1-Alkoxycarbonyl-3-bromoazetidin-2-ones as potential elastase inhibitors

Article abstract of DOI:10.1002/(SICI)1099-0690(199906)1999:6<1441::AID-EJOC1441>3.0.CO;2-K

Three 1-alkoxycarbonyl-3-bromoazetidin-2-ones have been prepared by reaction of (3S)-3-(tert-butoxycarbonyl)amino-azetidin-2-one with benzyl, trichloroethyl, and trifluoroethyl chloroformates followed by tBoc deprotection, diazotation of the exocyclic amino function and its substitution with potassium bromide. The 3-bromoazetidin-2-ones were obtained as racemic mixtures. Their hydroxide-catalyzed hydrolysis exclusively affords ring-opening products. Porcine pancreatic elastase (PPE) catalyzes the same reaction stereospecifically. Model building suggests that it is the (R) isomer that is enzymatically hydrolysed. The PPE-catalyzed hydrolysis is characterised by low k(cat) and K(m) values. Accordingly, these compounds behave as transient inhibitors of the enzyme.

Full text of DOI:10.1002/(SICI)1099-0690(199906)1999:6<1441::AID-EJOC1441>3.0.CO;2-K

FULL PAPER
1-Alkoxycarbonyl-3-bromoazetidin-2-ones as Potential Elastase Inhibitors
[a,b]
[a]
[c]
[d]
´
´
Gregory Tjoens, Roland Touillaux, Josette Lamotte-Brasseur ,
Cecile Beauve,
Jacqueline Marchand-Brynaert,[b] and Jacques Fastrez*[a]
Keywords: Azetidinones / Enzyme inhibitors / Kinetics / Molecular modelling
Three 1-alkoxycarbonyl-3-bromoazetidin-2-ones have been
hydrolysis exclusively affords ring-opening products. Porcine
prepared by reaction of (3S)-3-(tert-butoxycarbonyl)amino- pancreatic elastase (PPE) catalyzes the same reaction
azetidin-2-one with benzyl, trichloroethyl, and trifluoroethyl
chloroformates followed by tBoc deprotection, diazotation of
the exocyclic amino function and its substitution with
stereospecifically. Model building suggests that it is the (R)
isomer that is enzymatically hydrolysed. The PPE-catalyzed
hydrolysis is characterised by low kcat and Km values.
potassium bromide. The 3-bromoazetidin-2-ones were Accordingly, these compounds behave as transient inhibitors
obtained as racemic mixtures. Their hydroxide-catalyzed
of the enzyme.
reverse approach, i.e. to have a leaving group on N-1 and
an electron-withdrawing group at the C-4 position, was not
successfully developed.[4]
Introduction
β-Lactam derivatives have been intensively investigated
both to understand their chemical properties and because
of their interest as inhibitors of enzymes. Analysis of their
chemical reactivity has clarified their relationship with am-
ides and shown that release of strain on ring opening does
not lead, as initially expected, to a large increase in reactiv-
ity.[1] Bicyclic β-lactams derived from cephem or penam
antibiotics and several β-lactam derivatives have been
shown to efficiently inhibit the DD-peptidases involved in
the crosslinking of the bacterial walls and sometimes also
the β-lactamases used as a defense mechanism by bacteria.
Synthetic β-lactam derivatives have also been developed to
inhibit human leukocyte elastase, a serine protease whose
uncontrolled degradative action on connective tissue has
been implicated in several disease states (for reviews see
refs.[2Ϫ4]).
Several N-activated monocyclic β-lactams act on elastase
through a suicide inhibition mechanism:[5] The enzyme-cat-
alyzed ring opening leads to the departure of substituents
introduced in their structures and allows the unmasking of
a reactive electrophilic function into the enzymatic cavity.[6]
One family of inhibitors involves N-aryl-3-haloazetidin-2-
ones featuring a latent quinoniminium methide func-
tion.[7Ϫ13] Another family is based on N-acylazetidin-2-ones
bearing a good leaving group in the C-4 position (O-aryl);
usually, the endocyclic nitrogen atom is part of a lipophilic
urea function.[14Ϫ18] Processing of such β-lactams by the
target enzyme creates a Schiff base that is susceptible to
quenching by a nucleophilic residue of the active site. The
In this work, we investigate the hydroxide- and elastase-
catalyzed hydrolysis of monocyclic β-lactams bearing an al-
koxycarbonyl function on N-1, i.e. a substituent that could
significantly increase the β-lactam carbonyl reactivity
towards nucleophilic attack[19] while providing a potential
leaving group. We address the following questions: Which
of the β-lactam and urethane functions will be preferen-
tially hydrolyzed? If ring opening occurs, will it lead to leav-
ing-group expulsion? In the elastase active site, ring opening
by nucleophilic attack of Ser-195 followed by expulsion of
the leaving group could lead to a suicide inhibition by un-
masking an electrophilic cumulene (e.g. an isocyanate func-
tion) susceptible to accept a nucleophile (His-57) from the
enzyme (Scheme 1). Isoxazoline inhibitors of elastase fused
to N-(alkoxysulfonyl)succinimide motifs have been postu-
lated to follow such a mechanism in which the reactive iso-
cyanate function results from a rapid Lossen rearrange-
ment.[20]
To answer these questions, we selected the (3R)-1-alkoxy-
carbonyl-3-bromoazetidin-2-one family (Z ϭ O; L ϭ OR;
X ϭ Br). The choice of the C-3 substituent was based on
the work of WakselmanЈs group,[13] who demonstrated that
HLE was more efficiently inactivated with 3,3-dibromo-
and (3R)-3-bromo-3-fluoroazetidinones rather than with
3,3-difluoro- and (3S)-3-bromo-3-fluoroazetidinones. Three
potential leaving groups were considered; the benzyloxy-,
trichloroethyloxy-, and trifluoroethyloxy groups.
