4-Alkylidene-azetidin-2-ones: Novel inhibitors of leukocyte elastase and gelatinase

Article abstract of DOI:10.1016/j.bmc.2003.09.035

In addition to their antibiotic potency, β-lactams have recently been investigated as inhibitors of serine proteinase such as leukocyte elastase (LE), released by inflammatory cells. We describe the synthesis of a series of 4-alkylidene-β-lactams, and investigate how substitutions on C-3, C-4, and N-1 of the β-lactam ring affect the activity of human LE and gelatinases MMP-2 and MMP-9. LE activity was measured using a chromogenic substrate, while gelatin-zymography assay was used to evaluate gelatinase activity. We demonstrate that C-4 unsaturation on the β-lactam ring determines the degree of biological activity, with a selectivity over LE by 3-[1-(tert-butyldimethylsilyloxy)-ethyl] derivatives (lowest IC₅₀ was 4 μM), and over gelatinase MMP-2 by C-3-unsubstituted 4-[1-ethoxycarbonyl]-ethylidene-β-lactams (lowest IC₅₀ was 60 μM). (3S)-3-[(1R)-1-hydroxyethyl]-4-(1-ethoxycarbonyl)-ethylidene-azetidin-2- one inhibits gelatinase MMP-9. The compounds tested showed no cytotoxicity against NIH-3T3 murine fibroblasts. This is the first example of beta-lactams inhibiting metallo-proteinases instrumental in cancer invasion and angiogenesis. These molecules are good candidates for prototype drugs showing selective antibiotic, anti-inflammatory, and anti-invasion properties.

Full text of DOI:10.1016/j.bmc.2003.09.035

Bioorganic & Medicinal Chemistry11 (2003) 5391–5399
4-Alkylidene-azetidin-2-ones: Novel Inhibitors of Leukocyte
Elastase and Gelatinase
Gianfranco Cainelli,a Paola Galletti,a Spiridione Garbisa,b,*
Daria Giacomini,a,* Luigi Sartorb and Arianna Quintavallaa
aDepartment of Chemistry ‘G. Ciamician’, University of Bologna, Via Selmi 2, I-40126 Bologna, Italy
bDepartment of Experimental Biomedical Sciences, Medical School of Padova, Viale G. Colombo 3, 35121 Padova, Italy
Received 22 July2003; accepted 19 September 2003
Abstract—In addition to their antibiotic potency, b-lactams have recentlybeen investigated as inhibitors of serine proteinase such as
leukocyte elastase (LE), released by inflammatory cells. We describe the synthesis of a series of 4-alkylidene-b-lactams, and investigate
how substitutions on C-3, C-4, and N-1 of the b-lactam ring affect the activityof human LE and gelatinases MMP-2 and MMP-9. LE
activity was measured using a chromogenic substrate, while gelatin-zymography assay was used to evaluate gelatinase activity. We
demonstrate that C-4 unsaturation on the b-lactam ring determines the degree of biological activity, with a selectivity over LE by 3-
[1-(tert-butyldimethylsilyloxy)-ethyl] derivatives (lowest IC50 was 4 mM), and over gelatinase MMP-2 byC-3-unsubstituted 4-[1-
ethoxycarbonyl]-ethylidene-b-lactams (lowest IC50 was 60 mM). (3S)-3-[(1R)-1-hydroxyethyl]-4-(1-ethoxycarbonyl)-ethylidene-aze-
tidin-2-one inhibits gelatinase MMP-9. The compounds tested showed no cytotoxicity against NIH-3T3 murine fibroblasts. This is
the first example of beta-lactams inhibiting metallo-proteinases instrumental in cancer invasion and angiogenesis. These molecules
are good candidates for prototype drugs showing selective antibiotic, anti-inflammatory, and anti-invasion properties.
# 2003 Elsevier Ltd. All rights reserved.
Introduction
developed and widelyemployed as antimicrobial drugs,
theyhave been recentlyused as inhibitors of some ser-
ine enzymes produced by viruses, fungi and mammals.
In fact, studies have demonstrated the biological activ-
ityof b-lactams against human LE, cytomegalovirus
protein, prostate specific antigen, thrombin, herpes
virus, co-enzyme A independent transacylase, g-amino-
butyric acid (GABA) aminotransferase, and cytosolic
phospholipase A2.10 As regards monocyclic b-lactam
The clinical demand for new, more specific and potent
inhibitors of proteases is growing, especiallyfor
enzymes—such as human leukocyte elastase (LE) and
mammalian gelatinases—whose activities have proven
to be instrumental in a number of severe acute and
chronic pathologies, including inflammation and cancer
invasion.1,2 While a number of synthetic inhibitors of
gelatinases have been abandoned because of their side
effects,3 more recentlysome vegetable secondarymeta-
derivatives, some structural requirements have been
11
recognized as crucial for the inhibitoryactivity,
but
bolites—commonlypresent in human diet—showed good
4ꢀ8
there has been little investigation to date of the possi-
bilityof activating their azetidinone ring through
unsaturated systems directly linked to the b-lactam
ring.12ꢀ15
inhibitoryactivity.
A comparative investigation has
highlighted some chemical moieties crucial for the inhi-
bitorypotential of these molecules toward either LE or
gelatinases;9 this knowledge maybe useful in designing
new and more potent side-effect-free inhibitors.
In a continuation of our research on the synthesis of
16,17
biologicallyactive compounds
and in particular
A similar comparative approach should prove useful for
other classes of compounds, such as b-lactams. Initially
b-lactams,18ꢀ22 we describe here the synthesis of a series
of 4-alkylidene-b-lactams exhibiting activityagainst
LE and gelatinases MMP-2 and MMP-9 at micro-
molar concentration. We investigated the effect of sub-
stitution on C-3, C-4, and N-1 positions of the b-lactam
ring on inhibitoryactivity. In particular we studied 3-[1-
*Corresponding authors. Fax: +39-051-209-9456 (D. Giacomini); fax:
+39-049-827-6089 (S. Garbisa). E-mail: giacomin@ciam.unibo.it (D.
