Structure-based design of β-lactamase inhibitors. 1. Synthesis and evaluation of bridged monobactams

Article abstract of DOI:10.1021/jm980023c

Bridged monobactams are novel, potent, mechanism-based inhibitors of class C β-lactamases, designed using X-ray crystal structures of the enzymes. They stabilize the acyl-enzyme intermediate by blocking access of water to the enzyme-inhibitor ester bond.

Full text of DOI:10.1021/jm980023c

J . Med. Chem. 1998, 41, 3961-3971
3961
Str u ctu r e-Ba sed Design of â-La cta m a se In h ibitor s. 1. Syn th esis a n d Eva lu a tion
of Br id ged Mon oba cta m s
Ingrid Heinze-Krauss, Peter Angehrn, Robert L. Charnas, Klaus Gubernator,† Eva-Maria Gutknecht,
Christian Hubschwerlen,* Malgosia Kania, Christian Oefner, Malcolm G. P. Page,* Satoshi Sogabe,
J ean-Luc Specklin, and Fritz Winkler
Preclinical Research, F. Hoffmann-La Roche Ltd., CH-4070 Basle, Switzerland
Received J anuary 14, 1998
Bridged monobactams are novel, potent, mechanism-based inhibitors of class C â-lactamases,
designed using X-ray crystal structures of the enzymes. They stabilize the acyl-enzyme
intermediate by blocking access of water to the enzyme-inhibitor ester bond. Bridged
monobactams are selective class C â-lactamase inhibitors, with half-inhibition constants as
low as 10 nM, and are less effective against class A and class B enzymes (half-inhibition
constants > 100 µM) because of the different hydrolysis mechanisms in these classes of
â-lactamases. The stability of the acyl-enzyme complexes formed with class C â-lactamases
(half-lives up to 2 days were observed) enabled determination of their crystal structures. The
conformation of the inhibitor moiety was close to that predicted by molecular modeling,
confirming a simple reaction mechanism, unlike those of known â-lactamase inhibitors such
as clavulanic acid and penam sulfones, which involve secondary rearrangements. Synergy
between the bridged monobactams and â-lactamase-labile antibiotics could be observed when
such combinations were tested against strains of Enterobacteriaceae that produce large amounts
of class C â-lactamases. The minimal inhibitory concentration of the antibiotic of more than
64 mg/L could be decreased to 0.25 mg/L in a 1:4 combination with the inhibitor.
In tr od u ction
class C enzymes stems from the fact that hydrolysis of
the acyl intermediate (deacylation) can occur more
rapidly than the chemical rearrangement to the more
stable species in the class C active site.14 We set out to
use structural and kinetic information about class C
â-lactamases as a basis for the design of mechanism-
based inhibitors that would specifically block the deacy-
lation reaction of this refractory class of enzymes.
Infections caused by Gram-negative bacteria consti-
tute about one-half of the cases of life-threatening
nosocomial disease.1,2 The past decade has seen an
increase in the frequency of resistance to modern
antibiotics including the third-generation cephalospor-
ins.3,4 In many cases, the resistance to â-lactam anti-
biotics is due to a high level of expression of class C
â-lactamases.2,5,6 These enzymes are mostly chromo-
somally encoded, but recently class C enzymes have
appeared on plasmids, thus crossing interspecies bound-
aries.7 Available types of â-lactamase inhibitors such
as clavulanate (1), sulbactam (2a ), and tazobactam (2b)
are ineffective against these â-lactamases and are
therefore not suitable for use in combination with
â-lactamase-susceptible antibiotics against these emer-
gent organisms.2,3,8
Design of In h ibitor s of Cla ss C â-La cta m a ses
The starting point for inhibitor design was solution
of the crystal structures of Citrobacter freundii 1203
â-lactamase and the acyl-enzyme complex formed with
aztreonam (Scheme 1) at 2.0 and 2.5 Å, respectively.15
Comparison of the aztreonam moiety observed in the
complex with the intact â-lactam molecule indicated
that rotation about the C3-C4 bond by ∼70° had
occurred. The counterclockwise rotation about the C3-
C4 bond relaxes the eclipsed conformation of the intact
â-lactam to the energetically more favorable gauche
form (Scheme 1). Clockwise rotation, to adopt the
alternative gauche configuration, is less favorable be-
cause it would bring the C4 methyl substituent into
steric conflict with the side chains of tyrosine 150 (see
Figure 2) and leucine 119. The rearrangement in the
aztreonam complex leaves N1 and the sulfonate group
in a position where they block one face of the ester bond
formed with serine 64 (Scheme 1, upper reaction). They
also displace a water molecule (WAT192 in Figure 2)
that, in the native enzyme structure, is in a position
where it interacts with an extensive hydrogen-bonded
network of residues that is an essential part of the
catalytic mechanism of class C â-lactamases. The
residues involved in this network, comprising lysine 67,
tyrosine 150, asparagine 152, and lysine 315, all con-
Clavulanate and the penam sulfones undergo chemi-
cal rearrangements that are triggered by attack of the
enzyme on the â-lactam ring and which result in acrylic
esters that are intrinsically more stable to hydrolysis.9-14
The inefficiency of these compounds as inhibitors of the
* To whom correspondence should be addressed. Concerning syn-
theses: C. Hubschwerlen, fax, ++ 061 688 64 59; tel, 061 688 59 67.