[a]
`
Laboratoire de Biochimie Physique et des Biopolymeres,
[b]
[c]
Results
`
Laboratoire de Chimie Organique de Synthese,
Laboratoire de Chimie Physique et de Cristallographie,
Universite Catholique de Louvain, Departement de Chimie,
´
´
Syntheses
ˆ
Batiment Lavoisier,
1, place Louis Pasteur, B-1348 Louvain-la-Neuve, Belgium
Fax: (internat.) ϩ 32-10/472820
The (3S)-3-(tert-butoxycarbonyl)aminoazetidin-2-one 1
was prepared in five steps from -serine.[21] Reaction of 1
with lithium hexamethyldisilylazide (LiHMDS)[22Ϫ23] in
THF at Ϫ78°C followed by addition of chloroformates 2
E-mail: Fastrez@bioc.ucl.ac.be
[d]
´
´
Laboratoire dЈEnzymologie and Centre dЈIngenierie des
´
´
`
Proteines, Universite de Liege, Institut de Chimie,
`
B6, B-4000 Sart-Tilman (Liege 1), Belgium
Eur. J. Org. Chem. 1999, 1441Ϫ1447
WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
1434Ϫ193X/99/0606Ϫ1441 $ 17.50ϩ.50/0
1441

J. Fastrez et al.
FULL PAPER
Scheme 1. Potential suicide mechanism of elastase inhibition
gave the N-alkoxycarbonyl derivatives 3,[24Ϫ25] which could their 13C-NMR spectra by signals at δ ϭ 40 (C-3) and δ ϭ
1
be contaminated by the double acylation products 4. Com- 49 (C-4). In the H-NMR spectra, the three cyclic protons
pounds 3a (R ϭ CH2Ph), 3b (R ϭ CH2CCl3) and 3c (R ϭ gave a typical ABX pattern at δ ϭ 3.7Ϫ3.8 (4-H), 4.1Ϫ4.2
CH2CF3) were purified by chromatography on silica gel and (4Ј-H) and 4.8Ϫ4.9 (3-H).
characterized using standard spectroscopic techniques.
Typical spectroscopic values are the IR carbonyl stretching
near 1820 cmϪ1 for the activated azetidinone function, and
the H-NMR chemical shift near δ ϭ 4.8 for 3-H. Treat-
Chemical Hydrolysis of 6a؊c
1
ment with trifluoroacetic acid furnished the amines 5,
We have studied the chemical reactivity of compounds
which were directly treated with sodium nitrite and potas- 6a؊c (Scheme 2) towards hydrolysis. The products of
sium bromide in aqueous sulfuric acid at 6°C.[26] As shown chemical hydrolysis were analyzed by NMR spectroscopy
by GC analysis on a chiral capillary column, the 3-bromo- and mass spectrometry. Compounds 6a؊c were dissolved
azetidin-2-ones 6a؊c were obtained as racemic mixtures; in D2O (phosphate buffer pH ϭ 7.5)/5% [D6]DMSO and
racemization occurs presumably during the last step of the 1H-NMR spectra were recorded as a function of time; after
synthesis.
the disappearance of the starting material, the only prod-
In the cases of 6-aminopenam[27] and 7-aminocephem ucts detected were 7a؊c resulting from β-lactam ring open-
antibiotics,[28] similar substitutions involving the corre- ing (Scheme 3); no cleavage of the urethane was observed.
sponding 6- and 7-diazonium intermediates are stereoselec- The hydrolysis products are characterized by an AB pattern
tive giving the (6R)-6- and (7R)-7-halo derivatives. This sel- at δ ϭ 3.5Ϫ3.7 (3-H) and δ ϭ 3.4Ϫ3.6 (3Ј-H), with a large
ectivity results from the steric control exercised by the (5R)- geminal coupling constant J ϭ 13Ϫ15 Hz. The methylene
5- and (6R)-6-sulfur substituents. However, epimerization protons (3-H and 3Ј-H) split the resonance of the methine
of the (6R)-6-bromopenicillanic acid in aqueous solution proton (2-H) to give a doublet of doublet at δ ϭ 4.1Ϫ4.3. In
has been demonstrated to occur through an enolization the mass spectra of the crude mixture, peaks are observed at
process involving 6-H.[29] Accordingly, in the absence of in- a molecular mass corresponding to the addition of water to
duction due to a bulky C-3 substituent, the formation of the β-lactam ring.
racemic 1-alkoxycarbonyl-3-bromoazetidin-2-ones 6 is not
The rates of hydrolysis have been monitored by UV spec-
surprising. The compounds 6a؊c were characterized in trophotometry at 232 nm and the absorbance was found to
Scheme 2. Preparation of 1-alkoxycarbonyl-3-bromoazetidin-2-ones; reactions conditions: (i) LiHMDS, THF, Ϫ78°C, 30 min, then 2,
Ϫ78°C to 20°C, 1 h: 3a: R ϭ CH2Ph, yield 89%, 3b: R ϭ CH2CCl3, yield 69%, 3c: R ϭ CH2CF3, yield 83%; (ii) CF3COOH, 20°C,
30 min; (iii) KBr, 2.5  H2SO4/EtOH, NaNO2, 6Ϫ8°C, 3 h: 6a: R ϭ CH2Ph, yield 80%, 6b: R ϭ CH2CCl3, yield 23%, 6c: R ϭ CH2CF3,
yield 56%
1442
Eur. J. Org. Chem. 1999, 1441Ϫ1447

1-Alkoxycarbonyl-3-bromoazetidin-2-ones as Potential Elastase Inhibitors
FULL PAPER
Scheme 3. Hydrolysis product of 1-alkoxycarbonyl-3-bromoazetidin-2-ones
decrease by 40%. First-order rate constants (kobs) have been min. Plots of V/VI, ratios of initial rates of hydrolysis in the
measured at different pH values; buffer catalysis is negli- absence and in the presence of inhibitor versus [I] gave the
gible; plots of kobs versus hydroxide ion concentration are following inhibition constants: 6a: KI ϭ (1.5 ± 0.1) ϫ 10Ϫ6
linear, the following second-order rate constants are ob- ; 6b: KI ϭ (3.0 ± 0.36) ϫ 10Ϫ6 ; 6c: KI ϭ (2.4 ± 0.4) ϫ
Ϫ
Ϫ
tained: 6a: kOH ϭ (5.2 ± 1.1) ϫ 102 Ϫ1sϪ1; 6c: kOH
ϭ
10Ϫ5 .