Giacomini); garbisa@unipd.it (S. Garbisa)
0968-0896/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2003.09.035

5392
G. Cainelli et al. / Bioorg. Med. Chem. 11 (2003) 5391–5399
(tert-butyldimethylsilyloxy)-ethyl]-, 3-chloro-, 3-bromo-
and 3-H-4-alkylidene-azetidin-2-ones obtained from
diazoketones and diazoesters. For substitution on the b-
lactam nitrogen atom, together with N-H derivatives,
we synthesized N-acyl and N-alkyl-azetidin-2-ones.
Scheme 1 shows all b-lactams tested. To our knowledge,
no b-lactams with anti-gelatinase activityhave been
previouslydescribed.
isomers31 of the corresponding 4-alkylidene-b-lactams,
depending on the diazo-compound and the Lewis acid.
Critical to the success of the reaction was a stoichio-
metric amount of TiCl4 or AlCl3, and an excess of the
diazo-compound associated with a requirement for tri-
methylsilyl protection of the b-lactam nitrogen atom.
The diastereomeric products were easilyseparated by
column chromatography. Treatment of 1 and 8 with
HCl in acetonitrile produced deprotected derivatives 2
and 9 (Scheme 3).
Results and Discussion
The N-alkylated compounds 3, 10 were easilyobtained
bytreatment of 1 and 8 with benzyl bromide and solid
K2CO3 in acetone (Scheme 4).
The Lewis acid-catalyzed reaction of 4-acetoxy-azetidi-
nones with a-diazo carbonyls represents the key step in
the synthesis of 4 - alkylidene - azetidinones.23,24 Thus
compounds 1, 8, 11, 12, 13, 14, 15, and 18 were pre-
pared starting from commerciallyavailable (3 R,4R)-4-
acetoxy-3-[(1R)-1-(tert-butyldimethylsilyloxy)-ethyl]-
azetidin-2-one and 4-acetoxy-azetidin-2-one, while
compounds 16 and 17 were prepared starting from
(3R,4R)-4-acetoxy-3-bromo- and (3R,4R)-4-acetoxy-3-
chloro-azetidin-2-one,25ꢀ30 respectively( Scheme 2).
The synthetic pathway to obtain N-acylated-4-alkyli-
dene derivatives presented some problems and required
The reaction of 4-acetoxy-azetidinones with a-diazo-
carbonyls proceeds smoothly yielding E and/or Z
Scheme 3. Reagents: (i) HCl 1 M, CH3CN.
Scheme 1.
Scheme 2.

G. Cainelli et al. / Bioorg. Med. Chem. 11 (2003) 5391–5399
5393
careful analysis of reagents and reaction conditions. This
difficultyis due to higher reactivityof the b-lactam ring in
N-acylated compounds towards nucleophilic reagents.
Treatment of 1 with ethyl malonyl chloride in benzene at
reflux gave the corresponding acylated compound 4
(Scheme 5). Anyattempt to synthesize other N-acylated
analogues bythis procedure failed. We thus obtained
derivatives 5, 6 and 7 starting from E-isomer 1 and benzyl
chloroformate, benzyl isocyanate and acetic anhydride,
under lightlybasic conditions ( Scheme 5). Application of
this synthetic procedure to the Z-isomer 8 yielded only by-
products derived from a b-lactam ring expansion.32
Scheme 4. Reagents: (i) BnBr, K2CO3, acetone.
Even C-4-saturated analogues have been prepared:
compound 19 was prepared from (3R,4R)-4-acetoxy-3-
[(1R)-1-(tert-butyldimethylsilyloxy)-ethyl]-azetidin-2-one
using metallic Zn and ethyl 2-bromo-acetate. Its sub-
sequent acylation with ethyl malonyl chloride gave 20
(Scheme 6).33,34
All compounds were purified bysilica gel chromatography
and new derivatives fullycharacterized (see Experimental).
The inhibition exerted on human HLE and gelatinase
MMP-2 and MMP-9 bythe various molecules was
assayed as described in the Experimental; two examples
are given in Figure 1. The majorityof the tested com-
pounds showed inhibitoryactivityand the results are
shown in Table 1, which also gives the IC50 values.
Scheme 5. Reagents: (i) ethyl malonyl chloride, benzene reflux;
(ii) K2CO3, acetone and: CbzCl for 5, benzyl isocyanate for 6, and
Ac2O for 7.
The type of inhibition and the Ki over HLE was also
determined for compound 4, which showed the lowest
IC50: the inhibition exerted is dose-dependent and non-
competitive, as determined bydouble-reciprocal plotting
of the results obtained at different b-lactam concentra-
tions; the plots share a common ꢀ1/Michaelis constant
(Km) on the abscissa, and the calculated Ki is 10 mM.
For compounds 1 and 12 at concentrations ꢁ100 mM,
the absorbance of these molecules was interferring with
the colorimetric and/or zymographic assays; in addi-
tion, the inhibition exerted on HLE did not maintain a
constant slope throughout the measurements, and, for
this reason, in these cases the IC50 is not given.
Scheme 6. Reagents: (i) Zn, THF, BrCH2COOEt, reflux; (ii) ethyl
malonyl chloride, benzene reflux.
Figure 1. Inhibition of LE and metallo-proteinase-2 (MMP-2) by b-lactams 4 and 18. In the first case (a) a chromogenic substrate (0.4 mM) was
incubated at 37 ꢂC with HLE (5 mU) in the presence of increasing concentration of compound 4 (the most efficacious against LE), and the absor-
bance monitored at 405 nm (values at 60 min are plotted). Each point represents the mean of three digestions. In the second case (b) MMP-2-con-
taining medium was electrophoresed in gelatin-SDS-PAGE, and the zymography developed overnight in the presence of increasing amount of
compound 18 (the most efficacious against MMP-2). The MMP-2 digestion bands (inserts) were quantitated bydensitometry. Example of duplicate
experiment: double-reciprocal plot (c) demonstrating noncompetitive inhibition of LE (5 mU) byincreasing concentration of compound 4 (S=sub-
strate). The points represent the mean values of triplicate samples, with SD<10%.

