Design, synthesis and pharmacological evaluation of 3-benzylazetidine-2-one-based human chymase inhibitors

Article abstract of DOI:10.1016/S0968-0896(01)00209-7

3-Benzylazetidine-2-one derivatives were designed and evaluated as a novel series of chymase inhibitors. Structure-activity relationship studies of 3-benzylazetidine-2-ones led to compounds 23, which exhibited 3.1nM inhibition of human chymase and enhancement of stability in human plasma (t_1/2 6 h). Copyright

Full text of DOI:10.1016/S0968-0896(01)00209-7

Bioorganic & Medicinal Chemistry 9 (2001) 3065–3075
Design, Synthesis and Pharmacological Evaluation of 3-
Benzylazetidine-2-one-based Human Chymase Inhibitors
Yasunori Aoyama,a,* Masaaki Uenaka,a Makoto Kii,a Mamoru Tanaka,a
Toshiro Konoike,b,* Yoko Hayasaki-Kajiwara,c Noriyuki Nayac and Masatoshi Nakajimac
aShionogi Research Laboratories, Shionogi & Co., Ltd., Fukushima-ku, Osaka 553-0002, Japan
bShionogi Research Laboratories, Shionogi & Co., Ltd., Amagasaki, Hyogo 660-0813, Japan
cShionogi Research Laboratories, Shionogi & Co., Ltd., Futaba-cho, Toyonaka, Osaka 561-0825, Japan
Received 26 April 2001; accepted 22 June 2001
Abstract—3-Benzylazetidine-2-one derivatives were designed and evaluated as a novel series of chymase inhibitors. Structure–
activity relationship studies of 3-benzylazetidine-2-ones led to compounds 23, which exhibited 3.1 nM inhibition of human chymase
and enhancement of stability in human plasma (t1/2 6 h). # 2001 Elsevier Science Ltd. All rights reserved.
Introduction
insights we gained with regard to the inhibition of
human chyamse by 1-oxacephems9a and 1,3-diazetidine
2,4-diones,9b and the inhibition of human leukocyte
elastase by cephalosporin sulfones10 and 3,3-dialkylaze-
tidine-2-ones11 developed at Merck. Compound 10a
showed high activity against human chymase, but poor
stability in human plasma. Structure–activity relation-
ship studies of 3-benzylazetidine-2-ones such as 10a led
us to compound 23, which exhibited high potency
against human chymase and enhancement of stability in
human plasma. We describe herein the design, synthesis
and pharmacological evaluation of 3-benzylazetidine-2-
one-based human chymase inhibitors.
Human chymase is a chymotrypsin-like serine protease
that is stored in the secretory granules of mast cells.1
Although the physiological and pathological roles of
chymase have not been fully elucidated, recent studies
have demonstrated that this enzyme generates angio-
tensin II from angiotensin I with greater efficiency than
angiotensin I converting enzyme.2 Chymase has also
been shown to participate in histamine release from
mast cells,3 activation of precursor interleukin-1b,4 and
cleavage of type I procollagen5 and progelatinase B.6
Thus, chymase is speculated to play an important role
in cardiovascular diseases and chronic inflammation
following fibrosis, such as cardiac, renal, and pulmon-
ary fibrosis.7 Chymase inhibitors8,9 are thought to be
potentially useful as tools for elucidating the physi-
ological functions of chymase and therapeutic agents.
Chemistry
The molecular structures and synthetic schemes of 3-
benzylazetidine-2-one-based human chymase inhibitors
are summarized in Scheme 1. Compound 4 was pre-
pared from d-aspartic acid by a known procedure.12
Deprotonation of 4 with 2equiv of lithium diisopropy-
lamide (LDA) followed by benzylation with benzyl
bromide selectively introduced a benzyl group trans to
the carboxyl to afford 5 in 98% yield. Lead tetraacetate
(LTA) oxidation of the carboxylic acid 5 followed by
removal of the tert-butyldimethylsilyl (TBDMS) N-pro-
tecting group with n-tetrabutylammonium fluoride
(TBAF) provided the 4a-acetoxy derivative 6 in 53%
yield. Displacement of the acetoxy 6 with the preformed
sodium salt of diphenylmethyl 4-hydroxybenzoate,
In previous papers, we reported on a novel series of 1-
oxacephem-based9a and 1,3-diazetidine-2,4-dione-based9b
human chymase inhibitors. 1-Oxacephem 1 and 1,3-
diazetidine 2,4-dione 2 exhibited high activities against
human chymase (IC50: 27 and 4 nM), but had insuffi-
cient stability in human plasma (t1/2 1.5 and <0.5 h).
Then, we designed a new class of azetidine-2-one-based
human chymase inhibitor 10a (Fig. 1) on the basis of the
*Corresponding authors. Tel.: +81-6-6458-5861; fax: +81-6-6458-
0987; e-mail: yasunori.aoyama@shionogi.co.jp; toshiro.konoike@
shionogi.co.jp
0968-0896/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(01)00209-7

3066
Y. Aoyama et al. / Bioorg. Med. Chem. 9 (2001) 3065–3075
Scheme 1. Reagents and conditions: (a) see ref 12; (b) LDA (2 equiv), THF, BnBr; (c) (i) LTA, DMF, AcOH; (ii) TBAF, THF; (d) 4-HOC6H4-
CO2CHPh2, NaOH in H2O–acetone; (e) (R)-1-phenylethyl isocyanate, TEA, DMAP, CH2Cl2; (f) (S)-1-phenylethyl isocyanate, TEA, DMAP,
CH2Cl2; (g) TFA, anisole; (h) 1-methylpiperadine, WSCD, CH2Cl2; (i) LDA (2equiv), THF, EtI; (j) LDA (2equiv), THF, 2-(MeO)C 6H4CH2Br; (k)
LDA (2equiv), THF, 2-(EtO)C 6H4CH2Br; (l) LDA (2equiv), THF, 2-( i-PrO)C6H4CH2Br; (m) (R)-phenylpropyl isocyanate, TEA, DMAP, CH2Cl2;
(n) diphenylmethyl isocyanate, TEA, DMAP, CH2Cl2.
prepared from 4-hydroxybenzoic acid and diphenyldia-
zomethane, afforded 3,4-trans isomer 7a as the major
product (74% yield) and 3,4-cis isomer 7b as a minor
product (17% yield). Each isomer (7a and 7b) was indi-
vidually acylated with (R)- or (S)-1-phenylethyl isocya-
nate in the presence of triethylamine (TEA) and 4-
dimthylaminopyridine (DMAP) to afford the ureas 8a–
d. Conversion to 4-benzoic acids 9a–d was achieved by
trifluoroacetic acid (TFA) removal of the diphe-
nylmethyl ester in anisole. 4-Amides 10a–d were
Figure 1. Design of 3-benzylazetidine-2-one-based human chymase inhibitor.

Y. Aoyama et al. / Bioorg. Med. Chem. 9 (2001) 3065–3075
3067
obtained by treatment with 9a–d and 1-methylpiper-
adine in the presence of 1-(3-dimethylaminopropyl)-3-
ethylcarbodiimide (WSCD). Compounds 11a and 11b
were prepared from l-aspartic acid by the above proce-
dures. Alkylation of the dianion of compound 4 with
ethyl iodide provided 3b-ethyl derivative and second
alkylation with benzyl bromide selectively introduced a
benzyl group trans to the carboxyl to afford compound
12. 3,3-Disubstituted 2-azetidinone 13 was prepared
from compound 12 by the above procedures. Alkylation
of dianion of compound 4 with ortho-alkoxybenzyl
bromide, prepared from the corresponding benzyl alco-
hol and hydrogen bromide, provided carboxylic acids
14, 15 and 16. Compounds 17, 18 and 19 were prepared
from compounds 14, 15 and 16 respectively. Each iso-
mer (17, 18 and 19) was individually acylated with (R)-
1-phenylethyl isocyanate or (R)-1-phenylpropyl isocya-
nate or diphenylmethyl isocyanate, prepared from the
corresponding carboxylic acids by the Shioiri method,13
in the presence of triethylamine and 4-dimthylamino-
pyridine to afford the ureas 20–24.
kocyte elastase. The 1-substituent of azetidine-2-
one-based elastase inhibitors such as L-694,458
occupies the S0 region of elastase and the 4-sub-
stituent acts as the leaving group and does not
strictly interact with the enzyme.11 Based on the
above results, the substituent effects at the 1- and
4-positions of azetidine-2-one-based chymase inhi-
bitors are thought to be similar to those of azeti-
dine-2-one-based elastase inhibitors. We selected
(R)-1-phenylethyl urea easily prepared from com-
mercially available (R)-1-phenylethyl isocyanate
and the same substituent as that of L-694,458 as
the 1- and 4-substituents of azetidine-2-one-based
chymase inhibitors, respectively.