(1.0 ± 0.2) ϫ 103 Ϫ1sϪ1. The rate constant increases when
the pKa of the alcohol constituent of the urethane function
decreases: 6a: pKa ϭ 15.2; 6c: pKa ϭ 12.37.[30] These com-
pounds appear to be less stable to chemical hydrolysis (ca.
80 times) than the 3-ethyl-4-p-carbophenoxy-substituted
azetidinone described previously as a leukocyte elastase in-
hibitor.[31]
The observation of a transient inhibition suggested that
PPE catalyzes the hydrolysis of compounds 6a؊c, both the
urethane and the β-lactam functions being potentially sensi-
tive to enzymatic hydrolysis. 1H-NMR analysis of the prod-
ucts of enzymatic hydrolysis of a 2·10Ϫ2  solution of 6a in
the presence of PPE (6·10Ϫ6 ) in D2O/5% [D6]DMSO
showed that only ring opening occurs. When the course of
the elastase-catalyzed hydrolysis of 6c was monitored by
UV spectrophotometry, a biphasic curve was observed (Fig-
ure 1): The rate of the fast reaction is sensitive to the en-
Elastase Inhibition and Enzymatic Hydrolysis
Compounds 6a؊c were first evaluated for their potential zyme concentration while the slow phase reflects the hy-
inhibitory effect on porcine pancreatic elastase (PPE). The droxide-catalyzed reaction. Elastase catalyzes the hydrolysis
rates of elastase-catalyzed hydrolysis of N-succinyl-Ala-Ala- of only one enantiomer of the racemic mixture of 3-bromo-
Ala-p-nitroanilide were measured ([E] ϭ 2.10Ϫ7 , [S] ϭ azetidin-2-one derivatives. This interpretation was con-
10Ϫ4 , pH ϭ 7.5) in the presence of various concentrations firmed by chiral GC analysis of the mixture during hydroly-
of 6a؊c. At [I] ϭ 10Ϫ4 , transient inhibitions were ob- sis (Figure 1, insert). The following rate constants were ex-
served (percentage of inhibition: 6a: 98.5%; 6b: 97%; 6c: tracted for the enzymatic reaction by curve fitting of the
82%); at the relatively high enzyme concentration used, data: kcat ϭ 6.2·10Ϫ3 sϪ1 and Km ϭ 1.1·10Ϫ5 . As for the
however, complete enzymatic activity was restored after 20 fitting, the concentrations of the individual enantiomers are
Figure 1. Elastase-catalyzed hydrolysis of 6c; the absorbance of a 2·10Ϫ4  solution of 6c was monitored in the presence of 1.9·10Ϫ6
PPE at 232 nm at pH ϭ 7.67; inset: GC analysis on a Chrompack chiral-dex capillary column of the mixture during hydrolysis

Eur. J. Org. Chem. 1999, 1441Ϫ1447
1443

J. Fastrez et al.
FULL PAPER
Figure 2. Stereodiagram representation of the docking of 6a in the active site of PPE
used, the calculated Km is the concentration of half satu- drolysis, the enzymatic activity was fully restored within ex-
ration for the enantiomer whose hydrolysis is catalyzed by perimental error. This suggests that no cumulene function
the enzyme. When the inhibition of anilide hydrolysis was is generated within the active site during hydrolysis. A strat-
measured, the KI values were calculated on the basis of the egy to favor cumulene formation would be to incorporate
sum of the (R) and (S) isomer concentrations. The ratio less basic alkoxides in the urethane function but, unfortu-
between KI/Km determined for 6c, i.e. ca. 2, suggests that nately, the 3-bromoazetidin-2-one derivatives become then
only one enantiomer is recognized by the enzyme. This hy- too ustable (results not shown).
pothesis is confirmed by the fact that no inhibition of p-
When oriented in the PPE active site as described above,
nitroanilide hydrolysis is observed when this substrate is ad- and as observed in the crystallographic structure of the
ded to solutions of 6a؊c in which the enzymatic hydrolysis complex formed between the PPE and a β-lactam inhibitor
of the recognized isomers was complete.
complex,[35] the (R) derivative 6a fits easily in the enzyme
Docking of (R)- and (S)-6a؊c in the PPE active site cavity (Figure 2). The carbonyl oxygen atom of the β-lac-
(3EST)[32] followed by optimization by molecular mech- tam ring is in hydrogen-bonding interactions with the back-
anics using AMBER[33] (Figure 2) showed that the (R) de- bone NH groups of Ser-195 and Gly-193, which create the
rivatives fit nicely with their bromine atom in the S-1 bind- oxyanion hole. The esterified carboxylic group of the inhibi-
˚
ing pocket and the β-lactam carbonyl group is suitably tor is located about 4 A from the His-57 imidazole ring.
placed for nucleophilic attack by Ser-195 on the α face of The presence of the ester group is important as it prevents
the ring. The β-lactam oxygen atom is located in the oxy an electrostatic interaction that would otherwise occur be-
anion hole, the urethane carbonyl group is oriented towards tween the His side-chain and a free carboxylate group, thus
His-57.
perturbating the catalytic mechanism. The phenyl group of
6a protrudes outside the active site. With the (R) derivative,
the Br substituent is positioned in the S-1 cavity of the ser-
ine protease bordered by the side chains of Val-216 and
Thr-226, which leave room for favorable interactions only
with a small side chain like a Br atom. By contrast, with
the (S) derivative the Br atom is oriented towards Cys-191
and the disulfide bridge it forms with Cys-220, this induces
less favourable enzyme-ligand interactions. Replacement of
the phenyl CH2 group of 6a by the more bulky CF3 group
in 6c could lead to a weak steric hindrance with the
C42ϪC58 SϪS bridge or modify the electronic environment
in the vicinity of the oxy anion hole (Gly-193 NH), but
these effects would nevertheless be small.