5394
G. Cainelli et al. / Bioorg. Med. Chem. 11 (2003) 5391–5399
Table 1. LE, MMP-2 and MMP-9 inhibition of 4-alkylidene-b-lactams
Compd LE
MMP-2
MMP-9
Inhibition % at 100 mM IC50 (mM) Inhibition % at 200 mM IC50 (mM) Inhibition % at 200 mM IC50 (mM)
1
66
58
Nd
53
42
56
41
51
0
150
—
28
25
52
32
37
0
—
—
8
9
2
30
—
—
20
Nd
20
85
150
—
9
22
—
11
12
13
86
150
—
—
68
—
100
0
—
0
—
14
18
15
16
17
2
—
—
—
—
—
37
65
0
160
60
—
51
20
0
—
—
—
—
—
28
0
0
0
—
0
32
0
—
0
(continued on next page)

G. Cainelli et al. / Bioorg. Med. Chem. 11 (2003) 5391–5399
5395
Table 1 (continued)
Compd
LE
MMP-2
MMP-9
Inhibition % at 100 mM IC50 (mM) Inhibition % at 200 mM IC50 (mM) Inhibition % at 200 mM IC50 (mM)
4
82
89
89
44
0
4
37
—
—
—
—
—
—
—
—
37
—
—
—
—
—
—
—
—
5
30
0
0
6
6
0
0
7
25
—
—
—
—
0
0
3
0
0
10
19
20
23
0
0
0
0
0
20
0
0
Nd, not determined, due to non-constant slope of the inhibition exerted.
Cytotoxicity tests were performed only for two b-lac-
tams showing noteworthyinhibition of LE, compounds
4 and 8. The viabilityof NIH-3T3 murine fibroblasts
cultured in presence of b-lactams was not significantly
impaired bythe presence of the tested molecules. At
50 mM (concentration 5- and 100-fold the IC50 of epi-
gallocatechin-3-gallate (EGCG) against gelatinase
MMP-2 and LE, respectively),4ꢀ6 the ratio of dead cells
after 24 h was not increased compared with the control,
while the cell proliferation (as number of viable cells)
was lowered by19 and 8%, respectively, in triplicate
experiments (SD <10%).

5396
G. Cainelli et al. / Bioorg. Med. Chem. 11 (2003) 5391–5399
Regarding inhibition activity, we first investigated
4-alkylidene-b-lactams 1 and 8 with the hydroxy group
protected as tert-butyldimethylsilyl derivative (TBS).
Both stereoisomers inhibit LE, and in particular 8
showed a good IC50 value (9 mM), about 22-fold less
derivative 20: the former was totallyinactive over LE
and MMPs, while the latter showed a weak activityon
LE alone.
potent
than
epigallocatechin-3-gallate
(EGCG,
Conclusions
IC50=0.4 mM),6 but as potent as other b-lactam inhibi-
tors of LE, such as the azetidin-2-one L-680,833
(IC50=9 mM),35 selected as reference compounds. The
corresponding OH derivatives 2 and 9 exhibited a
potencylower than the OTBS precursors, indicating
that a lipophylic character is preferred for LE inhibi-
tion. As regards gelatinases, b-lactams 1 and 2 gave
In this report, we have described the synthesis of
4-alkylidene-b-lactams and their evaluation as inhibi-
tors of LE and gelatinases MMP-2 and MMP-9. We
demonstrated that the presence of the C-4 unsaturation
on the b-lactam ring was crucial for the biological
activity. Investigation on C-3, C-4 and N-1 substitution
disclosed a selectivityover LE by3-[1-( tert-butyldi-
methylsilyloxy)-ethyl]-4-[1-(ethoxycarbonyl)-ethylidene]-
azetidin-2-ones (compounds 1, 4, 5, 6, 7, and 8), and
over MMP-2 byC-3-unsubstituted 4-[(1-etho-xy
carbonyl)-ethylidene]-azetidin-2-ones (compounds 14
and 18). Onlycompound 2 effectivelyinhibits MMP-9.
N-Acyl-b-lactams were more effective than N-H analo-
gues over LE. The best results were obtained with com-
pound 18 (IC50 60 mM over MMP-2) and compound 4
(IC50 4 mM, and Ki 10 mM over LE), in the latter being
typical of non-competitive inhibition.
satisfactoryvalues of IC
over MMP-2 (IC50 150 and
50
85 mM, respectively), and 2 over MMP-9, with IC50
150 mM. These results are particularlyinteresting since,
to our knowledge, this is the first example of b-lactams
active as inhibitors of gelatinases.
The absence of a substituent on the C-3 position of the
b-lactam ring, as in compounds 14 and 18, causes a
severe loss of activityover LE, but theyexhibited an
interesting selectivityover MMP-2 (IC 160 mM for 14,
and 60 mM for 18). As regards the influence of the C-4
50
double bond geometryon the activityover MMP-2,
E
isomers (1, 2 and 18) were found to be more potent than
Z isomers (8, 9 and 14). In a tentative explanation of
this result, it should be recalled that we revealed the
presence in Z-alkylidene-b-lactams (but not in the E
isomers) of an intramolecular hydrogen bond between
the NH and the ethoxycarbonyl group on the C-4 side
chain.24 This could cause the reduction of activityin Z
isomers, suggesting the possibilityof a specific binding
interaction through the b-lactam NH group. This inter-
action should be more efficient in the E isomers given
the absence of the intramolecular H-bond.
This is the first example of b-lactams inhibiting metallo-
proteinases instrumental in cancer invasion and angio-
genesis.
These results mayrepresent a first step in designing new
b-lactams combining in the same molecule enhanced
and possiblydifferentiated antibiotic and anti-inflam-
matoryproperties, to be used in the treatment of var-
ious pathologies.
Routine tests for determining the antibiotic potential of
all tested molecules are in progress, and subsequent tests
will be performed on molecules of selection in order to
exclude cytotoxicity, as already reported for compounds
4 and 8.
The S-phenyl-carbonyl derivative 11 showed good
activityover LE and MMP-2, whereas the benzoyl ( 12)
and the cinnamoyl analogue (13) maintained the LE
activity, but they were found to be inactive over MMPs.
Indeed, the presence of an ester or a thioester on the C-4
side chain is crucial. This requirement occurs even for
C-3 unsubstituted b-lactams, where substitution of the
ethoxycarbonyl side chain (compound 14) with a ben-
zoyl one (compound 15) resulted in a complete loss of
activity.