First, we examined the inhibitory activity of the
designed compound 10a (Table 1). As expected, com-
pound 10a possessed high potency against human chy-
mase (IC50 0.46 nM). Next, diastereomers of 10a were
evaluated. (3b, 4b)-Isomer 10b displayed 3-fold increase
compared to (3b, 4a)-isomer 10a. (3a, 4a)-Isomer 11a
and (3a, 4b)-isomer 11b were 80- and 2-fold less potent
than (3b, 4a)-isomer 10a, respectively. This suggests
that 3b-isomers are more potent than 3a-isomers. (10S)-
Isomers 10c and 10d led to a greater loss of potency than
the corresponding (10R)-isomers 10a and 10b. Com-
pound 10b displayed the best IC50 value of 0.17 nM, but
stereoselective synthesis of the 3,4-cis isomer such as 10b
is more difficult than that of the 3,4-trans isomer such as
10a as shown in Scheme 1. The above results suggest the
optimal compound to be (10R, 3b, 4a)-isomer 10a.
However, compound 10a was unstable in human
plasma. We next investigated enhancement of the stabi-
lity in human plasma by modifying the 10-, 3- and 4-
positions of compound 10a (Table 2).
Results and Discussion
We designed compound 10a as an azetidin-2-one-based
human chymase inhibitor based on the following con-
siderations (Fig. 1).
i. Human chymase and human elastase belong to the
same family of serine proteases. With regard to b-
lactam-based chymase or elastase inhibitors, sub-
stituents at the a-position of b-lactam carbonyl are
situated in the S1 pocket of the enzymes.9ꢀ11 1,3-
Diazetidine 2,4-dione 39b and 3,3-diethylazetidine-
2-one L-694,45811c are elastase inhibitors, which
do not possess inhibitory activities against human
chymase.9b,14 The N-ethyl of 3 and the 3-ethyl of
L-694,458, substituents at the a-position of the b-
lactam carbonyl group, determine the selectivity
against human elastase. The 7b-phenyl ring of 1-
oxaxephems such as compound 1 and N-benzyl
substituent of 1,3-diazetidine 2,4-diones such as
compound 2 are essential for inhibition against
human chymase.9 These results agree with the pri-
mary substrate specificities of human chymase
(Phe, Thy)9b,15 and human elastase (Val, Leu).9b,16
Accordingly, the 3b-benzyl substituent is thought
to be required for azetidine-2-one-based human
chymase inhibitors.
First, 3,3-disubstituent compound 13 was prepared and
evaluated, because the Merck group reported that 3,3-
disubstituents improved the hydrolytic stability of aze-
tidin-2-one.11b As expected, the 3a-ethyl group of com-
pound 13 dramatically enhanced the stability in human
plasma, however, it markedly decreased the potency
against human chymase. Next, 4-substituent was mod-
ified. 4-Carboxylic acid 9a mildly improved the stability.
Table 1. Modifications at 10-, 3-, 4-positions
ii. The structure–activity relationships at the 30- and
4-positions of 1-oxacephem chymase inhibitors
displayed a similar tendency compared to those of
the cephalosporin elastase inhibitors.9a,10 4-Sub-
stituents of 1-oxacephems and cephalosporins
occupy the prime site region (S0 region) of each
enzyme and 3-substituents function as the leaving
group and do not strictly interact with enzymes.
This may suggest that the structure of the S0 region
of human chymase resembles that of human leu-
Compd
10-position
3-position
4-position
IC50
(nM)
Stability
t1/2 (h)
10a
10b
10c
10d
11a
11b
R
R
S
S
R
R
b
b
b
b
a
a
a
b
a
b
a
b
0.46
0.17
0.66
13
36
0.80
1.0
0.9
0.5
2.1
2.1
<0.5

3068
Y. Aoyama et al. / Bioorg. Med. Chem. 9 (2001) 3065–3075
The 3-substituent was then modified with the 4-car-
boxylic acid. Introduction of an alkoxy group at the
ortho-position of the 3-benzyl substituent was tried,
because an ortho-alkoxy group on 7b-benzamide of 1-
oxacephem-based human chymase inhibitors such as
compound 1 (Scheme 1) increased the potency against
human chymase and the stability in human plasma.9a
As expected, 3-ortho-(methoxy)benzyl 20 showed
enhancement of the stability. We next examined the
substituent effect at the 10-position. Replacement of the
methyl group by an ethyl group at the 10-position of 21
resulted in dramatic improvement of the stability.
Introduction of an ethoxy group at the 3-position of 22
led to further enhancement of the stability. Finally, 1-
diphenylmethyl urea 23, which did not possess an
asymmetric carbon atom at the 1-position, offered suf-
ficient potency against human chymase (IC50 3.1 nM)
and stability in human plasma (t1/2 6 h). 3-ortho-(iso-
Propoxy)benzyl 24 displayed remarkable improvement
of the stability, but considerable reduction of the activ-
ity. Additional enzymatic work showed that compound
23 was a selective inhibitor, causing weak or no inhibi-
tion of several other serine proteases (Table 3).
of human chymase activity (Fig. 3). Accordingly,
enhancement of the stability in human plasma is inversely
proportional to the increase of the potency against human
chymase (Table 2). This tendency was also observed for
1-oxacephem-based human chymase inhibitors.9a
A model for the inhibition by compound 23 of human
chymase, using atomic coordinates from the published
X-ray structure of the enzyme,17 is shown in Figure 4.18
In this model, the inhibitor 23 is bound to the active site
of the enzyme with the hydroxy of Ser195 and the
resulting hemiketal oxygen is in the oxyanion hole
formed by the backbone amide NH of Ser195 and
Gly193. The benzyl substituent at the 3-position is fully
enclosed by the residues of the S1 pocket and capped
with the side chain of Phe191. The phenoxy substituent
at the 4-position occupies the S2site with the carboxyl
group extending into the solution. The urea substituent
Table 2. Modifications at 10-, 3-, 4-positions
Table 2suggests that the a-branched substituent at the
1-benzylurea and the ortho-alkoky substituent on the 3-
benzene ring affect the activity against human chymase
and the stability in human plasma. Steric hindrance at
the 1- and 3-positions may prevent a variety of nucleo-
philes in human plasma from attacking the b-lactam
carbonyl group followed by cleavage of the b-lactam
ring and result in enhancement of the stability in human
plasma (Fig. 2). However, the steric hindrance also hin-
ders nucleophilic substitution of the active serine
hydroxyl group in human chymase followed by genera-
tion of an acyl-enzyme, which is necessary for inhibition
Compd
R
X
Y
Z
IC50
(nM)
Stability
t1/2 (h)
10a
13
9a
20
21
22
23
24
H
Et
H
H
H
H
H
H
A
A
H
H
H
OMe
OMe
OEt
OEt
OPr-i
Me
Me
Me
Me
Et
Et
Ph
Ph
0.46
190
11
1.0
0.5
2.1
3.1
40
1.0
>24
1.3
1.7
3.8
5
OH
OH
OH
OH
OH
OH
6
12
Table 3. Selectivity of 23 as an inhibitor of human chymase com-
pared to other serine proteases
Enzyme
IC50 (nM)
Enzyme
IC50 (nM)
Chymase
3.1
15.3
35.4
Thrombin
Elastase
Plasmin
>10,000
>10,000
>10,000
a-Chymotrypsin
CathepsinG
Trypsin
26,600
Figure 2. Schematic drawing of steric effects at 1- and 3-positions.
Figure 3. Proposed mechanism for inhibition of human chymase by compound 23.

Y. Aoyama et al. / Bioorg. Med. Chem. 9 (2001) 3065–3075
3069
Figure 4. Docking model of inhibitor 23 into human chymase: (a) surface representation; (b) schematic drawing.
at the 1-position occupies the hydrophobic prime site
region (S0) of the enzyme. The S0 region is made up of
two regions (S10 and S20). The two pairs of benzene
rings are projected into different hydrophobic regions
(S10 and S20).
fate. Flash chromatography was performed with Merck
silica gel 60 (230–400 mesh). Reactions were carried out
under a nitrogen atmosphere in anhydrous solvents
(dried over molecular sievses type 4A). Extractions were
routinely carried out twice with the given solvent and
each extract was washed with a portion of water and/or
sodium bicarbonate (NaHCO3) solution followed by a
portion of brine. The organic layers were then com-
bined, dried over sodium sulfate (Na2SO4) and con-
centrated in vacuo on a rotary evaporator.
Conclusion
We have described here the design, synthesis and phar-
macological evaluation of human chymase inhibition
based on 3-benzylazetidine-2-one. The designed com-
pound 10a possessed high potency against human chy-
mase. Modifications of compound 10a led to compound
23 which displayed high activity against human chy-
mase (IC50 3.1 nM) and sufficient stability in human
plasma (t1/2 6 h), and accordingly, is a promising candi-
date for in vivo testing.