Discussion
Both chemical and enzymatic hydrolyses afford β-lactam
ring-opening products without alkoxide expulsion. In the
pH range investigated, the chemical hydrolysis is a hydrox-
ide-catalyzed process with a negatively charged transition
state. These conditions are particularly favorable for expul-
sion of an alkoxide leaving group before protonation of the
nitrogen atom. As the hydrolysis product contains an intact
urethane unit, the alkoxide expulsion appears to be too
slow to generate an isocyanate function.
The 3-bromoazetidin-2-one derivatives described in this
work block the active site of PPE with apparent inhibition
constants in the micromolar range. The expected suicide in-
hibition was not observed, instead PPE was shown to cata-
lyze the hydrolysis of these compounds. As this hydrolysis
presumably occurs by an acyl-enzyme mechanism, the in-
hibitory effect is likely due to the fact that they act as slowly
deacylating substrates,[34] characterized by low kcat and Km
values. Suicide inhibition frequently arises from a minor
side pathway; its efficiency is then measured from the ratio
Experimental Section
Porcine pancreatic elastase (type 1) and N-succinyl--alanyl--ala-
nyl--alanyl-p-nitroanilide were purchased from Sigma Chemical
Co. The other reagents were purchased from Acros chimica, Ald-
rich, and Fluka. Tetrahydrofuran (THF) was dried with sodium/
benzophenone, then distilled. Dichloromethane and hexane were
dried with phosphorous pentoxide and distilled. Ethyl acetate was
distilled. Ϫ Column chromatography was carried out with silica gel
between suicide and turnover events. At the end of the hy- 60 (70Ϫ230 mesh ASTM) supplied by Merck. Analytical thin-layer
1444 Eur. J. Org. Chem. 1999, 1441Ϫ1447

1-Alkoxycarbonyl-3-bromoazetidin-2-ones as Potential Elastase Inhibitors
FULL PAPER
chromatography was performed on silica gel 60 plates F254
(Merck, 0.2 mm thick), on which the compounds were detected
with UV light, I2 and ninhydrin. Ϫ The optical rotations (± 0.1°)
1-Trifluoroethyloxycarbonyl Compound 3c: This compound was ob-
tained from 1 g (5.37 mmol) of 1 by treatment with 5.4 mL
(5.4 mmol) of 1  solution of LiHMDS and 1.3 g (8 mmol, 1.5
were determined with a PerkinϪElmer 241 MC polarimeter. Ϫ The equiv.) of 2,2,2-trifluoroethyl chloroformate (2c)[36] in 15 mL of
IR spectra were recorded with a PerkinϪElmer 1710 instrument
THF. Yield: 1.394 g (83%); CH2Cl2/EtOAc (80:20, v/v); RF ϭ 0.63;
m.p. 138.2Ϫ138.4°C; [α]D ϭ ϩ6.5 (c ϭ 1 g/100 mL, CH2Cl2). Ϫ
(Fourier Transform Infra Red Spectrometer), only the most signifi-
1
cant absorption bands being reported. Ϫ The H- and 13C-NMR IR (KBr): ν˜ ϭ 1818, 1747, 1735, 1522, 1172, 762 cmϪ1. Ϫ 1H NMR
spectra were obtained with a Bruker AM-500 spectrometer; chemi-
cal shifts are reported in ppm (δ) downfield from internal TMS. Ϫ
(500 MHz, [D6]DMSO): δ ϭ 1.37 (s, 9 H), 3.57 (dd, J ϭ 6.4 Hz
and 3.95 Hz, 1 H), 3 82 (dd, J ϭ 6.4 Hz and 6.7 Hz, 1 H), 4.75
The UV spectra were recorded with a Varian Cary 210 apparatus. (ddd, J ϭ 6.7 Hz, 3.95 Hz and 8.22 Hz, 1 H), 4.88 (qAB, J ϭ
Ϫ The mass spectra were obtained with a Finnigan MAT TSQ-70
instrument. The high resolution mass spectra were obtained at the
University of Mons, Belgium (Prof. R. Flammang). Ϫ The mi-
croanalyses were performed at the University College of London,
U.K. (Dr. A. Stones). Ϫ Melting points were determined with an
Electrothermal microscope and are uncorrected. Ϫ The gas chro-
matography (GC) analyses were recorded using a Varian Aerograph
1400 (detector FID, nitrogen) equipped with a chrompack chiral-
dex (25 m ϫ 0.25 mm) capillary column and connected to a Hitachi
D-2000 integrator.
8.85 Hz, 2 H), 7.66 (d, J ϭ 8.22 Hz, 1 H). Ϫ 13C NMR (125 MHz,
CDCl3): δ ϭ 28.08, 46.78, 57.77, 61.59, 81.28, 122.33 (CF3), 147.31,
154.43, 164.04. Ϫ MS (FABϩ); m/z: 313 [M ϩ 1]. Ϫ C11H15F3N2O5
(312): calcd. C 42.31, H 4.84, N 9.01; found C 42.34, H 4.61, N
8.85.