Experimental
General
Commercial reagents were used as received without
additional purification. Compounds 1, 2, 8, 9, 12, 13,
14, 15, 16, 17 and 18 have been prepared according to
refs 23 and 24. S-phenyl diazothioacetate was prepared
according to ref 36. Anhydrous solvents were obtained
commerciallyand used without further drying. 1H and
13C NMR values were recorded on a Varian Mercury
400, Inova 300 or Gemini 200 instrument using a 5-mm
probe. All chemical shifts have been quoted relative to
deuterated solvent signals, d in ppm, J in Hz. FT-IR:
Nicolet 205 measured as films or Nujol mull between
N-acyl and N-alkyl-b-lactams were then examined.
N-acyl derivatives showed an improvement of LE
inhibition activity; in particular compounds 4 and 6
showed the best values of IC50 in the series (4 and 6 mM,
10-fold that of EGCG and better than L-680,833). The
N-acetyl (7) and the N-carbobenzyloxy (5) derivatives
were less potent and N-alkyl-b-lactams 3 and 10 were
inactive over LE. Neither N-acyl nor N-alkyl-4-alkyli-
dene-b-lactams are active against gelatinases, thus rein-
forcing the hypothesis for a specific interaction of the
b-lactam NH group with gelatinases, as discussed above.
In order to attribute the inhibitoryactivityconfidentlyto
the presence of a C¼C double bond on the C-4 position
of the b-lactam ring, we tested the saturated analogues
on C-4, compound 19, and the corresponding N-acyl
NaCl plates and reported in cmꢀ1. TLC: Merck 60 F254
.
Column chromatography: Merck silica gel 200–300
mesh. GC–MS: HP 5890 II-HP 5971, column: HP 5
(30 m, 0.25 mm, 0.25 mm) temperature programme:
50 ꢂC (2 min), 10ꢂC/min, 250 ꢂC. Elemental analyses
were performed at the Istituto Geologia Marina C.N.R.

G. Cainelli et al. / Bioorg. Med. Chem. 11 (2003) 5391–5399
5397
Bologna, Italyand were within 0.4% of the calculated
2H), 4.20 (dd, J=4.4 Hz, J=1.8Hz, 1H), 4.66 (dq,
J=4.4 Hz, J=6.6 Hz, 1H), 5.33 (s, 2H), 6.28 (d,
J=1.8 Hz, 1H), 7.35 (m, 5H). 13C NMR (CDCl3,
75.45 MHz): d ꢀ5.1, ꢀ4.6, 14.3, 17.9, 20.2, 25.6, 60.3, 65.1,
65.4, 68.6, 99.1, 128.3, 128.7, 134.4, 147.3, 148.7, 163.6,
25
D
values. The ½aŠ were determined with a Perkin-Elmer
343 polarimeter. Melting points were determined using
a Buchi apparatus and are uncorrected.
165.9. ½aŠ25=+65 (c 0.4, CHCl3). IR (film): 2926, 1845,
D
Supporting information
1752, 1713, 1666cmꢀ1. Anal. (C23H33NO6Si) C, H, N.
1H NMR and 13C NMR for all new compounds (3, 4, 5,
6, 7, 10, 11, 19, 20) are available.
Ethyl {(2E),(3S)-1-benzylcarbamoyl-3-[(1R)-1-(tert-bu-
tyldimethylsilanyl oxy)-ethyl]-4-oxo-azetidin-2-ylidene}-
acetate (6). Compound 1 (0.31 g, 1 mmol) and benzyl
isocyanate (0.14 mL, 1 mmol) were dissolved in anhy-
drous acetone (10 mL) and then K2CO3 (0.14 g, 1 mmol)
was added, with a similar procedure as that described
for 3. Product 6 (0.268 g, 60% yield) was isolated as a
Ethyl {(2E),(3S)-1-benzyl-3-[(1R)-1-(tert-butyldimethyl-
silanyloxy)-ethyl]-4-oxo-azetidin-2-ylidene}-acetate (3).
Compound 1 (0.31 g, 1 mmol) and benzyl bromide
(0.12 mL, 1 mmol) were dissolved in anhydrous acetone
(10 mL); K2CO3 (0.14 g, 1 mmol) was added and the
reaction mixture was stirred for 2 h. Then K2CO3 was
filtered, the solvent was removed and the crude oily
residue was immediatelypurified byflash chromato-
graphy (ethyl acetate–cyclohexane 5:95). Product 3
(0.323 g, 80%) was isolated as a pale yellow oil; 1H
NMR (CDCl3, 200 MHz): d 0.11 (s, 3H), 0.12 (s, 3H),
0.92 (s, 9H), 1.17 (d, J=6.4 Hz, 3H), 1.26 (t, J=7.2 Hz,
3H), 4.13 (q, J=7.2 Hz, 2H), 4.18 (m, 1H), 4.45 (d,
J=16 Hz, 1H), 4.60 (d, J=16 Hz, 1H), 4.65 (dq, J=2.2
and J=6.4 Hz, 1H), 5.10 (d, J=1.4 Hz, 1H), 7.22–7.4
(m, 5H). 13C NMR (CDCl3, 50.29 MHz): d ꢀ4.8, ꢀ4.6,
14.4, 18.0, 19.8, 25.8, 44.2, 59.8, 65.0, 65.2, 90.5, 127.7,
128.0, 128.8, 134.1, 155.5, 165.7, 167.5. ½aŠ2D5=+76 (c
1.5, CHCl3). IR (film): 2925, 1806, 1706, 1653,
1262 cmꢀ1. GC–MS: Rt=23.4 min; m/e 388 (M+ꢀ15,
1), 346 (44), 185 (9), 143 (7), 115 (4), 91 (100), 73 (24),
59 (5). Anal. (C22H33NO4Si) C, H, N.