For general correspondence and specific questions concerning biological
results: M. G. P. Page, fax, ++ 061 688 27 29; tel, 061 688 38 13.
† Present address: CombiChem Inc., 9050 Camino Santa Fe, San
Diego, CA 92121.
S0022-2623(98)00023-5 CCC: $15.00 © 1998 American Chemical Society
Published on Web 09/16/1998

3962 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 21
Heinze-Krauss et al.
Sch em e 1. Changes in the Configuration of the â-Lactam Moiety Occurring during Reaction with â-Lactamases
tribute somewhat to activation of water for hydrolysis
of the acyl-enzyme complex.15-19 It was postulated
that it is this water, positioned near tyrosine 150, that
is involved in deacylation.15
carbonate 5 and treatment with NBu4Br22,23 (Scheme
2). The triflate derived from the alcohol 6 was reduced
with NaBH4, and the protecting groups were removed
by treatment with K2S2O8 and hydrogenolysis over Pd/C
to give the bridged â-lactam 7.
Molecular modeling, based on the C. freundii struc-
ture, suggested that with penicillins, cephalosporins,
and monobactams where the substituents are cis to one
another across the C3-C4 bond (Scheme 1, lower reac-
tion), clockwise rotation can occur without conflict with
protein side chains. The rotation leaves open the path
for the water molecule to attack the ester, and therefore
cis-substituted monobactams, as well as penicillins and
cephalosporins, are rapidly hydrolyzed by class C en-
zymes.20 Preventing this rotation in a suitable molecule
should block the access of the water molecule to the
ester bond and thus greatly stabilize the acyl-enzyme
complex. We investigated this possibility by synthesiz-
ing monobactams (3) that had N3 linked to C4 by a two-
carbon atom bridge.
The substituted monobactams 8, 10, and 12 were syn-
thesized as described in Scheme 2. In a series of reac-
tions, the 3,4-dimethoxybenzyl (DMB) protecting group
on the â-lactam nitrogen of 6 was replaced by the tert-
butyldimethylsilyl group (TBDMS) after transient THP-
ether protection of the alcohol function. The resulting
intermediate alcohol, activated as a triflate, was reacted
with sodium 5-mercapto-1-methyltetrazole. The TB-
DMS group was removed, and sulfonation with DMF‚
SO3 complex gave compound 8. The intermediate alco-
hol was also oxidized under Swern conditions24 to give
the ketone 9 that was subjected to Wittig reaction and,
after deprotection of the â-lactam nitrogen, was sul-
fonated to give 10. The alcohol 6 was transformed into
the isomeric alcohol 11 after Swern oxidation and
NaBH4 reduction. Compound 11, activated as triflate,
was reacted with NBu4F. Removal of the dimethoxy-
benzyl protecting group followed by sulfonation gave
compound 12.
