[15] Tzeng C.C., Chen I.L., Chen Y.L., Wang T.C., Chang Y.L.,
[3] Nishi T., Tabusa F., Tanaka T., Shimizu T., Kanbe T., Kimura
Teng C.M., Helv. Chim. Acta 83 (2000) 349–358.
Y., Nakagawa K., Chem. Pharm. Bull. 31 (1983) 1151–1157.
[16] Chen Y.L., Wang T.C., Liang S.C., Teng C.M., Tzeng C.C.,
[4] Nishi T., Tabusa F., Tanaka T., Ueda H., Shimizu T., Kanbe
T., Kimura Y., Nakagawa K., Chem. Pharm. Bull. 31 (1983)
852–860.
Chem. Pharm. Bull. 44 (1996) 1591–1599.
[17] Chen Y.L., Wang T.C., Lee K.H., Chang Y.L., Teng C.M.,
Tzeng C.C., Helv. Chim. Acta 79 (1996) 651–657.
[5] Nishi T., Tabusa F., Tanaka T., Shimizu T., Nakagawa K.,
[18] Wang T.C., Chen Y.L., Liou S.S., Chang Y.L., Teng C.M.,
Chem. Pharm. Bull. 33 (1985) 1140–1147.
Tzeng C.C., Helv. Chim. Acta 79 (1996) 1620–1626.
[6] Tominaga M., Yo E., Ogawa H., Yamashita S., Yabuuchi Y.,
[19] Liou S.S., Zhao Y.L., Chang Y.L., Teng C.M., Tzeng C.C.,
Nakagawa K., Chem. Pharm. Bull. 32 (1984) 2100–2110.
Chem. Pharm. Bull. 45 (1997) 1777–1781.
[7] Tominaga M., Ogawa H., Yo E., Yamashita S., Yabuuchi Y.,
[20] Tzeng C.C., Zhao Y.L., Chen Y.L., Liou S.S., Wang T.C.,
Chang Y.L., Teng C.M., Helv. Chim. Acta 80 (1997) 2337–
2344.
Nakagawa K., Chem. Pharm. Bull. 35 (1987) 3699–3704.
[8] Alabaster C.T., Bell A.S., Campbell S.F., Ellis P., Henderson
C.G., Roberts D.A., Ruddock K.S., Samuels G.M.R., Stefa-
niak M.H., J. Med. Chem. 31 (1988) 2048–2056.
[21] Wang T.C., Lee K.H., Chen Y.L., Liou S.S., Tzeng C.C.,
Bioorg. Med. Chem. Lett. 8 (1998) 2773–2776.
[9] Alabaster C.T., Bell A.S., Campbell S.F., Ellis P., Henderson
C.G., Morris D.S., Roberts D.A., Ruddock K.S., Samuels
G.M.R., Stefaniak M.H., J. Med. Chem. 32 (1989) 575–583.
[22] Lee K.H., Huang B.R., Tzeng C.C., Bioorg. Med. Chem. Lett.
9 (1999) 241–244.
[23] Tzeng C.C., Lee K.H., Wang T.C., Han C.H., Chen Y.L.,
[10] Fujioka T., Teramoto S., Mori T., Hosokawa T., Sumida T.,
Tominaga M., Yabuuchi Y., J. Med. Chem. 35 (1992) 3607–
3612.
[11] Uno T., Ozeki Y., Koga Y., Chu G.N., Okada M., Tamura K.,
Igawa T., Unemi F., Kido M., Nishi T., Chem. Pharm. Bull. 43
(1995) 1724–1733.
Pharm. Res. 17 (2000) 715–719.
[24] Teng C.M., Ko F.N., Thromb. Haemostasis 59 (1988) 304–
309.
[25] O’Brien J.R., J. Clin. Pathol. 15 (1962) 452–455.
[26] Makujina S.R., Olsson R.A., Esinhart J.D., Mustafa S.J., Eur.
[12] Tzeng C.C., Wang T.C., Chen Y.L., Wang C.J., Chang Y.L.,
J. Pharmacol. 243 (1993) 35–38.
Teng C.M., Helv. Chim. Acta 80 (1997) 1161–1168.