(3S)-1-Benzyloxycarbonyl-3-[(benzyloxycarbonyl)(tert-butyloxy-
carbonyl)amino]azetidin-2-one (4a): In the chromatographic purifi-
cation of 3a, the first fraction contained 4a: yield < 10%; RF
ϭ
0.94; m.p. 101.1°C. Ϫ IR (KBr): ν˜ ϭ 3040, 2977, 1817, 1729 cmϪ1
.
1
Ϫ H NMR (500 MHz, CDCl3): δ ϭ 1.41 (s, 9 H), 3.72 (dd, J ϭ
6.6 Hz and 4.04 Hz, 1 H), 3.85 (dd, J ϭ 6.6 Hz and 6.6 Hz, 1 H),
5.23 (s, 2 H), 5.24 (s, 2 H), 5.59 (dd, J ϭ 6.6 Hz and 4.04 Hz, 1 H),
7.3Ϫ7.45 (m, 10 H). Ϫ 13C NMR (125 MHz, CDCl3): δ ϭ 27.55,
45.14, 59.38, 68.06, 69.50, 85.16, 128.27, 128.53, 134.36, 134.84,
149.05, 149.95, 152.56, 162.93. Ϫ MS (FABϪ); m/z: 453 [M Ϫ 1]
(C24H26N2O7).
(3S)-1-Alkoxycarbonyl-3-(tert-butyloxycarbonyl)aminoazetidin-2-
ones (3): To a solution of lithium bis(trimethylsilyl)amide (1  solu-
tion in THF, 1 equiv.) in dry THF (2.5 mL/mmol), cooled at
Ϫ78°C under argon, was added (3S)-3-(tert-butyloxycarbonyl)ami-
noazetidin-2-one (1) (1 equiv.), dissolved in THF (1 mL/mmol).
The mixture was stirred for 30 min at Ϫ78°C, then chloroformate
2 (1 equiv.) was added with a syringe through a rubber stopper.
Stirring was continued for 1 h at Ϫ78°C and a further 45 min at
room temperature. After addition of water, the mixture was ex-
tracted with CH2Cl2. The organic layer was washed with brine,
dried with Na2SO4, and concentrated under vacuum. The crude
product was purified by column chromatography on silica gel with
CH2Cl2/EtOAc as eluent. If necessary, the product could be pre-
cipitated from hexane.
(3S)-1-Alkoxycarbonyl-3-aminoazetidin-2-ones 5: Boc deprotection
of 3a؊c was performed by dissolution in trifluoroacetic acid (2
mL/100 mg of 3). After 30 min at 20°C, the solvent was evaporated
under vacuum and the crude amine (as the trifluoroacetate salt)
was precipitated from ether. Yield: 90Ϫ100%.
1-Alkoxycarbonyl-3-bromoazetidin-2-ones 6: To 2.5
 aqueous
H2SO4 (1 mL), cooled at 5°C, were successively added, with stir-
ring, 3-aminoazetidin-2-one (5) (0.15 mmol, 1 equiv.), KBr (5
equiv.), and ethanol (1 mL/0.3 mmol of 5). A solution of NaNO2
(1.5 equiv.) in water (1 mL/0.5 mmol) was added dropwise within
1 h, and the mixture was stirred at 6Ϫ8°C for a further 3Ϫ4 h. The
mixture was extracted with CHCl3 (several times) and the organic
layers were washed with cold brine, dried with Na2SO4 and concen-
trated under vacuum. The product was purified by column chroma-
tography on silica gel with CH2Cl2 as eluent.
1-Benzyloxycarbonyl Compound 3a: This compound was obtained
from 450 mg (2.4 mmol) of 1 by treatment with 2.4 mL (2.4 mmol)
of 1  LiHMDS solution and 0.337 mL (2.4 mmol) of benzyl
chloroformate (2a), in 7 mL of THF. Yield: 687 mg (89%); CH2Cl2/
EtOAc (60:40, v/v); RF ϭ 0.73; m.p. 148.1°C; [α]D ϭ Ϫ1.9 (c ϭ
1 g/100 mL, CH2Cl2). Ϫ IR (KBr): ν˜ ϭ 3366, 2977, 1816, 1713
cmϪ1. Ϫ 1H NMR (500 MHz, [D6]acetone): δ ϭ 1.42 (s, 9 H), 3.71
(dd, J ϭ 6.4 Hz and 3.73 Hz, 1 H), 3.91 (dd, J ϭ 6.4 Hz and 5.8 Hz,
1 H), 4.85 (ddd, J ϭ 7.9 Hz, 3.73 Hz and 5.8 Hz, 1 H), 5.25 (s, 2
H), 6.88 (d, J ϭ 7.9 Hz, 1 H), 7.25Ϫ7.5 (m, 5 H). Ϫ 13C NMR
(125 MHz, [D6]acetone): δ ϭ 28.38, 46.77, 58.06, 68.05, 80.18,
1-Benzyloxycarbonyl Compound 6a: This compound was obtained
from 164 mg (0.49 mmol) of 5a, 300 mg (2.5 mmol) of KBr, and
54 mg (0.68 mmol) of NaNO2. Yield: 112 mg (80%); RF ϭ 0.36;
m.p. 46.5°C. Ϫ IR (KBr): ν˜ ϭ 1816, 1733, 1122, 954, 764 cmϪ1. Ϫ
1H NMR (500 MHz, CDCl3): δ ϭ 3.69 (dd, J ϭ 7.82 Hz and
2.94 Hz, 1 H), 4.11 (dd, J ϭ 7.82 Hz and 5.86 Hz, 1 H), 4.75 (dd,
J ϭ 5.86 Hz and 2.94 Hz, 1 H), 5.21 (s, 2 H), 7.35 (sharp m, 5 H).
128.92, 129.07, 129.3, 136.74, 149.86, 155.75, 165.8.
Ϫ
C16H20N2O5·H2O (338): calcd. C 56.80, H 5.92, N 8.28; found C
56.94, H 5.96, N 7.96. Ϫ HRMS (FABϩ); m/z: 321.14504 [M ϩ 1]
(calcd. for C16H21N2O5: 321.14283).