Compound 7 was treated with di-tert-butyl dicarbon-
ate, sulfonated with DMF‚SO3, and hydrolyzed to give
the betaine 13 (Scheme 2). Treatment of 13 with
O-succinyl-activated esters (method A) and O-succinyl-
activated carbamates (method B) gave the correspond-
Ch em istr y
The key bicyclic derivatives 6 and 7 were obtained
from the easily accessible21 diol 4 by formation of the

Design of â-Lactamase Inhibitors. 1
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 21 3963
Sch em e 2a
a
Reagents: (a) carbonyldiimidazole; (b) Bu4NBr; (c) (1) Tf2O, (2) NaBH4, (3) K2S2O8, (4) H2, Pd/C; (d) (1) K2S2O8, (2) dihydropyran, (3)
TBDMS-Cl, imidazole, (4) pTSA; (e) (1) Tf2O, (2) 5-mercapto-1-methyltetrazole Na salt, (3) DMF‚SO3; (f) Swern oxidation; (g) (1)
Ph3PCHCONH2, (2) NBu4F, (3) DMF‚SO3; (h) NaBH4; (f) (1) Tf2O, (2) NH4F, (3) K2S2O8, (4) DMF‚SO3; (h) (1) (BOC)2O, (2) DMF‚SO3 then
H2O; (i) R1COOSucc (method A) or RNHCOOSucc (method B); (j) ROCOX (method C) or R1COCl (method D) or R1COOH, DCC (method
E); (k) DMF‚SO3; (l) Nuc (method F), see Table 2. DMB, 3,4-dimethoxybenzyl.
Sch em e 3a
Ta ble 1. Preparation of Bridged Monobactams 14-20
a
Reagents: (a) Ph3PCHCHO; (b) (1) NaBH4, (2) H2, Pd/C, (3)
(BOC)2O; (c) (1) MsCl, (2) NaH, (3) K2S2O8, (4) DMF‚SO3 then
NaHCO3.
hydrogenation over Pd/C and treatment with di-tert-
butyl dicarbonate. The mesylate derived from the
alcohol 41 was cyclized with NaH. Deprotection and
sulfonation on the â-lactam nitrogen gave derivative 42.
a
b
EI (M - CONH). ISP (M + H)+.
Resu lts a n d Discu ssion
ing derivatives 21-28. In another approach, compound
7 was first derivatized to give the intermediates 14-
20 (methods C-E, Scheme 2) and then sulfonated with
DMF‚SO3 to give the derivatives 29-35. The bro-
moacetyl derivative 31, obtained following method D,
was reacted with nucleophiles to give compounds 36-
38 (method F, Scheme 2; Tables 1 and 2).
The synthesis of the bridged â-lactam 42 is described
in Scheme 3. The aldehyde 39 obtained from 422 was
subjected to Wittig reaction leading to the unsaturated
aldehyde 40. Reduction to the corresponding alcohol 41
was achieved by treatment with NaBH4 followed by
In h ibition of Cla ss C â-La cta m a ses. The inhibi-
tors, which we term bridged monobactams, were highly
efficacious inhibitors of a wide variety of class C
â-lactamases (Table 3). The reaction with class C
â-lactamases was typically biphasic (Figure 1), reflecting
two kinetically distinct conformations of the protein, as
shown in Scheme 4.18,19,25 The steady-state level of
occupation of the active site, which is dependent on the
ratio of deacylation rate to the acylation rate (k4/k2),26
was greater than 0.999. The low deacylation rate was
reflected in a very low net hydrolysis rate, with kcat/KM
< 1 M-1 s-1 (Figure 1).

3964 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 21
Heinze-Krauss et al.
Sch em e 4. Minimum Reaction Scheme for the
Interaction of Bridged Monobactams with Class C
â-Lactamases
with the bridged monobactams is attributable solely to
the denial of access to the water molecule. It does not
involve deactivation of the ester as a result of either
chemical rearrangement or moving the carbonyl oxygen
out of the oxyanion hole.
A rotation about the amide bond of the acyl side chain
by approximately 35°, compared to the aztreonam
complex, is observed in the inhibitor complexes. This
different vector is imposed by the pyrrolidine ring. It
has the consequence of placing the side chain deeper in
the active site. In turn, this restricts the options for
side chain variation to compounds that have small
substituents in the R-position and that are relatively
flexible. Indeed, bridged monobactams with rigid side
chains such as that of aztreonam have very low affinities
for the enzymes (Table 3).
The side chain of the inhibitor lies parallel to the
strand of the â-sheet that forms one edge of the catalytic
center. A hydrogen bond is possible between the car-
bonyl oxygen of the side chain and the amide nitrogen
of asparagine 152, but its stereochemistry is far from
ideal (Figure 2). There are two other interactions that
appear to influence side chain specificity and affinity.