1
pale yellow oil; H NMR (CDCl3, 200 MHz): d 0.11 (s,
3H), 0.12 (s, 3H), 0.90 (s, 9H), 1.23 (d, J=6.4 Hz, 3H),
1.30 (t, J=7.0 Hz, 3H), 4.19 (q, J=7.0 Hz, 2H), 4.24
(dd, J=4.4 Hz, J=1.6 Hz, 1H), 4.44 (dd, J=5.8 Hz,
J=15.8 Hz, 1H), 4.52 (dd, J=5.8 Hz, J=15.8 Hz, 1H),
4.70 (dq, J=4.4 Hz and J=6.4 Hz), 1H), 6.44 (d,
J=1.6 Hz, 1H), 6.92 (t, J=5.8 Hz, 1H), 7.35 (m, 5H).
13C NMR (CDCl3, 50.29 MHz): d ꢀ5.0, ꢀ4.6, 14.3,
17.9, 19.9, 25.6, 43.7, 60.2, 64.8, 65.1, 98.6, 127.6, 127.7,
128.7, 137.0, 148.5,37 165.9, 167.7. ½aŠ2D5=+69 (c 1.5,
CHCl3). IR (film): 3377, 2933, 1819, 1739, 1666,
1534 cmꢀ1. GC–MS: Rt=17.7 min; m/e 256 (M+ꢀ148,
63), 210 (100), 184 (19), 168 (6), 143 (43), 115 (6), 99
(16), 75 (50), 59 (8). Anal. (C23H34N2O5Si) C, H, N.
Ethyl {(2E),(3S)-1-acetyl-3-[(1R)-1-(tert-butyldimethylsil-
anyloxy)-ethyl]-4-oxo-azetidin-2-ylidene}-acetate (7).
Compound 1 (0.31 g, 1 mmol) and acetic anhydride
(0.10 mL, 1 mmol) were dissolved in anhydrous acetone
(10 mL). Then K2CO3 (0.14 g, 1 mmol) was added with a
similar procedure as that described for 3. Product 7
(0.284 g, 80% yield) was isolated as a pale yellow oil; 1H
NMR (CDCl3, 200 MHz): d ꢀ0.07 (s, 3H), 0.03 (s, 3H),
0.80 (s, 9H), 1.30 (t, J=6.8 Hz, 3H), 1.40 (d, J=6.4 Hz,
3H), 2.42 (s, 3H), 4.05 (dd, J=1.4 Hz, J=1.6 Hz, 1H),
4.20 (q, J=6.8 Hz, 2H), 4.73 (dq, J=6.4 Hz, J=1.6 Hz,
1H), 6.52 (d, J=1.4 Hz, 1H).13C NMR (CDCl3,
50.29 MHz): d ꢀ5.5, ꢀ4.4, 14.4, 17.7, 22.0, 24.1, 25.5,
60.3, 65.0, 65.6, 99.6, 150.3, 165.7, 166.5, 166.7.
Ethyl {(2E),(3S) - 3 - [(1R) - 1 - (tert - butyldimethylsilanyl-
oxy)-ethyl]-2-ethoxy carbonylmethylene-4-oxo-azetidin-
1-yl}-3-oxo-propionate (4). Compound
1
(0.13 g,
0.415 mmol) and ethyl malonyl chloride (0.09 mL,
0.715 mmol) were dissolved in anhydrous benzene
(3 mL). The reaction mixture was refluxed for 3 h, then
the solvent was removed and the crude oilyresidue was
immediatelypurified byflash chromatography(ethly
acetate–cyclohexane 10:90). Product 4 (0.035 g, 20%
yield) was isolated as a pale yellow oil; 1H NMR
(CDCl3, 300 MHz): d 0.11 (s, 3H), 0.12 (s, 3H), 0.90 (s,
9H), 1.22 (d, J=6.4 Hz, 3H), 1.26–1.32 (m, 6H), 3.80 (s,
2H), 4.21 (m, 4H), 4.28 (dd, J=4.5 Hz, J=1.2 Hz, 1H),
4.72 (m, 1H), 6.55 (d, J=1.2 Hz, 1H). 13C NMR
(CDCl3, 75.45 MHz): d ꢀ5.0, ꢀ4.7, 14.0, 14.2, 17.9,
19.7, 25.6, 43.5, 60.4, 61.9, 65.2, 65.2, 100.9, 147.9,
½aŠ25=ꢀ189 (c 1.2, CHCl3). IR (film): 2959, 1832, 1726,
D
1660, cmꢀ1. GC–MS: Rt=18.2 min; m/e 340 (M+ꢀ15,
1), 298 (96), 256 (100), 228 (16), 184 (21), 103 (20), 75
(53), 59 (7). Anal. (C17H29NO5Si) C, H, N.
162.1, 165.3, 165.7, 165.9. ½aŠ25=+ 61 (c 1.06, CHCl3).
Ethyl {(2Z),(3S)-1-benzyl-3-[(1R)-1-(tert-butyldimethyl-
silanyloxy)-ethyl]-4-oxo-azetidin-2-ylidene}-acetate (10).
From compound 8 (0.31 g, 1 mmol) and benzyl bro-
mide (0.12 mL, 1 mmol), a similar procedure as that
described for 3, gave product 10 (0.31 g, 77% yield) as a
D
IR (film): 2932, 1836, 1724, 1665 cmꢀ1
(C20H33NO7Si) C, H, N.
.
Anal.
Benzyl (2E),(3S)- 3-[(1R)-1-(tert-butyldimethylsilanyl-
oxy)-ethyl]-2-ethoxy carbonylmethylene-4-oxo-azetidine-
1-carboxylate (5). Compound 1 (0.31 g, 1 mmol) and
benzyl chloroformate (0.15 mL, 1 mmol), were dissolved
in anhydrous acetone (10 mL) and then K2CO3 (0.14 g,
1 mmol) was added with a similar procedure as that
described for 3. Product 5 (0.309 g, 69% yield) was iso-
1
pale yellow oil; H NMR (CDCl3, 200 MHz): d 0.03 (s,
3H), 0.07 (s, 3H), 0.87 (s, 9H), 1.21 (t, J=7.0 Hz, 3H),
1.32 (d, J=6.2 Hz, 3H), 3.59 (d, J=5.8 Hz, 1H), 4.10 (q,
J=7.0 Hz, 2H), 4.10–4.22 (m, 1H), 5.04 (d, J=15.8 Hz,
1H), 5.14 (d, J=15.8 Hz, 1H), 5.20 (s, 1H), 7.20–7.40
(m, 5H). 13C NMR (CDCl3, 75.45 MHz): d ꢀ4.8, ꢀ4.3,
14.3, 17.9, 22.3, 25.7, 46.6, 59.8, 64.1, 65.8, 92.2, 127.2,
127.7, 128.4, 136.8, 152.4, 165.0, 168.4. ½aŠ2D5=+21 (c
1
lated as a pale yellow oil; H NMR (CDCl3, 200 MHz):
d 0.09 (s, 3H), 0.10 (s, 3H), 0.87 (s, 9H), 1.23 (d,
J=6.6 Hz, 3H), 1.30 (t, J=7 Hz, 3H), 4.19 (q, J=7 Hz,
2.57, CHCl3). IR (film): 2926, 1806, 1713, 1646 cmꢀ1
.