1-Trichloroethyloxycarbonyl Compound 3b: This compound was ob-
tained from 500 mg (2.69 mmol) of 1 by treatment with 2.75 mL
(2.7 mmol) of 1  LiHMDS solution and 0.4 mL (2.7 mmol) of
2,2,2-trichloroethyl chloroformate (2b) in 8 mL of THF. Yield:
Ϫ
13C NMR (125 MHz, CDCl3): δ ϭ 40.34, 48.86, 68.57, 128.37,
128.58, 128.66, 134.44, 148.56, 160.25. Ϫ MS (EI); m/z: 285 and
283 [M]. HRMS (EI); m/z: 282.98416 [M] (calcd. for
Ϫ
C11H10BrNO3: 282.98440).
670 mg (69%); CH2Cl2/EtOAc (95:5, v/v); RF
ϭ 0.84; m.p.
133.2Ϫ133.4°C; [α]D ϭ ϩ4.7 (c ϭ 1 g/100 mL, CH2Cl2). Ϫ IR
1-Trichloroethyloxycarbonyl Compound 6b: This compound was ob-
1
(KBr): ν˜ ϭ 1822, 1711, 1330, 1155, 1022, 719 cmϪ1. Ϫ H NMR tained from 345 mg (0.92 mmol) of 5b, 568 mg (4.77 mmol) of KBr
(500 MHz, CDCl3): δ ϭ 1.44 (s, 9 H), 3.86 (dd, J ϭ 6.7 Hz and and 100 mg (1.43 mmol) of NaNO2. Yield: 70 mg (23%); RF ϭ 0.9;
4.5 Hz, 1 H), 3.98 (dd, J ϭ 6.7 Hz and 6.7 Hz, 1 H), 4.76 (ddd,
J ϭ 6.7 Hz, 4.5 Hz and 6.4 Hz, 1 H), 4.82 (s, 2 H), 5.16 (d, J ϭ
m.p. 70.3Ϫ70.6°C. Ϫ IR (KBr): ν˜ ϭ 1802, 1741, 1339, 1127, 814,
773 cmϪ1. Ϫ 1H NMR (500 MHz, CDCl3): δ ϭ 3.82 (dd, J ϭ
6.4 Hz, 1 H). Ϫ 13C NMR (125 MHz, CDCl3): δ ϭ 28.09, 46.98, 7.96 Hz and 3.06 Hz, 1 H), 4.25 (dd, J ϭ 7.96 Hz and 6.01 Hz, 1
57.80, 74.98, 81.19, 94.18, 147.25, 154.52, 164.07. Ϫ MS (FABϪ); H), 4.81 (s, 2 H), 4.88 (dd, J ϭ 6.01 Hz and 3.06 Hz, 1 H). Ϫ 13C
m/z: 359, 361, 363, and 365 [M Ϫ 1]. Ϫ C11H15Cl3O5N2 (361.6):
calcd. C 36.53, H 4.18, N 7.74; found C 36.24, H 3.90, N 7.39.
NMR (125 MHz, CDCl3): δ ϭ 40.55, 49.19, 75.08, 93.85 (CCl3),
146.75, 160.12. Ϫ MS (FABϩ); m/z: 332, 330, 328, 326, and 324 [M
Eur. J. Org. Chem. 1999, 1441Ϫ1447
1445

J. Fastrez et al.
FULL PAPER
[6]
[7]
[8]
R. A. Firestone, P. L. Barker, J. M. Pisano, B. M. Ashe, M. E.
Dahlgren, Tetrahedron 1990, 46, 2255Ϫ2262.
J.-L. Maillard, C. Favreau, M. Reboud-Ravaux, R. Kobaiter, R.
Joyeau, M. Wakselman, Eur. J. Cell. Biology 1990, 52, 213Ϫ218.
M. Wakselman, R. Joyeau, R. Kobaiter, N. Boggetto, I. Vergely,
J. Maillard, V. Okochi, J.-J. Montagne, M. Reboud-Ravaux,
FEBS Lett. 1991, 282, 377Ϫ381.
N. Boggetto, A.-C. Vilain, J.-J. Montagne, M. Reboud-Ravaux,
J.-P. Mazaleyrat, J. Xie, M. Wakselman, Bull. Soc. Chim. Fr.
1994, 131, 152Ϫ164.
M. Wakselman, J. Xie, J.-P. Mazaleyrat, N. Bogetto, A.-C. Vi-
lain, J.-J. Montagne, M. Reboud-Ravaux, J. Med. Chem. 1993,
36, 1539Ϫ1547.
ϩ 1]. Ϫ C6H5BrCl3NO3 (325.5): calcd. C 22.15, H 1.55, N 4.30;
found C 22.50, H 1.51, N 4.28.
1-Trifluoroethoxycarbonyl Compound 6c: This compound was ob-
tained from 98 mg (0.3 mmol) of 5c, 186 mg (1.5 mmol) of KBr
and 33.5 mg (0.45 mmol) of NaNO2. Yield: 49 mg (55.9%); RF
0.9; oil. Ϫ IR (film): ν˜ ϭ 1830, 1751, 1350, 1126, 839, 759 cmϪ1
1H NMR (500 MHz, CDCl3): δ ϭ 3.78 (dd, J ϭ 7.97 Hz and
ϭ
[9]
.
Ϫ
3.05 Hz, 1 H), 4.21 (dd, J ϭ 7.97 Hz and 5.97 Hz, 1 H), 4.56 (sharp
q, AB, J ϭ 8.02 Hz, 2 H), 4.85 (dd, J ϭ 5.97 Hz and 3.05 Hz, 1
H). Ϫ 13C NMR (125 MHz, CDCl3): δ ϭ 40.50, 49.16, 61.89,
122.16 (CF3), 145.87, 159.96. Ϫ MS (FABϩ); m/z: 276 and 278 [M
ϩ 1]. Ϫ C6H5BrF3NO3 (276): calcd. C 26.11, H 1.82, N 5.07; found
C 26.19, H 2.15, N 4.85.