First, there is the possibility to form a hydrogen bond
between a donor, such as the R-amino group of 21, and
the backbone carbonyl group of serine 318 in the
â-strand. This interaction appears to make a consider-
ably smaller contribution to the overall binding energy
than that involving asparagine 152 (Table 2). This may
be because the side chains of tyrosine 221, valine 211,
and threonine 319 provide a rather apolar protein
environment at this position that counteracts the favor-
able hydrogen bond interaction. The second possibility
is to form hydrophobic contacts with the aromatic ring
of tyrosine 221, which lies at the bottom of the side chain
binding pocket. The influence of this interaction was
clear in a series of compounds having simple aromatic
and aliphatic side chains (Table 2).
Optimization of the interactions described above
through modification of the acyl side chain increased
the affinity over 10000-fold, from a K2 value > 10 mM
(when none of the interactions are satisfied) to a K2
value of 120 nM. The stability of the acyl-enzyme
complex also increased (Table 4), possibly because
tighter binding of the inhibitor decreases the flexibility
of the acyl-enzyme complex and thus the rate of
occasional access of water to the ester.
F igu r e 1. Time course of the reaction of â-furylacrylamido-
bridged monobactam with â-lactamases. (a) Acylation of C.
freundii 1982 class C â-lactamase. The protein concentration
was 0.5 µM, and the inhibitor was present at 10 µM for
acylation (O) and at 0.6 µM for deacylation (b). Free enzyme
activity was determined by measuring residual nitrocefin
hydrolysis.25 The buffer was 0.1 M sodium phosphate, pH 7.0
at 37 °C. (b) Tracings of the absorption change at 310 nm
during the reaction of the inhibitor with P. aeruginosa 18 SH
class C â-lactamase (10 µM inhibitor), TEM-3 class A â-lac-
tamase (1 mM inhibitor), and B. licheniformis class A â-lac-
tamase (1 mM inhibitor). The enzymes were present at 0.1
µM; the buffer was 0.1 M sodium phosphate, pH 7.0 at 37 °C.
The reactions were started by mixing equal volumes of
inhibitor and enzyme solutions using an Applied Photophysics
RX 1000 rapid mixing spectrophotometer accessory.
Str u ctu r e of â-La cta m a se-In h ibitor Com p lexes.
The stability of the acyl-enzyme complexes formed with
C. freundii class C â-lactamase allowed solution of their
structures by X-ray crystallography (Figure 2). No
significant changes in protein structure occurred, except
that the side chain of aspartate 123 moved to accom-
modate the inhibitor bound to a neighboring protein
molecule. The conformation adopted by the inhibitor
moiety shows that the â-lactam ring has been opened
to form an ester with the catalytic serine residue. The
N-sulfonate moiety forms a salt bridge with lysine 315.
N1 blocks the trajectory that an incoming water mol-
ecule would have to use for attack on the ester, while
the sulfonate displaces the water from the position
where it can be activated. The carbonyl oxygen of the
ester occupies the “oxyanion hole”, forming good hydro-
gen bonds to the main chain NH groups of serine 70
and serine 318. In this position, the reactivity of the
ester toward attack by a water molecule should be
comparable to that of a rapidly hydrolyzed substrate.
Therefore, the stability of the acyl-enzymes formed
Rea ction w ith Cla ss A â-La cta m a ses. The class
A enzymes generally exhibited low affinity for the
bridged monobactams, although significant inhibition
was observed with some side chains (Table 3). At high
concentration, acylation of class A â-lactamases was also
rapid, but the rate of breakdown of the acyl-enzyme
was also considerably higher. Thus, lower steady-state
levels of occupancy of the active site and more rapid

Design of â-Lactamase Inhibitors. 1
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 21 3965
F igu r e 2. Binding mode of compound 23 in the active site of C. freundii â-lactamase. (a) Schematic representation. Hydrogen-
bonding interactions of the covalently bound inhibitor moiety with the enzyme and with ordered solvent molecules are indicated
by dashed lines (distances in Å). CR positions are emphasized by filled circles. (b) Stereoview showing the inhibitor (thick lines)
together with an omit electron density map (contour level 0.12 e/Å3) calculated using the final model with the inhibitor moieties
omitted. Hydrogen-bonding interactions as in panel a are indicated with broken lines.