5398
G. Cainelli et al. / Bioorg. Med. Chem. 11 (2003) 5391–5399
GC–MS: Rt=23.4 min m/e 346 (M+ꢀ57, 47), 185 (10),
170 (3), 143 (7), 103 (6), 91 (100), 65 (6). Anal.
(C22H33NO4Si) C,H,N.
0.32 mmol) and ethyl malonyl chloride (0.07 mL,
0.54 mmol), with a procedure similar to that described
for 4, product 20 (0.051 g, 37% yield) was isolated as a
1
colorless oil; H NMR (CDCl3, 200 MHz): d 0.05 (s,
S-Phenyl-{(2Z), (3S)-3-[(1R)-1-(tert-butyldimethylsilanyl-
oxy)-ethyl]-4-oxo-azetidin-2-ylidene}-thioacetate (11).
To
3H), 0.08 (s, 3H), 0.85 (s, 9H), 1.21–1.32 (m, 9H), 2.76
(dd, J=15.4 Hz, J=8.0 Hz, 1H), 3.10–3.20 (m, 2H),
3.61 (d, J 4 Hz, 1H), 4.49 (m, 1H). 13C NMR (CDCl3,
75.45 MHz): d ꢀ5.2, ꢀ4.3, 14.1, 14.0, 17.8, 22.0, 25.6,
36.4, 43.3, 49.3, 60.9, 61.5, 62.9, 65.0, 163.3, 166.0,
a
solution of [3R(10R,4R)4-acetoxy3-[1-(tbutyl-
dimethylsilyloxy) ethyl]-azetidin-2-one (0.959 g, 3.343
mmol) in CH2Cl2 (33 mL), Et3N (0.64 mL, 4.346 mmol)
and Me3SiCl (0.465 mL, 3.677 mmol) were added at
room temperature. At complete conversion of the start-
ing material (GC monitoring) the solution was brought
to 0 ꢂC and 3.34 mL (3.34 mmol) of a 1 M solution of
TiCl4 in CH2Cl2 were added. After 1 min, a solution of
2 equiv of S-phenyl diazothioacetate35 in 6 mL of
CH2Cl2 was slowlyadded dropwise (over 30 min) at
0 ꢂC. The reaction mixture was stirred at room tem-
perature and the conversion monitored byTLC. After
4 h the reaction was quenched in ice cold water and
extracted with CH2Cl2 (3 ꢃ 40 mL), dried on Na2SO4
and concentrated. The residue was purified byflash
chromatography (cyclohexane–ethylacetate 95:5) to give
product 11 (0.62 g, 50% yield) as a pale yellow solid; mp
129–131 ꢂC. 1H NMR (CDCl3, 300 MHz): d 0.11 (s,
3H), 0.12 (s, 3H), 0.92 (s, 9H), 1.36 (d, J=6.3 Hz, 3H),
3.73 (d, J=5.4 Hz, 1H), 4.27 (quintet, J=6.3 Hz, 1H),
5.73 (s, 1H), 7.45–7.53 (m, 5H), 8.62 (bs, 1H). 13C NMR
(CDCl3, 75.45 MHz): d ꢀ5.0, ꢀ4.2, 17.9, 22.4, 25.7,
65.1, 65.4, 97.2, 127.7, 129.2, 129.5, 134.8, 151.4, 166.4,
166.2, 169.8. ½aŠ25=ꢀ63 (c 1.46, CHCl3). IR (film):
D
2933, 1799, 1746, 1720 cmꢀ1. Anal. (C20H35NO7Si) C,
H, N.
Materials for biological assays
Elastase from human leukocytes and elastase substrate
N-methoxysuccinyl-ala-ala-pro-val p-nitroanilide were
purchased from Sigma.
Cytotoxicity test
NIH-3T3 murine fibroblasts were used for cytotoxi-
citytests. Cells (3 ꢃ 106) were seeded onto 75 cm2
cell culture flasks and incubated at 37 ꢂC in Dulbecco
Modified Eagle’s Medium supplemented with 10%
fetal calf serum, with or without b-lactams 50 mM
(>10-fold the IC50 of compound 4, the most active
against LE). After 24 h, adherent cells were harvested
with trypsin-EDTA (0.05–0.2%, Sigma), and added
to those previouslypelleted from suspension; follow-
ing Trypan Blue (Sigma) vital staining, alive and
dead cells were counted. In order to discriminate
between possible cytotoxic and cytostatic effects, the
values were subtracted from those from control
experiments.
187.5. ½aŠ25=ꢀ24 (c 1, CHCl3). IR (film): 3185, 1819,
D
1786, 1679, 1633 cmꢀ1. Anal. (C19H27NO3Si) C,H,N.