(3S,4R)-3-Benzyl-1-(tert-butyldimethylsilyl)-2-oxoazeti-
dine-4-carboxylic acid (5). To a solution of (4R)-1-(tert-
butyldimethylsilyl)-2-oxoazetidine-4-carboxylic acid (4)12
(12.8 g, 56mmol) in tetarahydrofuran (THF) (64 mL) at
ꢀ55ꢁC was added 2.0 M lithium diisopropylamide in
heptane/THF/ethylbenzene (58.8 mL, 0.12mol) over
15 min. After 20 min, benzylbromide (14.7 mL, 0.12 mol)
was added at ꢀ55 ꢁC. The reaction mixture was kept
between ꢀ40 and ꢀ15 ꢁC for 1.5 h and then poured into
1 M NaHSO4 aqueous solution. The layers were sepa-
rated, and the aqueous layer was extracted with ethyl
acetate (EtOAc). The organic layers were combined and
extracted with six portions of 5% sodium bicarbonate
solution. The aqueous layers were combined, adjusted
to pH 2.0 and extracted with EtOAc. The organic layer
was washed with brine, dried over Na2SO4, evaporated
under reduced pressure to dryness and treated with
hexane. The precipitate was collected by filtration,
washed with hexane and dried at room temperature
under atmospheric pressure to give 17.5 g (98%) of 5.
1H NMR (CDCl3): d 0.01 (3H, s), 0.21 (3H, s), 0.78 (9H,
s), 3.03 (1H, dd, J=14.5, 5.5 Hz), 3.15 (1H, dd, J=14.5,
6.2Hz), 3.65 (1H, ddd, J=6.0, 5.7, 2.8 Hz), 3.78 (1H, d,
J=2.8 Hz), 6.36 (1H, br s), 7.20–7.40 (5H, m).
Experimental
General
IR spectra were recorded in CHCl3 or 99.5% EtOH on
1
a JASCO FT/IR-700. H NMR spectra were recorded
on a Varian GEMINI 2000 at 300 MHz with tetra-
methylsilane as internal standard (d scale) and in
CDCl3. J values are given in Hertz. Optical rotations
were measured on a Perkin-Elmer polarimeter. HR-
FAB/MS were recorded on a JEOL LMS-SX/SX 102A.
Analytical thin layer chromatography (TLC) was car-
ried out on Merck precoated TLC plates of silica gel 60
F254 and visualized with UV light or 10% H2SO4 con-
taining 5% ammonium molybdate and 0.2% ceric sul-

3070
Y. Aoyama et al. / Bioorg. Med. Chem. 9 (2001) 3065–3075
(3S,4S)-4-Acetoxy-3-benzylazetidin-2-one (6). To a solu-
tion of 5 (17.3 g, 54 mmol) in dimethylformamide
(DMF; 50 mL) and acetic acid (HOAc, 10 mL) under N2
was added lead tetraacetate (25.2 g, 54 mmol). The
reaction mixture was heated at 50–55 ꢁC for 40 min and
then cooled to 25 ꢁC. To the mixture was added 1.0 M n-
tetrabutylammonium fluoride in THF (43 mL, 43 mmol)
at 25 ꢁC. The reaction mixture was stirred at room tem-
perature for 1 h before it was poured into dilute HCl
and extracted with EtOAc. The EtOAc layer was washed
with sodium bicarbonate solution, and brine, dried over
Na2SO4, and evaporated. The residue was purified by
elution (EtOAc/toluene) through a silica gel column and
gave 6.24 g (53%) of 6 as an oil, 0.65 g (6%) of (3S,4R)-
4-acetoxy-3-benzylazetidin-2-one and 1.07 g (9%) of a
(CDCl3): d 1.53 (3H, d, J=6.9 Hz), 3.02(1H, dd,
J=14.7, 9.1 Hz), 3.22 (1H, dd, J=14.9, 6.2Hz), 3.65
(1H, ddd, J=8.9, 6.0, 1.2Hz), 5.01 (1H, quintet,
J=7.4 Hz), 5.72(1H, d,
J=1.3 Hz), 6.58 (1H, d,
J=8.0 Hz), 6.94 (2H, d, J=8.8 Hz), 7.07 (1H, s), 7.12–
7.48 (20H, m), 7.99 (2H, d, J=8.9 Hz).
Preparation of 8b–d. Using essentially the same proce-
dures as for the preparation of 8a, the following com-
pounds were prepared from 7a or 7b.
(3S,4R)-3-Benzyl-4-[4-(1,1-diphenylmethoxycarbonyl)-
phenoxy]-1-[(R)-1-(phenylethyl)aminocarbonyl]azetidin-2
-one (8b). This was derived from 7b and (R)-1-phenyl-
ethyl isocyanate; 1H NMR (CDCl3): d 1.54 (3H, d,
J=6.4 Hz), 3.19 (1H, dd, J=14.8, 8.6 Hz), 3.23 (1H, dd,
J=14.9, 6.1 Hz), 3.87 (1H, ddd, J=8.4, 7.0, 4.5 Hz),
5.04 (1H, quintet, J=7.4 Hz), 6.13 (1H, d, J=4.6 Hz),
6.85 (1H, d, J=8.3 Hz), 7.08 (1H, s), 7.12–7.48 (22H,
m), 8.06 (2H, d, J=9.0 Hz).
1
mixture of C-4 isomers. H NMR (CDCl3): d 2.08 (SH,
s), 3.02(1H, dd, J=14.8, 8.0 Hz), 3.14 (1H, dd, J=14.7,
6.0 Hz), 3.51 (1H, ddd, J=7.5, 6.2, 1.3 Hz), 5.51 (1H, d,
J=1.1 Hz), 6.42(1H, br s), 7.18–7.40 (5H, m).
(3S,4S)- and (3S,4R)-3-Benzyl-4-[4-(1,1-diphenylmethoxy-
carbonyl)phenoxy]azetidin-2-one (7a,b). To a solution of
4-hydroxybenzoic acid (27.6 g, 0.20 mol) in methylene
chloride (221 mL) and methanol (69 mL) was added a
solution of diphenyldiazomethane (58.3 g, 0.30 mol) in
methylene chloride (200 mL) in an ice bath. The reac-
tion mixture was stirred at room temperature for 16 h
and concentrated. The residue was crystallized from n-
hexane/EtOAc to give 50.0 g (82%) of diphenylmethyl
4-hydroxybenzoate. To a solution of diphenylmethyl 4-
hydroxybenzoate (2.77 g, 9.1 mmol) in acetone (11 mL)
and 1 N sodium hydroxide (8.4 mL, 8.4 mmol) was
slowly added a solution of 6 (1.54 g, 7.0 mmol) in ace-
tone (8 mL) at 5–10 ꢁC. The mixture was stirred at 10–
15 ꢁC for 3 h and then diluted with water and extracted
with two portions of EtOAc. The EtOAc layers were
each washed with brine and then combined, dried over
Na2SO4 and evaporated to give 4.22 g of oil. The residue
was purified by elution (EtOAc/toluene) through a silica
gel column and gave 2.38 g (74%) of 7a as a solid and
(3S,4S)-3-Benzyl-4-[4-(1,1-diphenylmethoxycarbonyl)-
phenoxy]-1-[(S)-1-(phenylethyl)aminocarbonyl]azetidin-2-
one (8c). This was derived from 7a and (S)-1-phenyl-
ethyl isocyanate; 1H NMR (CDCl3): d 1.53 (3H, d,
J=7.6 Hz), 3.05 (1H, dd, J=14.6, 9.0 Hz), 3.25 (1H, dd,
J=14.9, 6.2Hz), 3.63 (1H, ddd, J=7.5, 6.9, 1.3 Hz),
5.02(1H, quintet, J=7.4 Hz), 5.77 (1H, d, J=1.3 Hz),
6.79 (1H, d, J=8.0 Hz), 6.92(2H, d, J=8.9 Hz), 7.07
(1H, s), 7.12–7.48 (20H, m), 7.98 (2H, d, J=8.9 Hz).
(3S,4R)-3-Benzyl-4-[4-(1,1-diphenylmethoxycarbonyl)-
phenoxy]-1-[(S)-1-(phenylethyl)aminocarbonyl]azetidin-2-
one (8d). This was derived from 7b and (S)-1-phenyl-
ethyl isocyanate; 1H NMR (CDCl3): d 1.53 (3H, d,
J=7.0 Hz), 3.22 (1H, dd, J=14.8, 8.4 Hz), 3.25 (1H, dd,
J=14.8, 7.0 Hz), 3.84 (1H, ddd, J=8.6, 7.1, 4.7 Hz),
5.03 (1H, quintet, J=7.4 Hz), 6.09 (1H, d, J=4.7 Hz),
6.88 (1H, d, J=8.2 Hz), 7.08 (1H, s), 7.12–7.48 (22H,
m), 8.07 (2H, d, J=9.0 Hz).
1
0.56 g (17%) of 7b as a solid. 7a, H NMR (CDCl3): d
3.07 (1H, dd, J=14.8, 8.4 Hz), 3.23 (1H, dd, J=14.9,
5.7 Hz), 3.63 (1H, ddd, J=8.3, 5.2, 1.2 Hz), 5.40 (1H, d,
J=1.2Hz), 6.55 (1H, br s), 6.87 (2H, d, J=8.6 Hz), 7.08
(1H, s), 7.13–7.48 (15H, m), 8.02(2H, d, J=8.5 Hz). 7b,
1H NMR (CDCl3): d 3.20 (2H, d, J=7.6 Hz), 3.71–3,85
(1H, m), 5.75 (1H, d, J=4.2Hz), 6.70 (1H, br s), 6.86
(2H, J=8.9 Hz), 7.09 (1H, s), 7.10–7.49 (15H, m), 8.04
(2H, d, J=8.9 Hz).