3966 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 21
Ta ble 2. Preparation of Bridged Pyrrolidino Monobactams
Heinze-Krauss et al.
a
Exp refers to special cases described in the Experimental Section. Methods A-F refer to general procedures described in the
Experimental Section. Refers to a two-step procedure; see Experimental Section. c (M - Na)-. (M + H)+.
b
d
hydrolysis were observed (Figure 1), with kcat/KM e 104
M-1 s-1
Inspection of class A â-lactamase crystal
structures indicated that a water molecule in an equiva-
lent position would be much less activated because there
is a serine residue replacing the tyrosine of class C
enzymes.15,27 Attack in a class A enzyme occurs from
.
the opposite side of the ester, with activation of the
water molecule occurring through an extension of the
hydrogen bond network that is not present in class C
enzymes.27-30 Hence, rotation about C3-C4 is not so

Design of â-Lactamase Inhibitors. 1
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 21 3967
MICb (µg/mL) ceftriaxone:inhibitor ) 1:4
Ta ble 3. Enzyme Inhibition and in Vitro Synergy with Ceftriaxone
a
IC50 (nM)
C. freundii
P. aeruginosa
E. coli
C. freundii
P. aeruginosa
E. coli
compd
1982 (class C)
18 SH (class C)
TEM-3 (class A)
1982
18 SH
TEM-3
refc
8
128
64
>16
>16
2
128
64
>16
>16
8
8
8
16
4
4
4
8
128
8
8
8
128
8
16
16
16
16
8
220
65
13
500
220
109
10
613
1035
380
90
420
155
24
35
31
511
10
12
21
22
23
24
25
26
27
28
29
30
32
33
34
35
36
37
38
42
1d
>100000
NA
2
1
1
0.5
0.25
1
2
8
4
8
>100000
4000
16
16
16
16
16
16
16
8
16
8
16
8
54
12
217
48
>100000
100000
>100000
>100000
>100000
>100000
>100000
>100000
>100000
>100000
700
364
2200
100
225
85
122
283
50
23
6
3
5700
900
2246
3900
200
1500
200
612
853
110
27
55
25
145000
800
8
0.5
1
1
0.5
0.5
0.25
8
4
>100000
16
128
>16
>32
16
8
16
0.25
79200
>100000
15
>16
8
a
b
Inhibition of the isolated enzymes. With the strains that produce the â-lactamases. c MICs of ceftriaxone in the absence of any
d
inhibitor. Values obtained with tazobactam, reference inhibitor.
Ta ble 4. Parameters Influencing the Efficacy of Selected
Bridged Monobactams
showed no synergy apparently because of limited pen-
etration through the outer membrane (Table 4). Lim-
kinetic parameters
for P. aeruginosa
ited permeability also seems to be a limiting factor with
P. aeruginosa, as the magnitude of the synergy observed
with the monosubstituted derivatives (21-38) was not
as strong as expected. Several mechanisms might be
operating in this organism,31 but it seems that some
derivatives (e.g., 23 or 30) have difficulties in entering
the periplasmic space of P. aeruginosa (Table 4). When
this barrier could be overcome, as with Ro 48-1256 (28),
better synergy could be obtained.32 Penetration of the
outer membrane is not such a barrier in the Entero-
bacteriaceae, and the monosubstituted derivatives (21-
38) exhibited a strong synergy with ceftriaxone against
strains producing class C â-lactamases. With the
exception of compounds 29, 30, and 38, the MIC values
correlated rather well with the IC50 values.
relative
18 SH â-lactamasea
permeability (P)b
KS
(µM)
kon
koff
C. freundii P. aeruginosa
compd
(s-1
)
(10-5 s-1
)
1982
18 SH
8
21
23
28
30
36
42
15
6.9
50
1480
127
3.8
2.7
0.4
1.5
4.1
<1
10
68
8.7
0.6
1.7
NA
0.8
NA
8.2
38
1.0
24
1.2
NA
5.3
120
3.7
11
65
3500
0.9
0.5
0.9
70
a
Kinetic parameters for acylation and deacylation were deter-
mined by the methods described in ref 8. P is defined as the ratio
b
of the observed IC50 for â-lactamase inhibition in intact, growing
cells to the observed IC50 in the same cells treated with 0.7% final
concentration of octyl-poly(oxyethylene). This treatment results
in exposure of â-lactamase and IC50 values identical within
experimental error to those obtained with isolated enzyme in free
solution.