[10]
[11]
I. Vergely, N. Bogetto, V. Okochi, S. Golpayegani, M. Reboud-
Ravaux, R. Kobaiter, R. Joyeau, M. Wakselman, Eur. J. Med.
Chem. 1995, 30, 199Ϫ208.
[12]
[13]
I. Vergely, P. Laugaa, M. Reboud-Ravaux, J. Mol. Graphics
1996, 14, 158Ϫ167.
Chemical Hydrolysis: For product analyses, 2·10Ϫ2  solutions of
6a؊c in [D6]DMSO were added to a phosphate buffer in D2O
(pH ϭ 7.5; final concentration of 10Ϫ3 ); 1H-NMR spectra
(500 MHz) were recorded every 30 min (the hydrolysis of 6a was
completed within 1.5 h).
C. Doucet, I. Vergely, M. Reboud-Ravaux, J. Guilhem, R. Ko-
baiter, R. Joyeau, M. Wakselman, Tetrahedron: Asymmetry
1997, 8, 739Ϫ751.
[14]
[15]
[16]
S. K. Shah, C. P. Dorn, P. E. Finke, J. J. Hale, W. K. Hagmann,
K. A. Brause, G. O. Chandler, A. L. Kissinger, B. M. Ashe, H.
Weston, W. B. Knight, A. L. Maycock, P. S. Dellea, D. S.
Fletcher, K. M. Hand, R. A. Mumford, D. J. Underwood, J. B.
Doherty, J. Med. Chem. 1992, 35, 3745Ϫ3754.
3-[(Benzyloxycarbonyl)amino]-2-bromopropionic Acid (7a): δ ϭ 3.38
(dd, J ϭ 7.3 Hz and 14.6 Hz, 1 H), 3.47 (dd, J ϭ 6.1 Hz and
14.6 Hz, 1 H), 4.12 (dd, J ϭ 6.1 Hz and 7.3 Hz, 1 H).
W. K. Hagmann, A. L. Kissinger, S. K. Shah, P. E. Finke, C. P.
Dorn, K. A. Brause, B. M. Ashe, H. Weston, A. L. Maycock,
W. B. Knight, P. S. Dellea, D. S. Fletcher, K. M. Hand, D.
Osinga, P. Davies, J. B. Doherty, J. Med. Chem. 1993, 36,
771Ϫ777.
2-Bromo-3-[(trichloroethoxycarbonyl)amino]propionic Acid (7b): δ ϭ
3.59 (dd, J ϭ 6.4 Hz and 13.7 Hz, 1 H), 3.69 (dd, J ϭ 6.4 Hz and
13.7 Hz, 1 H), 4.30 (t, J ϭ 6.4 Hz, 1 H).
R. Chabin, B. G. Green, P. Gale, A. L. Maycock, H. Weston,
C. P. Dorn, P. E. Finke, W. K. Hagmann, J. J. Hale, M. Mac-
Coss, S. K. Shah, D. Underwood, J. B. Doherty, W. B. Knight,
Biochemistry 1993, 32, 8970Ϫ8980.
2-Bromo-3-[(trifluoroethoxycarbonyl)amino]propionic Acid (7c): δ ϭ
3.58 (dd, J ϭ 6.4 Hz and 13.9 Hz, 1 H), 3.68 (dd, J ϭ 6.4 Hz and
13.9 Hz, 1 H), 4.30 (t, J ϭ 6.4 Hz, 1 H).
[17] [17a]
B.G. Green, R. Chabin, S. Mills, D. J. Underwood, S. K.
Shah, D. Kuo, P. Gale, A. L. Maycock, J. Liesch, C. S. Burgey,
J. B. Doherty, C. P. Dorn, P. E. Finke, W. K. Hagmann, J. J.
For kinetic analyses, spectra of 1·10Ϫ3  solutions of 6a؊c in 0.025
 phosphate or borate buffers (pH ϭ 7.2Ϫ8.75) containing 5%
DMSO were recorded between 220 and 400 nm. The disappearance
of the 234-nm band associated with the β-lactam was used to ob-
tain first-order rate constants.
Hale, M. MacCoss, W. M. Westler, W. B. Knight, Biochemistry
[17b]
1995, 34, 14331Ϫ14343. Ϫ
D. J. Underwood, B. G. Green,
R. Chabin, S. Mills, J. B. Doherty, P. E. Finke, M. MacCoss, S.
K. Shah, C. S. Burgey, T. A. Dickinson, P. R. Griffin, T. E.
Lee, K. M. Swiderek, T. Covey, W. M. Westler, W. B. Knight,
Biochemistry 1995, 34, 14344Ϫ14355.
[18]
P. E. Finke, S. K. Shah, D. S. Fletcher, B. M. Ashe, K. A.
Brause, G. O. Chandler, P. S. Dellea, K. M. Hand, A. L. May-
cock, D. G. Osinga, D. J. Underwood, H. Weston, P. Davies, J.
B. Doherty, J. Med. Chem. 1995, 38, 2449Ϫ2462.
C. Palomo, J. M. Aizpuruna, C. Cuevas, A. Mielgo, R. Galarza,
Tetrahedron Lett. 1995, 36, 9027Ϫ9030.