(3S,4S)-3-Benzyl-4-(4-carboxy)phenoxy-1-[(R)-1-(phenyl-
ethyl)aminocarbonyl]azetidin-2-one (9a). To a solution of
8a (1.88 g, 3.1 mmol) in anisole (9.4 mL) was added
precooled trifluoroacetic acid (TFA; 2.4 mL) at 5 ꢁC.
After 1.5 h at 5 ꢁC, the reaction mixture was poured into
sodium bicarbonate solution and extracted with two
portions of EtOAc. The EtOAc layers were each washed
with brine, combined, dried over Na2SO4 and evapo-
rated to give 10 g of oil. The residue was crystallized
from hexane to give 1.23 g (90%) of 9a, [a]D +89.4ꢁ (c
0.51, CHCl3). IR nmax cmꢀ1: 1779, 1708, 1691; 1H NMR
(CDCl3): d 1.54 (3H, d, J=6.9 Hz), 1.80 (1H, br s), 3.03
(1H, dd, J=14.6, 9.5 Hz), 3.24 (1H, dd, J=14.4,
6.0 Hz), 3.66 (1H, ddd, J=9.2, 6.0, 1.2 Hz), 5.02 (1H,
quintet, J=7.4 Hz), 5.74 (1H, d, J=1.2Hz), 6.83 (1H, d,
J=8.4Hz), 6.90 (2H, d, J=9.3Hz), 7.18–7.40 (10H, m),
7.92(2H, d, J=8.7Hz); HR-MS (FAB, [M+H]+) calcd
for [C26H24O5N2+H]+ 445.1763, found 445.1768.
(3S,4S)-3-Benzyl-4-[4-(1,1-diphenylmethoxycarbonyl)-
phenoxy]-1-[(R)-1-(phenylethyl)aminocarbonyl]azetidin-2
- one (8a). To a solution of 7a (1.85 g, 4.0 mmol) in
methylene chloride (18 mL) were added (R)-1-pheny-
lethyl isocyanate (1.13 mL, 8.0 mmol), triethylamine
(1.12mL, 8.0 mmol) and 4-(dimethylamino)pyridine
(DMAP) (25 mg, catalyst). The reaction mixture was
stirred at room temperature for 16 h and then poured
into dilute HCl and extracted twice with EtOAc. The
EtOAc layers were washed with brine, combined, dried
over Na2SO4, and evaporated. The residue was purified
by elution (EtOAc/toluene) through a silica gel column
and gave 2.02 g (83%) of 8a as a solid. 1H NMR
Preparation of 9b–d. Using essentially the same proce-
dures as for the preparation of 9a, the following com-
pounds were prepared from 8b, 8c or 8d.

Y. Aoyama et al. / Bioorg. Med. Chem. 9 (2001) 3065–3075
3071
(3S,4R)-3-Benzyl-4-(4-carboxy)phenoxy-1-[(R)-1-(phenyl-
ethyl)aminocarbonyl]azetidin - 2 - one (9b). This was
derived from 8b; 1H NMR (CDCl3): d 1.55 (3H, d,
J=6.9 Hz), 2.20 (1H, br s), 3.16 (1H, dd, J=14.9,
8.9 Hz), 3.25 (1H, dd, J=14.6, 7.1 Hz), 3.89 (1H, ddd,
J=10.2, 7.2, 4.5 Hz), 5.06 (1H, quintet, J=7.2Hz), 6.17
(1H, d, J=4.5 Hz), 6.89 (1H, d, J=7.8 Hz), 7.11 (2H, d,
J=8.7 Hz), 7.16–7.45 (10H, m), 7.96 (2H, d, J=8.7 Hz).
[M+H]+) calcd for [C31H34O4N4+H]+ 527.2658,
found 527.2659.
(3S,4S)-3-Benzyl-4-(4-methylpiperazincarbonyl)phenoxy-
1-[(S)-1-(phenylethyl)aminocarbonyl]azetidin-2-one (10c).
1
This was derived from 9c; H NMR (CDCl3): d 1.53
(3H, d, J=6.6 Hz), 2.25–2.55 (4H, br), 2.32 (3H, s), 3.03
(1H, dd, J=14.9, 9.5 Hz), 3.10–3.90 (4H, br), 3.24 (1H,
dd, J=14.7, 5.7 Hz), 3.63 (1H, ddd, J=9.2, 6.2, 1.2 Hz),
5.02(1H, quintet, J=7.2Hz), 5.70 (1H, d, J=1.2Hz),
6.76 (1H, d, J=8.1 Hz), 6.87 (2H, d, J=8.4 Hz), 7.15–
7.38 (12H, m); HR-MS (FAB, [M+H]+) calcd for
[C31H34O4N4+H]+ 527.2658, found 527.2661.
(3S,4S)-3-Benzyl-4-(4-carboxy)phenoxy-1-[(S)-1-(phenyl-
ethyl)aminocarbonyl]azetidin - 2 - one (9c). This was
derived from 8c; 1H NMR (CDCl3): d 1.53 (3H, d,
J=6.9 Hz), 1.85 (1H, br s), 3.06 (1H, dd, J=14.6,
9.5 Hz), 3.27 (1H, dd, J=14.4, 6.0 Hz), 3.66 (1H, ddd,
J=9.2, 6.0, 1.2 Hz), 5.02 (1H, quintet, J=7.4 Hz), 5.76
(1H, d, J=1.2Hz), 6.83 (1H, d, J=8.4 Hz), 6.90 (2H, d,
J=9.0 Hz), 7.18–7.40 (10H, m), 7.92(2H, d, J=9.0 Hz).
(3S,4R)-3-Benzyl-4-(4-methylpiperazincarbonyl)phenoxy]
- 1 - [(S) - 1 - (phenylethyl)aminocarbonyl]azetidin - 2 - one
1
(10d). This was derived from 9d; H NMR (CDCl3): d
1.54 (3H, d, J=6.9 Hz), 2.33 (3H, s), 2.33–2.53 (4H, br),
3.23 (1H, dd, J=14.7, 8.4 Hz), 3.27 (1H, dd, J=14.9,
7.1 Hz), 3.40–3.85 (4H, br), 3.84 (1H, ddd, J=8.6, 7.4,
4.6 Hz), 5.03 (1H, quintet, J=7.4 Hz), 6.02(1H, d,
(3S,4R)-3-Benzyl-4-(4-carboxy)phenoxy]-1-[(S)-1-(phenyl-
ethyl)aminocarbonyl]azetidin - 2 - one (9d). This was
derived from 8d; 1H NMR (CDCl3): d 1.55 (3H, d,
J=7.2Hz), 1.80 (1H, br s), 3.19 (1H, dd, J=14.9,
8.6 Hz), 3.27 (1H, dd, J=14.7, 7.2Hz), 3.86 (1H, ddd,
J=8.6, 7.1, 4.5 Hz), 5.04 (1H, quintet, J=7.2Hz), 6.12
(1H, d, J=4.8 Hz), 6.91 (1H, d, J=8.4 Hz), 7.16 (2H, d,
J=9.0 Hz), 7.18–7.38 (10H, m), 7.99 (2H, d, J=9.0 Hz).
J=4.5 Hz), 6.92(1H, d,
J=8.4 Hz), 7.16 (2H, d,
J=8.7 Hz), 7.19–7.38 (12H, m); HR-MS (FAB,
[M+H]+) calcd for [C31H34O4N4+H]+ 527.2658,
found 527.2669.
Preparation of 11a,b. Using essentially the same proce-
dures as for the preparation of 10a,b, the following
compounds were prepared from l-aspartic acid.
(3S,4S)-3-Benzyl-4-(4-methylpiperazincarbonyl)phenoxy-
1-[(R)-1-(phenylethyl)aminocarbonyl]azetidin-2-one (10a).
To a solution of 9a (120 mg, 0.27 mmol) in methylene
chloride (1.2mL) were added 1-methylpiperazine
(36 mL, 0.32mmol) and 1-[3-(dimethylamino)propyl]-3-
ethylcarbodiimide hydrochloride (62mg, 32mmol) at
5 ꢁC. The reaction mixture was stirred at room tem-
perature for 5 h and then poured into sodium bicarbo-
nate solution and extracted with two portions of
EtOAc. The EtOAc layers were each washed with brine,
combined, dried over Na2SO4 and evaporated to give
143 mg of oil. The residue was crystallized from hexane
to give 130 mg (92%) of 10a as an amorphous solid, [a]D
+43.2ꢁ (c 0.77, CHCl3); IR nmax cmꢀ1: 1780, 1710,
1620; 1H NMR (CDCl3): d 1.54 (3H, d, J=6.9 Hz), 2.32
(3H, s), 2.30–2.54 (4H, br), 3.04 (1H, dd, J=14.7,
9.3 Hz), 3.25 (1H, dd, J=14.7, 6.0 Hz), 3.32–3.90 (4H,
br), 3.64 (1H, ddd, J=9.2, 6.0, 1.2 Hz), 5.03 (1H, quin-
tet, J=7.3 Hz), 5.70 (1H, d, J=1.2Hz), 6.83 (1H, d,
J=7.5 Hz), 6.88 (2H, d, J=8.4 Hz), 7.19–7.40 (12H, m);
HR-MS (FAB, [M+H]+) calcd for [C31H34O4N4+H]+
527.2658, found 527.2662.