Con clu sion
The bridged monobactams are potent and selective
inhibitors of class C â-lactamases. The activity arises
from the formation of a stable acyl-enzyme complex
that blocks the enzymes. The synergy with â-lacta-
mase-sensitive antibiotics was less pronounced against
P. aeruginosa than expected and could be partly at-
tributed to a reduced penetration of the inhibitors
through the outer membrane. Strong synergism with
ceftriaxone and other â-lactam antibiotics was observed
against Enterobacteriaceae.
critical to the mechanism of deacylation in these en-
zymes, and restricting the rotation should have less
impact on the stability of the acyl-enzyme. Thus, these
observations are entirely consistent with the predictions
based on the crystal structures of representative en-
zymes from the two classes and highlight the differences
between the hydrolysis mechanisms of the two enzyme
classes.
In Vitr o Effica cy of th e In h ibitor s in Com bin a -
tion w ith â-La cta m An tibiotics. The inhibitors had
no intrinsic antibacterial activity (data not shown).
They were tested in combination with ceftriaxone, a
third-generation cephalosporin, against strains produc-
ing class A and C â-lactamases (Table 3). As expected
from the mechanism of the inhibitors, there was little
synergy against strains producing class A â-lactamases.
Surprisingly, despite relatively potent enzyme inhibi-
tion, the disubstituted derivatives (e.g., 8, 10, and 12)
Exp er im en ta l Section
Gen er a l Syn th etic Meth od s. Solvents were dried by
filtration through Al2O3 (neutral, Brockmann-number 1) when
necessary and stored over a bed of molecular sieves (3 Å). All
the organic solutions obtained after extraction were worked
up by washing with water and brine, dried over MgSO4 (unless
indicated otherwise), and evaporated under reduced pressure
with a water bath temperature below 35 °C. Chromatography
was performed using Merck silica gel 60 (particle size 40-63

3968 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 21
Heinze-Krauss et al.
µm). Water-soluble compounds were purified by gel filtration
(Mitsubishi Kasei Corp. MCI gel CHP20P 75-150 µm) using
a gradient of MeOH in H2O. The relevant fractions were
pooled and lyophilized. TLC was performed on Merck TLC
plates (silica gel 60 F254) and were visualized with Cl2 and a
0.5% aqueous solution of o-tolidine (in the presence of 2%
AcOH and 0.05% KI). Proton NMR spectra were recorded on
a Bruker AC250 (250 MHz) spectrophotometer. Chemical
shifts (δ) are reported in ppm relative to Me4Si as internal
standard, and J values are in hertz (Hz). IR spectrum were
recorded on a Nicolet FTIR spectrometer as KBr pellets (unless
indicated otherwise). Ion spray mass spectra were recorded
on a Finnigan MAT SSQ 7000 instrument, and EI mass
spectra were recorded on a Perkin-Elmer Siex API III instru-
ment. Optical rotation data were obtained on a Perkin-Elmer
241 polarimeter at 20 °C. Melting points were determined on
a Buchi 510 melting point apparatus and are uncorrected.
Elemental analyses are indicated by the symbols of the
elements; analytical results were within 0.4% of the theoretical
values.
warmed to 100 °C, and K2S2O8 (164 g, 0.6 mol) was added while
the pH was maintained at 4.5 with a saturated NaHCO3
solution. The mixture was cooled, saturated with NaCl, and
extracted with AcOEt. The aqueous layer was extracted with
AcOEt, and the combined organic phases were worked up,
chromatographed (AcOEt/hexane, 2:1, then AcOEt), and crys-
tallized giving 50.4 g (68%) of 7c, mp 59-60 °C (AcOEt/
hexane). [R]D -234°(CHCl3, c 1). IR: 1755, 1689 cm-1. NMR
(CDCl3): 1.69-2.03 (m, 2H); 3.42 (m, 1H); 4.1 (br, 1H); 4.32
(br, 1H); 5.05-5.35 (br, 3H); 6.08 (br, 1H); 7.36 (m, 5H). Anal.
(C13H14N2O3) C, H, N.
(d ) (1S,5R)-2,6-Dia za bicyclo[3.2.0]h ep ta n -7-on e (7). A
solution of 7c (13.7 g, 55.6 mmol) in MeOH (0.15 L) was
hydrogenated over Pd/C. After removal of the catalyst, the
filtrate was evaporated and crystallized under Ar leaving 4.5
g (72%) of 7, mp 148-49 °C (EtOH). [R]D -0.2°(EtOH, c 1).