Ethyl {(3S,2R)-3-[(1R)-1-(tert-butyldimethylsilanyloxy)-
ethyl]-4-oxo-azetidin-2-yl}-acetate (19). In a double-
necked round-bottom flask metallic Zn was warmed
under vacuum and, after cooling, [3R(10R,4R)4-ace-
toxy3-[1-(tbutyldimethylsilyloxy) ethyl]-azetidin-2-one
(0.861 g, 3 mmol) and THF (30 mL) were added. The
solution was heated to reflux and 2-bromoethyl acetate
(0.5 mL, 4.5 mmol) was added dropwise. The reaction
mixture was stirred for 2.5 h and then quenched in a
buffer solution (pH 7.5), the zinc salt was filtered on
Celite and the solution was extracted with AcOEt (2 ꢃ
20 mL), dried on Na2SO4 and concentrated. The residue
was purified byflash chromatography(ccylohexane-
ethylacetate 7:3) to give product 19 (0.210 g, 23%) as a
white crystalline solid; mp=56–59 ꢂC. 1H NMR
(CDCl3, 300 MHz): d 0.06 (s, 6H), 0.86 (s, 9H), 1.20 (d,
J=6.3 Hz, 3H), 1.26 (t, J=7.2 Hz, 3H), 2.55 (dd,
J=9.6 Hz, J=16.2 Hz, 1H), 2.72 (dd, J=3.9 Hz,
J=16.2 Hz, 1H), 2.80 (dd, J=2.1 Hz, J=5.1 Hz, 1H),
3.95 (ddd, J=2.1 Hz, J=3.9 Hz, J=9.6 Hz, 1H), 4.15
(q, J=7.2 Hz, 2H), 4.18 (dq, J=5.1 Hz, J=6.3 Hz, 1H),
6.23 (bs, 1H). 13C NMR (CDCl3, 75.45 MHz): d ꢀ5.3,
ꢀ4.5, 13.9, 17.7, 22.2, 25.5, 39.5, 46.6, 60.6, 63.8, 65.2,
Substrate degradation by HLE
LE was solubilized (250 mU/mL) in Hepes buffer (0.1 M
Hepes, 0.5 m NaCl, 10% DMSO, pH 7.8). All the com-
pounds and elastase substrate were freshlyprepared
20ꢃ in DMSO. Dilutions of the compounds were pre-
mixed with the enzyme in micrometers wells, and main-
tained 15 min at 4 ꢂC. Then, 5 mL of the substrate (8 mM
concentrated) were added to 100 mL final volume, and
the mixture was incubated at 37 ꢂC. At 20-min intervals,
the intensityof the color developed bythe digested
substrate was measured at 405 nm using a Titertek
Mutiskan (Flow Laboratories), and the control back-
ground subtracted (in triplicate experiments). The reac-
tions developed linearlyfor as long as 120 min; data
from 60 min were used for determining all IC50s. Dou-
ble-reciprocal plot of the results allowed the type and Ki
of inhibition exerted over LE bycompound 4 to be
deduced.
168.0, 170.9. ½aŠ25=+12 (c 1.43, CHCl3). IR (film):
D
3262, 2930,1763,1736 cmꢀ1. GC–MS: Rt=21.2 min m/e
300 (M+ꢀ15, 2), 258 (100), 214 (25), 170 (23), 115 (25),
75 (70). Anal. (C15H29NO4Si) C, H, N.
Zymographic analysis
Aliquots of gelatinase-containing medium conditioned
byhuman neuroblastoma cells (for MMP-2) or
HT-1080 human fibrosarcoma cells (for MMP-9) were
assayed as described.5 Without heating the samples,
zymography was performed by electrophoresing 15 mL
Ethyl {(3S,2R)-3-[(1R)-1-(tert-butyldimethylsilanyloxy)-
ethyl]-2-ethoxycarbonylmethyl-4-oxo-azetidin-1-yl}-3-
oxo-propionate (20). Obtained from 19 (0.1 g,

G. Cainelli et al. / Bioorg. Med. Chem. 11 (2003) 5391–5399
5399
of medium in 0.1% gelatin-containing 8% poly-
acrylamide, in presence of SDS. After electrophoresis,
the gels were washed twice for 15 min with 2.5% Triton
X-100, incubated overnight at 37 ꢂC in Tris buffer
(50 mM Tris–HCl, 200 mM NaCl, 10 mM CaCl2, pH
7.4). For gelatinase inhibition assays, compounds were
freshlysolubilized in DMSO and diluted in Tris buffer
used for developing the zymogram. The gel slab was
then cut into slices corresponding to the lanes and then
put in different tanks containing the stated concentra-
tions of inhibitors.
11. Beauve, C.; Bouchet, M.; Touillaux, R.; Fastrez, J.;
Marchand-Brynaert, J. Tetrahedron 1999, 55, 13301.
12. Prasad, K.; Kneussel, P.; Schulz, G.; Stuetz, P. Tetra-
hedron Lett. 1982, 23, 1247.
13. Greengrass, C. W.; Hoople, D. W. T. Tetrahedron Lett.
1981, 22, 1161.
14. Bachi, M. D.; Goldberg, O.; Gross, A.; Vaya, J. J. Org.
Chem. G 45, 1481–1485.
15. Battaglia, A.; Cainelli, G.; Giacomini, D.; Martelli, G.;
Panunzio, M. Tetrahedron Lett. 1987, 28, 4347.
16. Cainelli, G.; Giacomini, D.; Galletti, P. Eur. J. Org.
Chem. 1999, 61.
17. Cainelli, G.; Galletti, P.; Giacomini, D.; Orioli, P. Angew.
Chem., Int. Ed. Enl. 2000, 39, 523.
18. Cainelli, G.; Giacomini, D.; Galletti, P. Synthesis 2000,
289.
19. Cainelli, G.; Galletti, P.; Giacomini, D. Tetrahedron Lett.
1998, 39, 7779.
The gels then were stained for 30 min with 30% metha-
nol/10% acetic acid containing 0.5% Coomassie Bril-
liant Blue R-250, and destained in the same solution
without dye. Clear bands on the blue background
represent areas of gelatinolysis. Digestion bands were
quantitated using an image analyzer system with Gel-
Doc 2000 and QuantityOne software (Bio-Rad), and
the densitometric values were expressed as percentage of
control bands on the same gel.
20. Cainelli, G.; Umani Ronchi, A.; Contento, M.; Galletti,
P.; Giacomini, D.; Da Col, M.; Dall’Asta L., Resta I. PCT Int.
Appl., 1997; Coden: PIXXD2 WO 9742170 A1 19971113.
21. Bandini, E.; Cainelli, G.; Giacomini, D.; Martelli, G.;
Panunzio, M.; Spunta, G. Bio. Med. Chem. Lett. 1993, 3, 2347.