(3R,4S)-3-Benzyl-4-(4-methylpiperazincarbonyl)phenoxy-
1 - [(R) - 1 - (phenylethyl)aminocarbonyl]azetidin - 2 - one
(11a). 1H NMR (CDCl3) d 1.54 (3H, d, J=6.9 Hz),
2.33–2.55 (4H, br), 2.34 (3H, s), 3.23 (1H, dd, J=14.7,
8.7 Hz), 3.26 (1H, dd, J=15.0, 7.5 Hz), 3.36–3.90 (4H,
br), 3.84 (1H, ddd, J=8.7, 7.2, 4.6 Hz), 5.04 (1H, quin-
tet, J=7.4 Hz), 6.02(1H, d, J=4.5 Hz), 6.92(1H, d,
J=8.4 Hz), 7.16 (2H, d, J=9.0 Hz), 7.19–7.38 (12H, m);
HR-MS (FAB, [M+H]+) calcd for [C31H34O4N4+H]+
527.2658, found 527.2660.
(3R,4R)-3-Benzyl-4-(4-methylpiperazincarbonyl)phenoxy
- 1 - [(R) - 1 - (phenylethyl)aminocarbonyl]azetidin - 2 - one
(11b). 1H NMR (CDCl3) d 1.54 (3H, d, J=6.9 Hz),
2.30–2.54 (4H, br), 2.33 (3H, s), 3.04 (1H, dd, J=14.7,
9.3 Hz), 3.25 (1H, dd, J=14.7, 6.0 Hz), 3.32–3.90 (4H,
br), 3.64 (1H, ddd, J=9.2, 6.0, 1.2 Hz), 5.03 (1H, quin-
tet, J=7.3 Hz), 5.70 (1H, d, J=1.2Hz), 6.83 (1H, d,
J=7.5 Hz), 6.88 (2H, d, J=8.4 Hz), 7.19–7.40 (12H, m);
HR-MS (FAB, [M+H]+) calcd for [C31H34O4N4+H]+
527.2658, found 527.2660.
Preparation of 10b–d. Using essentially the same proce-
dures as for the preparation of 10a, the following com-
pounds were prepared from 9b, 9c or 9d.
(3S,4R)-3-Benzyl-3-ethyl-1-(tert-butyldimethylsilyl)-2-ox-
oazetidine-4-carboxylic acid (12). To a solution of diiso-
propylamine (9.8 mL, 70 mmol) in THF (53 mL) at
ꢀ35 ꢁC was added 1.6 M n-butyllithium in hexane
(42.6 mL, 68 mmol). After 15 min, the solution was
cooled to ꢀ55 ꢁC and a solution of 4 (7.11 g, 31 mmol)
in tetarahydrofuran (THF) (31 mL) was added over
5 min. The solution was warmed to ꢀ25 ꢁC for 30 min
before ethyl iodide (6.20 mL, 78 mmol) was added. The
reaction mixture was kept between ꢀ25 and 0 ꢁC for
1.5 h and then diluted with ether and poured into a
(3S,4R)-3-Benzyl-4-(4-methylpiperazincarbonyl)phenoxy-
1 - [(R) - 1 - (phenylethyl)aminocarbonyl]azetidin - 2 - one
1
(10b). This was derived from 9b; H NMR (CDCl3): d
1.54 (3H, d, J=6.9 Hz), 2.33–2.56 (4H, br), 2.34 (3H, s),
3.21 (1H, dd, J=15.3, 8.4 Hz), 3.24 (1H, dd, J=14.7,
6.9 Hz), 3.32–3.90 (4H, br), 3.86 (1H, ddd, J=8.6, 6.9,
4.6 Hz), 5.05 (1H, quintet, J=7.4 Hz), 6.07 (1H, d,
J=4.8 Hz), 6.89 (1H, d, J=7.5 Hz), 7.14 (2H, d,
J=8.4 Hz), 7.18–7.38 (12H, m); HR-MS (FAB,

3072
Y. Aoyama et al. / Bioorg. Med. Chem. 9 (2001) 3065–3075
mixture of ice and 1 N HCl. The layers were separated,
and the aqueous layer was extracted with ether. The
ether layers were each washed with brine, combined,
dried over Na2SO4 and evaporated. The residue was
crystallized from hexane to give 7.00 g (84%) of (3S,4R)
-1-(tert-butyldimethylsilyl)-3-ethyl-2-oxoazetidine-4-car-
tions of 7% sodium bicarbonate solution. The aqueous
layer was adjusted to pH 2.0 and extracted with two
portions of EtOAc. The organic layers were each
washed with brine, combined, dried over Na2SO4 and
1
evaporated to give 17.5 (95%) g of 14 as an oil. H
NMR (CDCl3): d 0.03 (3H, s), 0.24 (3H, s), 0.83 (9H, s),
2.97 (1H, dd, J=14.2, 7.7 Hz), 3.22 (1H, dd, J=14.2,
5.3 Hz), 3.62(1H, ddd, J=7.8, 5.3, 2.4 Hz), 3.82 (3H, s),
3.83 (1H, d, J=2.7 Hz), 5.60 (1H, br s), 6.80–6.95 (2H,
m), 7.15–7.25 (2H, m).
1
boxylic acid. H NMR (CDCl3): d 0.14 (3H, s), 0.32
(3H, s), 0.97 (9H, s), 1.06 (3H, t, J=7.4 Hz), 1.62–1.99
(2H, m), 3.29 (1H, ddd, J=8.3, 5.7, 2.6 Hz), 3.79 (1H, d,
J=2.7 Hz), 6.90 (1H, br s). To a solution of diisopro-
pylamine (9.8 mL, 70 mmol) in THF (53 mL) at ꢀ35 ꢁC
was added 1.6 M n-butyllithium in hexane (42.6 mL,
68 mmol). After 15 min, the solution was cooled to
ꢀ55 ꢁC and a solution of (3S,4R)-1-(tert - butyldi-
methylsilyl)-3-ethyl-2-oxoazetidine-4-carboxylic acid
(7.00 g, 26 mmol) in THF (31 mL) was added over
5 min. The solution was warmed to ꢀ25 ꢁC for 30 min
before benzyl bromide (8.13 mL, 68 mmol) was added.
(3S,4R)-1-(tert-Butyldimethylsilyl)-3-(2-ethoxybenzyl)-2-
oxoazetidine - 4 - carboxylic acid (15). Using essentially
the same procedures as for the preparation of 14, com-
pound 15 was prepared from 2-ethoxybenzyl alcohol,
hydrobromic acid and 4. 1H NMR (CDCl3): d 0.01 (3H,
s), 0.23 (3H, s), 0.81 (9H, s), 1.42 (3H, t, J=7.1 Hz),
2.96 (1H, dd, J=14.1, 7.2Hz), 3.25 (1H, dd, J=14.3,
5.3 Hz), 3.63 (1H, ddd, J=7.7, 5.4, 3.0 Hz), 3.88 (1H, d,
J=2.7 Hz), 4.04 (2H, q, J=7.1 Hz), 6.83 (1H, d,
J=8.4 Hz), 6.87 (1H, td, J=7.5, 1.2Hz), 7.18 (1H, td,
J=7.2, 1.5 Hz), 7.19 (1H, d, J=7.2Hz).
ꢁ
The reaction mixture was kept between ꢀ2 5 and 0C
for 1.5 h and then poured into a mixture of ice and 1N
HCl and extracted with EtOAc. The organic layer was
extracted with eleven portions of 5% sodium bicarbo-
nate solution. The aqueous layer was adjusted to pH 1.0
and extracted with two portions of EtOAc. The organic
layers were each washed with brine, combined, dried
over Na2SO4 and evaporated to give 8.74 g (79%) of 12.
1H NMR (CDCl3): d ꢀ0.10 (3H, s), 0.26 (3H, s), 0.92
(9H, s), 1.29 (3H, t, J=7.5 Hz), 1.70–2.10 (2H. m), 2.96
(1H, d, J=14.1 Hz), 3.40 (1H, d, J=14.0 Hz), 4.11 (1H,
s), 6.30 (1H, br s), 7.28–7.50 (5H, m).