Assay of Elastase: Solutions of 6a؊c in N-methylpyrrolidone were
diluted twenty-fold into a solution of elastase in 100 m  buffer
at pH ϭ 7.5. Samples (3 µL) from these incubation mixtures were
mixed with 300 µL of 0.3 m N-succinyl--alanyl--alanyl--alanyl-
[19]
[20]
p-nitroanilide (Km ϭ 2.4 mM
[37]). The appearance of p-nitroaniline
W. C. Groutas, R. Venkataraman, L. S. Chong, J. E. Yoder, J.
B. Epp, M. A. Stanga, E.-H. Kim, Bioorg. Med. Chem. 1995,
3, 125Ϫ128.
was observed by recording with a Cary 210 the absorbance change
at 410 nm (∆ε ϭ 8480 Ϫ1cmϪ1) for 10 min.
[21]
[22]
[23]
[24]
D. M. Floyd, A. W. Fritz, J. Pluscec, E. R. Weaver, C. M. Cima-
rusti, J. Org. Chem. 1982, 47, 5160Ϫ5167.
T. Kobayashi, Y. Iwano, K. Hirai, Chem. Pharm. Bull. 1978,
26, 1761.
Acknowledgments
M. Ihara, A. Nakayama, K. Fukumoto, Tetrahedron 1982, 38,
2489Ϫ2504.
`
The authors thank Michele Bouchet for technical assistance. C. B.
W. K. Hagmann, S. K. Shah, C. P. Dorn, L. A. OЈGrady, L. J.
Hale, P. E. Finke, K. R. Thompson, K. A. Brause, B. M. Ashe,
W. Hazel, M. E. Dahlgren, A. L. Maycock, P. S. Dellea, K. M.
Hand, D. G. Osinga, R. J. Bonney, P. Davies, D. S. Fletcher, J.
B. Doherty, Bioorg. Med. Chem. Lett. 1991, 1, 545Ϫ550.
I. Ojima, O. Duclos, G. Dorman, B. Simonot, G. D. Prestwich,
J. Med. Chem. 1995, 38, 3891Ϫ3894.
`
acknowledges a fellowship form the Fonds pour la Formation a la
Recherche dans lЈIndustrie et lЈAgriculture (FRIA, Belgium). This
work was supported in part by the Fonds National de la Recherche
Scientifique (FNRS, Belgium) and by the Belgian Program of In-
teruniversity Poles of Attraction (PAI N°P4/03).
[25]
[26]
R. G. Micetich, S. N. Maiti, P. Spevak, M. Tanaka, T. Yama-
zaki, K. Ogawa, Synthesis 1986, 292Ϫ296.
[27]
[28]
I. McMillan, R. J. Stoodley, J. Chem. Soc. C 1968, 2533Ϫ2537.
J. B. Doherty, B. M. Ashe, P. L. Barker, T. J. Blacklock, J. W.
Butcher, G. O. Chandler, M. E. Dahlgren, P. Davies, C. P. Dorn,
P. E. Finke, R. A. Firestone, W. K. Hagmann, T. Halgren, W.
B. Knight, A. L. Maycock, M. A. Navia, L. OЈGrady, J. M.
Pisano, S. K. Shah, K. R. Thompson, H. Weston, M. Zimmer-
man, J. Med. Chem. 1990, 33, 2513Ϫ2521.
[1]
M. I. Page, Adv. Phys. Org. Chem. 1987, 23, 165Ϫ270.
[2]
R. F. Pratt, in The chemistry of ß-lactams (Ed.: M. I. Page),
Blackie Academic & Professional, London, 1992, p. 229.
[3]
P. D. Edwards, P. R. Bernstein, Med. Res. Rev. 1994, 14,
127Ϫ194.
[4]
O. A. Mascaretti, C. E. Boschetti, G. O. Danelon, E. G. Mata,
[29]
[30]
M. J. Loosemore, R. F. Pratt, J. Org. Chem. 1978, 43,
3611Ϫ3613.
Handbook of Organic Chemistry (Ed.: J. A. Dean), McGraw Hill
Book Company, New York, 1987.
O. A. Roveri, Curr. Med. Chem. 1995, 1, 441Ϫ470.
[5]
M. A. Ator, P. R. Ortiz de Montellano, in The Enzymes (Ed.:
P. D. Boyer), Academic Press, New York, 1990, vol. XIX, pp.
214Ϫ282.
1446
Eur. J. Org. Chem. 1999, 1441Ϫ1447

1-Alkoxycarbonyl-3-bromoazetidin-2-ones as Potential Elastase Inhibitors
FULL PAPER
[31]
[35]
W. K. Hagmann, K. R. Thompson, S. K. Shah, P. E. Finke, B.
M. Ashe, H. Weston, A. L. Maycock, J. B. Doherty, Bioorg.
Med. Chem. Lett. 1992, 2, 681Ϫ684.
E. F. Meyer, G. M. Cole, R. Radhakrishnan, O. Epp, Acta Crys-
tallogr., Sect. B.: Struct. Sci. 1988, B44, 26Ϫ38.
D. A. Pearlman, D. A. Case, J. W. Caldwell, W. S. Ross, T. E.
Cheatham III, D. E. Ferguson, G. L. Seibel, U. C. Singh, P. K.
Weiner, P. A. Kollman, AMBER 4.1, University of California,
San Francisco, 1995.
M. A. Navia, J. P. Springer, T.-Y. Lin, H. R. Williams, R. A.
Firestone, J. M. Pisano, J. B. Doherty, P. E. Finkle, K. Hoogs-
teen, Nature 1987, 327, 79Ϫ82.
[32]
[33]
[36]
[37]
I. Abe, N. Fujimoto, T. Nishiyama, K. Terada, T. Nakahara, J.
Chromatogr. A 1996, 722, 221Ϫ227.
J. Bieth, B. Spiess, C. G. Wermuth, Biochem. Med. 1974, 11,
350Ϫ357.
Received November 20, 1998
[O98536]
[34]
R. Baggio, Y. Q. Shi, Y. Q. Wu, R. H. Abeles, Biochemistry
1996, 35, 3351Ϫ3353.
Eur. J. Org. Chem. 1999, 1441Ϫ1447
1447