22. Cainelli, G.; Panunzio, M.; Giacomini, D.; Martelli, G.;
Spunta, G. J. Am. Chem. Soc. 1988, 110, 6879.
23. Cainelli, G.; Galletti, P.; Gazzano, M.; Giacomini, D.;
Quintavalla, A. Tetrahedron Lett. 2002, 43, 233.
24. Cainelli, G.; Giacomini, D.; Galletti, P.; Quintavalla, A.
Eur. J. Org. Chem. 2003, 1765.
25. Gunda, T. E. Org. Lett. 2000, 2, 103.
26. Stoodley, R. J.; Whitehouse, N. R. J. Chem. Soc., Perkin
Trans. 1 1973, 32.
27. Brennan, J.; Hussain, F. H. S. Synthesis 1985, 749.
28. Imuta, M.; Uyeo, S.; Nakano, M.; Yoshida, T. Chem.
Pharm. Bull. 1985, 33, 4371.
Acknowledgements
We thank Ms. Silvia Spiga for assistance with experi-
mental work. This research was supported byMURST
(60% and COFIN), the Universityof Bologna (funds
for selected topics), and bygrants from the ‘Associa-
zione Italiana per la Ricerca sul Cancro’. We are grate-
ful to Dr S. Biggin for revision of the English
manuscript.
29. Imuta, M.; Uyeo, S.; Nakano, M.; Yoshida, T. Chem.
Pharm. Bull. 1985, 33, 4371.
30. John, D. I.; Tyrrell, N. D.; Thomas, E. J. Tetrahedron
1983, 39, 2477.
References and Notes
31. Refs 23 and 24 refer to the attribution of E/Z stereo-
chemistrywhich was fullyascertained by 1H/13C NMR and X-
raydiffraction analysis.
32. Cainelli, G.; Giacomini, D.; Gazzano, M.; Galletti, P.;
Quintavalla, A. Tetrahedron Lett. 2003, 44, 6269. Alajarin, M.;
Vidal, A.; Sanchez-Anprada, P.; Tovar, F.; Ochoa, G. Org.
Lett. 2000, 2, 965.
1. Sternlicht, M. D.; Werb, Z. Neutrophil Elastase and
Cathepsin G. In Extracellular Matrix, Anchor, and Adhesion
Proteins, Kreis, T., Vale, R., Eds.; Oxford UniversityPress:
Oxford UK, 1999; p 543.
2. Egeblad, M.; Werb, Z. Nat. Rev. 2002, 2, 161.
3. Coussens, L. M.; Fingleton, B.; Matrisian, L. M. Science
2002, 295, 2387.
33. Brodney, M. A.; Padwa, A. J. Org. Chem. 1999, 64, 556.
34. Padwa, A.; Prein, M. Tetrahedron 1998, 54, 6957.
35. (a) Durette, P. L.; Chabin, R. M.; Fletcher, D. S.; Green,
B. G.; Hanlon, W. A.; Humes, J. L.; Knight, W. B.; Lanza,
T. J., Jr.; Mumford, R. A.; Pacholok, S.; MacCoss, M. Bioorg.
Med. Chem. Lett. 1995, 5, 271. (b) Doherty, J. B.; Shah, S. K.;
Finke, P. E.; Dorn, C. P., Jr.; Hagmann, W. K.; Hale, J. J.;
Kissinger, A. L.; Thompson, K. R.; Brause, K.; Chandler,
G. O.; Knight, W. B.; Maycock, A. L.; Ashe, B. M.; Weston,
H.; Gale, P.; Mumford, R. A.; Andersen, O. F.; Williams,
H. R.; Nolan, T. E.; Frankenfield, D. L.; Underwood, D.;
Vyas, K. P.; Kari, P. H.; Dahlgren, M. E.; Mao, J.; Fletcher,
D. S.; Dellea, P. S.; Hand, K. M.; Osinga, D. G.; Peterson,
L. B.; Williams, D. T.; Metzger, J. M.; Bonney, R. J.; Humes,
J. L.; Pacholok, S. P.; Hanlon, W. A.; Opas, E.; Stolk, J.;
Davies, P. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 8727.
36. Lawlor, M. D.; Thomas, W. L.; Danheiser, R. L. J. Org.
Chem. 2000, 65, 4375.
4. Garbisa, S.; Biggin, S.; Cavallarin, N.; Sartor, L.; Benelli,
R.; Albini, A. Nat. Med. 1999, 5, 1216.
5. Garbisa, S.; Sartor, L.; Biggin, S.; Salvato, B.; Benelli, R.;
Albini, A. Cancer 2001, 91, 822.
6. Sartor, L.; Pezzato, E.; Garbisa, S. J. Leukoc. Biol. 2002,
71, 73.
7. Dell’Aica, I.; Dona, M.; Sartor, L.; Pezzato, E.; Garbisa, S.
Lab. Invest. 2002, 82, 1685.
8. Dona, M.; Dell’Aica, I.; Calabrese, F.; Benelli, R.; Morini,
M.; Albini, A.; Garbisa, S. J. Immunol. 2003, 4335.
9. Sartor, L.; Pezzato, E.; Dell’Aica, I.; Caniato, R.; Biggin,
S.; Garbisa, S. Biochem. Pharmacol. 2002, 64, 229.
10. (a) Konaklieva, M. I. Curr. Med. Chem. Anti-Infective
Agents 2002, 1, 215. (b) As recent examples of b-lactam inhi-
bitors of HLE, see: Singh, R.; Micetich, R. G. IDrugs 2000, 3,
512. (c) Clemente, A.; Domingos, A.; Grancho, A. P.; Iley, J.;
Moreira, R.; Neres, J.; Palma, N.; Santana, A. B.; Valente, E.
Bioorg. Med. Chem. Lett. 2001, 11, 1065. (d) Koteva, K. P.;
Cantin, A. M.; Neugebauer, W. A.; Escher, E. Can. J. Chem.
2001, 79, 377.
37. This signal is the sum of two carbons that have the same
chemical shift. This evidence derives from an integrated 13C
spectrum recorded with a delayof 60 s and bygHMBC analyses.