(3S,4R)-1-(tert-Butyldimethylsilyl)-2-oxo-3-(2-iso-pro-
poxybenzyl)azetidine - 4 - carboxylic acid (16). Using
essentially the same procedures as for the preparation of
14, compound 16 was prepared from 2-iso-propoxy-
benzyl alcohol, hydrobromic acid and 4. 1H NMR
(CDCl3): d 0.01 (3H, s), 0.23 (3H, s), 0.81 (9H, s), 1.35
(6H, d, J=6.0 Hz), 2.92 (1H, dd, J=14.0, 7.4 Hz), 3.26
(1H, dd, J=14.3, 5.4 Hz), 3.63 (1H, ddd, J=7.7, 5.4,
3.0 Hz), 3.91 (1H, d, J=2.7 Hz), 4.58 (1H, quintet,
J=6.0 Hz), 6.85 (1H, d, J=8.4 Hz), 6.86 (1H, td,
J=6.8, 1.5 Hz), 7.18 (1H, td, J=7.6, 1.5 Hz), 7.20 (1H,
d, J=7.6 Hz).
(3S,4S)-3-Benzyl-3-ethyl-4-(4-methylpiperazincarbonyl)-
phenoxy-1-[(R)-1-(phenylethyl)aminocarbonyl]azetidin-2-
one (13). Using essentially the same procedures as for
the preparation of 10a, compound 13 was prepared
1
from 12. H NMR (CDCl3): d 1.19 (3H, t, J=7.5 Hz),
1.50 (3H, d, J=6.6 Hz), 1.84 (1H, quintet, J=7.2Hz),
2.05 (1H, quintet, J=7.3 Hz), 2.28–2.53 (4H, br), 2.33
(3H, s), 2.95 (1H, d, J=14.1 Hz), 3.15 (1H, d,
J=14.1 Hz), 3.30–3.90 (4H, m), 4.94 (1H, quintet,
J=7.5 Hz), 5.62(1H, s), 6.82(1H, d, J=8.1 Hz), 7.03
(2H, d, 8.4 Hz), 7.11–7.39 (12H, m); HRMS (FAB,
[M+H]+) calcd for [C33H38O4N4+H]+ 555.2971,
found 555.2967.
Preparation of 17, 18 and 19. Using essentially the same
procedures as for the preparation of 7a, the following
compounds were prepared from 14, 15 or 16.
(3S,4S)-4-[4-(1,1-Diphenylmethoxycarbonyl)phenoxy]-3-
(2-methoxybenzyl)azetidin-2-one (17). This was derived
1
from 14; H NMR (CDCl3): d 3.00 (1H, dd, J=14.4,
8.8 Hz), 3.28 (1H, dd, J=14.4, 5.2Hz), 3.59 (1H, ddd,
J=8.8, 5.3, 1.0 Hz), 3.76 (3H, s), 5.46 (1H, d,
J=1.0 Hz), 6.52(1H, s), 6.76 (2H, d, J=8.9 Hz), 6.85
(1H, dd, J=8.4, 1.2Hz), 6.92 (1H, td, J=7.8, 1.2Hz),
7.08 (1H, s), 7.12–7.48 (12H, m), 8.01 (2H, d,
J=8.9 Hz).
(3S,4R)-1-(tert-Butyldimethylsilyl)-3-(2-methoxybenzyl)-
2-oxoazetidine-4-carboxylic acid (14). The mixture of 2-
methoxybenzyl alcohol (25.4 g, 0.18 mol) and 48%
hydrobromic acid (64.2mL, 0.57 mol) was vigorously
stirred at room temperature for 2.5 h. The layers were
separated and the organic layer was dried over CaCl2 to
give 32.6 g (90%) of 2-methoxybenzyl bromide. To a
solution of diisopropylamine (15.8 mL, 0.11 mol) in
THF (85 mL) at ꢀ35 ꢁC was added 1.54 M n-butyl-
lithium in hexane (71.4 mL, 0.11 mol). After 15 min, the
solution was cooled to ꢀ55 ꢁC and a solution of 4
(11.5 g, 0.050 mol) in tetarahydrofuran (THF) (50 mL)
was added over 15 min. The solution was warmed to
ꢀ25 ꢁC for 30 min before 2-methoxybenzyl bromide
(25.1 g, 0.12 mol) was added. The reaction mixture was
(3S,4S)-4-[4-(1,1-Diphenylmethoxycarbonyl)phenoxy]-3-
(2-ethoxybenzyl)azetidin-2-one (18). This was derived
1
from 15; H NMR (CDCl3): d 1.35 (3H, t, J=6.9 Hz),
3.01 (1H, dd, J=14.6, 8.9 Hz), 3.30 (1H, dd, J=14.4,
5.4 Hz), 3.62(1H, ddd, J=8.9, 5.3, 1.0 Hz), 4.00 (2H, q,
J=6.9 Hz), 5.50 (1H, d, J=0.9 Hz), 6.36 (1H, s), 6.74
(2H, d, J=8.7 Hz), 6.84 (1H, dd, J=8.4, 1.2Hz), 6.91
(1H, td, J=7.8, 1.2Hz), 7.08 (1H, s), 7.19–7.45 (12H,
m), 8.00 (2H, d, J=9.0 Hz).
ꢁ
kept between ꢀ2 5 and 0 C for 1.5 h and then poured
(3S,4S)-4-[4-(1,1-Diphenylmethoxycarbonyl)phenoxy]-2-
oxo - 3 - (2 - iso - propoxybenzyl)azetidin (19). This was
derived from 16; 1H NMR (CDCl3): d 1.23 (3H, d,
into a mixture of ice and 1 N HCl and extracted with
EtOAc. The organic layer was extracted with six por-

Y. Aoyama et al. / Bioorg. Med. Chem. 9 (2001) 3065–3075
3073
J=6.0 Hz), 3.01 (1H, dd, J=14.6, 8.9 Hz), 3.30 (1H, dd,
J=14.4, 5.4 Hz), 3.62(1H, ddd, J=8.9, 5.3, 1.0 Hz),
4.52(1H, quintet, J=6.0 Hz), 5.50 (1H, d, J=0.9 Hz),
6.36 (1H, s), 6.74 (2H, d, J=8.7 Hz), 6.84 (1H, dd,
J=8.4, 1.2Hz), 6.91 (1H, td, J=7.8, 1.2Hz), 7.08 (1H,
s), 7.19–7.45 (12H, m), 8.00 (2H, d, J=9.0 Hz).
m), 3.66 (3H, s), 4.77 (1H, q, J=7.5 Hz), 5.80 (1H, d,
J=1.2Hz), 6.81 (1H, d, J=8.4 Hz), 6.90 (1H, t,
J=7.5 Hz), 6.91 (2H, d, J=9.0 Hz), 7.13–7.39 (8H, m),
7.91 (2H, d, J=9.0 Hz); HR-MS (FAB, [M+H]+) calcd
for [C28H28O6N2+H]+ 489.2026, found 489.2024.
(3S,4S)-3-(2-Ethoxybenzyl)4-(4-carboxy)phenoxy-1-[(R)-
1 - (phenylpropyl)aminocarbonyl]azetidin - 2 - one (22).
Using essentially the same procedures as for the pre-
paration of 21, the following compound was prepared
from 18. White amorphous solid, [a]D +2.0ꢁ (c 0.50,
99.5% EtOH); IR nmax cmꢀ1: 1778, 1709, 1691; 1H
NMR (CDCl3): d 0.92(3H, t, J=7.2Hz), 1.28 (3H, t,
J=7.1 Hz), 1.84 (1H, qdd, J=7.4, 13.8, 3.3 Hz), 1.87
(1H, qdd, J=7.4, 13.8, 3.3 Hz), 2.00 (1H, br s), 2.98
(1H, dd, J=13.8, 9.3 Hz), 3.27 (1H, dd, J=14.0,
5.6 Hz), 3.70 (1H, ddd, J=9.0, 5.7, 1.4 Hz), 3.93 (2H, q,
J=7.1 Hz), 4.77 (1H, q, J=7.8 Hz), 5.78 (1H, d,
J=1.2Hz), 6.80 (1H, d, J=7.8 Hz), 6.88 (1H, t,
J=7.5 Hz), 6.89 (2H, d, J=9.0 Hz), 7.16 (1H, dd,
J=7.5, 1.8 Hz), 7.19–7.39 (7H, m), 7.90 (2H, d,
J=9.0 Hz); HR-MS (FAB, [M+H]+) calcd for
[C29H30O6N2+H]+ 503.2182, found 503.2188.
(3S,4S)-4-(4-Carboxy)phenoxy-3-(2-methoxybenzyl)-1-
[(R)-1-(phenylethyl)aminocarbonyl]azetidin-2-one (20).
Using essentially the same procedures as for the pre-
paration of 9a, the following compound was prepared
from 17. White amorphous solid; [a]D +102.3ꢁ (c 0.51,
CHCl3). IR nmax cmꢀ1: 1778, 1708, 1691; 1H NMR
(CDCl3): d 1.54 (3H, d, J=7.2Hz), 1.80 (1H, br s), 3.00
(1H, dd, J=14.3, 8.7 Hz), 3.24 (1H, dd, J=14.4,
5.4 Hz), 3.64 (1H, ddd, J=8.7, 5.4, 1.2Hz), 3.68 (3H, s),
5.03 (1H, quintet, J=7.4 Hz), 5.81 (1H, d, J=1.2Hz),
6.83 (1H, d, J=8.1 Hz), 6.85 (1H, t, J=8.3 Hz), 6.92
(2H, d, J=8.7 Hz), 7.17 (1H, dd, J=7.5, 1.5 Hz), 7.23–
7.40 (7H, m), 7.92(2H, d, J=9.0 Hz); HR-MS (FAB,
[M+H]+) calcd for [C27H26O6N2+H]+ 475.1869,
found 475.1870.
(3S,4S)-4-(4-Carboxy)phenoxy-3-(2-methoxybenzyl)-1-
[(R)-1-(phenylpropyl)aminocarbonyl]azetidin-2-one (21).
To a solution of (R)-phenylbutyric acid (616 mg,
3.8 mmol) in methylene chloride (6 mL) were added
triethylamine (523 mL, 3.8 mmol) and diphenylphos-
phoryl azide (808 mL, 3.7 mmol) and then the reaction
mixture was stirred at room temperature for 2.5 h. To
the reaction mixture were added a solution of 17
(740 mg, 1.5 mmol) in methylene chloride (4 mL), trie-
thylamine (523 mL, 3.8 mmol) and DMAP (10 mg, cata-
lyst). The reaction mixture was stirred at 45 ꢁC for 5 h,
diluted with EtOAc and poured into a mixture of ice
and 1 N HCl. The layers were separated and the aqu-
eous layer was reextracted with EtOAc. The organic
layers were each washed with brine, combined, dried
over Na2SO4 and evaporated. The residue was purified
by elution (toluene/EtOAc) through a silica gel column
and gave 880 mg (90%) of (3S,4S) - 4 - (4 - diphenyl-
methoxycarbonyl)phenoxy-3-(2-methoxybenzyl)-1-[(R)-1-
(phenylpropyl)aminocarbonyl]azetidin - 2- one 25. 1H
NMR (CDCl3): d 0.91 (3H, t, J=7.4 Hz), 1.81 (1H, qdd,
J=7.6, 13.8, 1.4 Hz), 1.88 (1H, qdd, J=7.6, 13.8,
1.4 Hz), 3.00 (1H, dd, J=14.1, 9.1 Hz), 3.21 (1H, dd,
J=14.3, 5.5 Hz), 3.59–3.70 (1H, m), 3.64 (3H, s), 4.76
(1H, q, J=7.1 Hz), 5.78 (1H, d, J=1.3 Hz), 6.78 (1H, d,
J=8.1 Hz), 6.87 (1H, t, J=7.8 Hz), 6.95 (2H, d,
J=9.0 Hz), 7.07 (1H, s), 7.10–7.48 (18H, m), 7.98 (2H,
d, J=9.0 Hz). To a solution of 25 (655 mg, 1.0 mmol) in
anisole (3.2mL) was added precooled TFA (0.8 mL) at
5 ꢁC. After 1.5 h at 5 ꢁC, the reaction mixture was
poured into sodium bicarbonate solution and extracted
with two portions of EtOAc. The EtOAc layers were
each washed with brine, dried over Na2SO4 and evapo-
rated. The residue was crystallized from hexane to give
415 g (85%) of 21 as a white amorphous solid, [a]D
+2.0ꢁ (c 0.50, 99.5% EtOH); IR nmax cmꢀ1: 1778, 1709,
(3S,4S)-4-(4-Carboxy)phenoxy-1-[diphenylmethyl)amino-
carbonyl]-3-(2-ethoxybenzyl)azetidin-2-one (23). To a
solution of diphenylacetic acid (182mg, 0.86 mmol) in
methylene chloride (2mL) were added triethylamine
(144 mL, 1.0 mmol) and diphenylphosphoryl azide
(222 mL, 1.0 mmol) and then the reaction mixture was
stirred at room temperature for 2.5 h. To the reaction
mixture were added
a solution of 18 (300 mg,
0.59 mmol) in methylene chloride (5 mL), triethylamine
(144 mL, 1.0 mmol) and DMAP (10 mg, catalyst). The
reaction mixture was stirred at 45 ꢁC for 5 h, diluted
with EtOAc and poured into a mixture of ice and 1 N
HCl. The layers were separated and the aqueous layer
was reextracted with EtOAc. The organic layers were
each washed with brine, combined, dried over Na2SO4
and evaporated. The residue was purified by elution
(toluene/EtOAc) through a silica gel column and gave
398 mg (94%) of (3S,4S) - 4 - (4 - diphenylmethoxy-
carbonyl)phenoxy-1-[(diphenylmethyl)aminocarbonyl]-3
-(2-ethoxybenzyl)azetidin-2-one 26. 1H NMR (CDCl3): d
1.28 (3H, t, J=6.9 Hz), 3.00 (1H, dd, J=14.3, 8.6 Hz),
3.29 (1H, dd, J=14.1, 6.1 Hz), 3.65–3.75 (1H, m), 3.94
(2H, q, J=6.8 Hz), 5.83 (1H, d, J=1.5 Hz), 6.15 (1H, d,
J=8.4 Hz), 6.81 (1H, d, J=8.0 Hz), 6.89 (1H, t,
J=7.8 Hz), 6.94 (2H, d, J=8.9 Hz), 7.07 (1H, s), 7.10–
7.48 (18H, m), 7.97 (2H, d, J=8.9 Hz). To a solution of
26 (392mg, 0.55 mmol) in anisoꢁle (0.42mL) was added
precooled TFA (0.42mL) at 5 C. After 1.5 h at 5 ꢁC,
the reaction mixture was poured into sodium bicarbo-
nate solution and extracted with two portions of
EtOAc. The EtOAc layers were each washed with brine,
dried over Na2SO4 and evaporated. The residue was
crystallized from hexane to give 242 g (80%) of 23 as a
white amorphous solid, [a]365ꢀ2.0ꢁ (c 0.51, CHCl3). IR
1
nmax cmꢀ1: 1778, 1709, 1691; H NMR (CDCl3): d 1.29
1
1691; H NMR (CDCl3): d 0.92(3H, t, J=7.5 Hz), 1.84
(3H, t, J=7.0 Hz), 2.97 (1H, dd, J=14.3, 9.3 Hz), 3.30
(1H, dd, J=13.7, 5.9 Hz), 3.66–3.76 (1H, m), 3.95 (2H,
q, J=7.0 Hz), 5.84 (1H, d, J=1.3 Hz), 6.16 (1H, d,
J=8.4 Hz), 6.79–6.95 (4H, m), 7.15–7.40 (12H, m), 7.91
(1H, qdd, J=7.7, 13.8, 3.0 Hz), 1.87 (1H, qdd, J=7.7,
13.8, 3.0 Hz), 2.00 (1H, br s), 3.00 (1H, dd, J=14.1,
9.3 Hz), 3.23 (1H, dd, J=14.1, 5.7 Hz), 3.59–3.70 (1H,

3074
Y. Aoyama et al. / Bioorg. Med. Chem. 9 (2001) 3065–3075
(2H, d, J=8.9 Hz); HRMS (FAB, [M+H]+) calcd for
[C33H30O6N2+H]+ 551.2182, found 551.2175.
period and the results were fitted to a first-order decay
curve from which the reported t1/2 value was derived.
(3S,4S)-4-(4-Carboxy)phenoxy-1-[(diphenylmethyl)ami-
nocarbonyl]-2-oxo-3-(2-iso-propoxybenzyl)azetidine (24).
Using essentially the same procedures as for the pre-
paration of 23, the following compound was prepared
from 19. White amorphous solid, [a]Dꢀ1.2ꢁ (c 0.51,
99.5% EtOH); IR nmax cmꢀ1: 1780, 1711, 1691; 1H
NMR (CDCl3): d 1.21 (3H, d, J=6.1 Hz), 1.26 (3H, d,
J=6.1 Hz), 2.97 (1H, dd, J=14.3, 9.3 Hz), 3.30 (1H, dd,
J=13.7, 5.9 Hz), 3.73 (1H, ddd, J=9.0, 5.7, 1.4 Hz),
4.52(1H, quintet, J=6.3 Hz), 5.84 (1H, d, J=1.2Hz),
6.15 (1H, d, J=8.4 Hz), 6.80–6.95 (4H, m), 7.15–7.40
(12H, m), 7.89 (2H, d, J=8.9 Hz); HR-MS (FAB,
[M+H]+) calcd for [C34H32O6N2+H]+ 565.2339,
found 565.2344.
Molecular modeling
Molecular modeling was performed using the SYBYL
6.6 software package. The three-dimensional model of
compound 23 was generated by CONCORD and opti-
mized by the Tripos force field. The model of inhibitor
23 was manually docked into the active site of human
chymase on the basis of the structure–activity relation-
ships of 3-benzylazetidine-2-one derivatives against
human chymase and the docking study for the binding
of 3,3-diethyl-2-azetidinone to human leukocyte elasta-
se.11b The resulting hemiketal intermediate (human
chymase-23 complex) was energy-minimized using the
Tripos force field.
Enzyme assay
Acknowledgements
The human chymase assay was performed as follows.
First, human chymase was purified according to the
method of Takai.19 The purified chymase was pre-
incubated with test compounds dissolved in DMSO at
37 ꢁC for 30 min in 0.1 M Tris–HCl (pH 8.0) containing
1.8 M NaCl, then the chymase reaction was started by
adding succinyl-Ala-Ala-Pro-Phe-p-nitroanilides (Sigma
Chemical Co.). The change of absorbꢁance was measured
at 405 nm after 2h incubation at 37 C. The IC50 value
was calculated from the inhibition of p-nitroaniline for-
mation at each concentration of the test compound.
We wish to thank Professor. M. Miyazaki, Osaka
Medical College, for the kind gift of human chymase.
References and Notes
1. (a) Caughey, G. H. J. Respir. Crit. Care Med. 1994, 150,
S138. (b) Wang, Z.; Walter, M.; Selwood, T.; Rubin, H.;
Schechter, N. M. Biol. Chem. 1998, 379, 167.
2. Ihara, M.; Urata, H.; Kinoshita, A.; Suzumiya, J.; Sasa-
guri, M.; Kikuchi, M.; Ideishi, M.; Arakawa, K. Hypertension
1999, 33, 1399.
The inhibitory effects of compound 23 on the enzymatic
activities of seven serine proteases were evaluated using
the purified enzymes and chromogenic substrates. The
enzymes and substrates used here were as follows: N-suc-
cinyl-Ala-Ala-Pro-Phe-pNA (Bachem) for bovine pan-
creatic a-chymotrypsin (Sigma) and human cathepsin G
(Wako); Chromozyme TH (Boehringer Mannheim) for
human thrombin (Sigma); N-succinyl-Ala-Ala-Phe-Arg-
pNA (Bachem) for bovine pancreatic trypsin (Sigma);
N-succinyl-Ala-Ala-Val-pNA (Bachem) for human
neutrophil elastase (Athens Research and Technology,
Inc.); Chromozym PL (Boehringer Mannheim) for
human plasmin (Sigma). The assay buffer used here was
as follows: 50 mM Tris–HCl (pH=8.0) containing
2mM CaCl 2 for a-chymotrypsin, trypsin and elastase;
50 mM Tris–HCl (pH=7.5) containing 2mM CaCl 2
for cathepsin G; 50 mM Tris–HCl (pH=7.5) contain-
ing 50 mM NaCl for plasmin; 0.1 M Tris–HCl
(pH=8.0) containing 10 mM CaCl2 and 0.1 M NaCl for
thrombin.
3. He, S.; Gaca, M. D. A.; McEuen, A. R.; Walls, A. F. J.
Pharmacol. Exp. Ther. 1999, 291, 517.
4. Mizutani, H.; Schechter, N.; Lazarus, G.; Black, R. A.;
Kupper, T. S. J. Exp. Med. 1991, 174, 821.
5. Kofford, M. W.; Schwartz, L. B.; Schechter, N. M.; Yager,
D. R.; Diegelmann, R. F.; Graham, M. F. J. Biol. Chem. 1997,
272, 7127.
6. Fang, K. C.; Raymond, W. W.; Blount, J. L.; Caughey,
G. H. J. Biol. Chem. 1997, 272, 25628.
7. (a) Shiota, M.; Fukamizu, A.; Takai, S.; Okunishi, H.;
Murakami, K.; Miyazaki, M. J. Hypertens. 1997, 15, 431. (b)
Hamada, H.; Terai, M.; Kimura, H.; Hirano, K.; Oana, S.;
Niimi, H. Am. J. Respir. Crit. Care. Med. 1999, 160, 1303. (c)
Takai, S.; Yuda, A.; Jin, D.; Nishimoto, M.; Sakaguchi, M.;
Sasaki, S.; Miyazaki, M. FEBS Lett. 2000, 467, 141.
8. (a) Hayashi, Y.; Iijima, K.; Katada, J.; Kiso, Y. Bioorg.
Med. Chem. Lett. 2000, 10, 199. (b) Groutas, W. C.; Schech-
ter, N. M.; He, S.; Yu, H.; Huang, P.; Tu, J. Bioorg. Med.
Chem. Lett. 1999, 9, 2199. (c) Iijima, K.; Katada, J.; Hayashi,
Y. Bioorg. Med. Chem. Lett. 1999, 9, 413. (d) Fukami, H.;
Okunishi, H.; Miyazaki, M. Curr. Pharm. Des. 1998, 4, 439
and references therein.
9. (a) Aoyama, Y.; Uenaka, M.; Konoike, T.; Iso, Y.; Nishi-
tani, Y.; Kanda, A.; Naya, N.; Nakajima, M. Bioorg. Med.
Chem. Lett. 2000, 10, 2397 and 2403. (b) Aoyama, Y.;
Uenaka, M.; Konoike, T.; Hayakawa-Kajiwara, Y.; Naya, N.;
Nakajima, M. Bioorg. Med. Chem. Lett. 2001, 11, 1691.
10. (a) Doherty, J. B.; Ashe, B. M.; Barker, P. L.; Blacklock,
T. J.; Butcher, J. W.; Chandler, G. O.; Dahlgren, M. E.;
Davies, P.; Dorn, C. P.; Finke, P. E., Jr.; Firestone, R. A.;
Hagmann, W. K.; Halgren, T.; Knight, W. B.; Maycock, A. L.;
Navia, M. A.; O’Grady, L.; Pisano, J. M.; Shah, S. K.;
Thompson, K. R.; Weston, H.; Zimmerman, M. J. Med.
Stability assay
The stability assay in human plasma was performed as
follows. Test compounds dissolved in DMSO were
added at 20 mg/mL to human plasma and incubated at
37 ꢁC. Aliquots were removed at intervals of 30 min,
quenched with 3-fold volume of AcOH–MeOH (1/1) or
AcOH–MeCN (1/1), and analyzed by reverse-phase
HPLC after removal of the proteins. The disappearance
of test compounds was then monitored for the half-life

Y. Aoyama et al. / Bioorg. Med. Chem. 9 (2001) 3065–3075
3075
Chem. 1990, 33, 2513. (b) Finke, P. E.; Ashe, B. M.; Knight,
W. B.; Maycock, A. L.; Navia, M. A.; Shah, S. K.; Thompson,
K. R.; Underwood, D. J.; Weston, H.; Zimmerman, M.;
Doherty, J. B. J. Med. Chem. 1990, 33, 2522. (c) Shah, S. K.;
Brause, K. A.; Chandler, G. O.; Finke, P. E.; Ashe, B. M.;
Weston, H.; Knight, W. B.; Maycock, A. L.; Doherty, J. B. J.
Med. Chem. 1990, 33, 2529.
12. (a) Baldwin, J. E.; Adlington, R. M.; Gollins, D. W.;
Schofield, C. J. Tetrahedron 1990, 13, 4733. (b) Ohtani, M.;
Watanabe, F.; Narisada, M. J. Antibiot. 1985, 38, 610.
13. Shioiri, T.; Ninomiya, K.; Yamada, S. J. Am. Chem. Soc.
1972, 94, 6203.
14. We examined the activity of L-694,458 against human
chymase and found that it was not potent.
11. (a) Shah, S. K.; Dorn, C. P.; Finke, P. E., Jr.; Hale, J. J.;
Hagmann, W. K.; Brause, K. A.; Chandler, G. O.; Kissinger,
A. L.; Ashe, B. M.; Weston, H.; Knight, W. B.; Maycock,
A. L.; Dellea, P. S.; Fletcher, D. S.; Hand, K. M.; Mumford,
R. A.; Underwood, D. J.; Doherty, J. B. J. Med. Chem. 1992,
35, 3745. (b) Finke, P. E.; Shah, S. K.; Fletcher, D. S.; Ashe,
B. M.; Brause, K. A.; Chandler, G. O.; Dellea, P. S.; Hand,
K. M.; Maycock, A. L.; Osinga, D. G.; Underwood, D. J.;
Weston, H.; Davies, P.; Doherty, J. B. J. Med. Chem. 1995,
38, 2449. (c) Cvetovich, R. J.; Chartrain, M.; Hartner, F. W.;
Roberge, C., Jr.; Amato, J. S.; Grabowski, E. J. J. J. Org.
Chem. 1996, 61, 6575.
15. Kinoshita, A.; Urata, H.; Bumpus, F. M.; Husain, A. J.
Biol. Chem. 1991, 266, 19192.
16. Stein, R. L.; Strimpler, A. M.; Mori, H.; Powers, J. C.
Biochemistry 1987, 26, 1301.
17. Pereira, P. J. B.; Wang, Z. M.; Rubin, H.; Huber, R.;
Bode, W.; Schechter, N. M.; Strobl, S. J. Mol. Biol. 1999, 286,
163 (PDB code, 1pjp).
18. See ref 11b for the docking study for the binding of 3,3-
diethyl-2-azetidinones to human leukocyte elastase performed
at Merck.
19. Takai, S.; Siota, N.; Sakaguchi, M.; Muraguchi, H.; Mat-
sumura, E.; Miyazaki, M. Clinica Chimica Acta 1997, 265, 13.