IR: 1735, 1713 cm-1
. NMR (DMSO-d6): 1.24 (m, 1H); 1.68
(dd, J ) 5 and 13, 1H); 2.65 (m, 1H); 3.01 (br, 1H); 3.08 (dd, J
) 7.5 and 11.5, 1H); 3.93 (dd, J ) 3.5 and 5, 1H); 4.35 (dd br,
J ) 3.5 and 5, 1H); 7.74 (br, 1H). Anal. (C5H8N2O) C, H, N.
(1S,4R,5S)-2-(Ben zyloxyca r bon yl)-4-[(1-m eth yl-1H-tet-
r a zol-5-yl)su lfa n yl]-7-oxo-2,6-d ia za bicyclo[3.2.0]h ep ta n e-
6-su lfon ic Acid Sod iu m Sa lt (8). (a ) Ben zyl (1S,4S,5S)-
4-h y d r o x y -7-o x o -2,6-d i a z a b i c y c lo [3.2.0]h e p t a n e -2-
ca r boxyla te was prepared in 40% yield from 6 according to
the procedure described for 7c, mp 192-193 °C (MeOH). IR:
1764, 1720, 1660 cm-1. Anal. (C13H14N2O4) C, H, N.
(3S,4S)-[1-(3,4-Dim et h oxyb en zyl)-2-oxo-4-[(R)-2-oxo-
1,3-d ioxolola n -3-yl]a zetid in -3-yl]ca r ba m ic Acid Ben zyl
Ester (5). A solution of 422 (215.3 g, 0.5 mol) in THF (3 L)
and 1,1′-carbonyldiimidazole (121.6 g, 0.75 mol) was refluxed
for 4 h. The solvent was evaporated, and the residue was
dissolved in CH2Cl2, washed with 1 N HCl, and worked up
leaving 220.3 g (96%) of 5, mp 135-136 °C. IR: 1836, 1850,
1765, 1721, 1687 cm-1
.
MS (EI): 456 (M+). NMR (CDCl3):
(b) 1:1 Mixtu r e of (1S,4S,5S)-2-(Ben zyloxyca r bon yl)-
4-[(R)- an d (S)-tetr ah ydr opyr an -2-yloxy]-2,6-diazabicyclo-
[3.2.0]h ep ta n -7-on e. The previous alcohol 8a (40.0 g, 153
mmol) was treated with p-toluenesulfonic acid (2.9 g, 15 mmol)
and dihydropyran (28 mL, 305 mmol) in DMF (0.5 L). After
24 h the solvent was removed under reduced pressure, and
the residue was purified by chromatography (EtOAc/hexane,
3.68 (dd, br, 1H); 3.88 (s, 6H); 3.97-4.25 (m, br, 3H); 4.68-
4.81 (m, br, 2H); 4.97-5.20 (m, br, 3H); 5.87 (NH, d, br, 1H);
6.81 (s, 1H); 6.83 (s, 1H); 7.35 (5H). Anal. (C23H24N2O8) C,
H, N.
Ben zyl (1S,4S,5S)-6-(3,4-Dim eth oxyben zyl)-4-h yd r oxy-
7-oxo-2,6-d ia za bicyclo[3.2.0]h ep ta n e-1-ca r boxyla te (6). A
solution of 5 (220 g, 0.48 mol) and Bu4NBr (25.4 g, 0.12 mol)
in DMF (3 L) was vigorously stirred under argon for 5 h at
140 °C. The solvent was evaporated, and the residue was
dissolved in AcOEt, worked up, chromatographed (AcOEt/
hexane, 3:1, then AcOEt), and crystallized yielding 172 g (86%)
of 6, mp 100-102 °C (AcOEt). [R]D -172°(CHCl3, c 1). IR:
1731, 1707 cm-1. NMR (CDCl3): 2.24 (s br, 1H); 3.35 (dd, J
) 3.5 and 13, 1H); 3.86 (s, 3H); 3.87 (s, 3H); 3.91 (br, 3H);
4.21 (d, J ) 15, 1H); 4.40 (d, J ) 15, 1H); 5.13 (s, 2H); 5.26
(br, 1H); 6.76 (s, 1H); 6.81 (s, 3H); 7.32 (m, 5H). Anal.
(C22H24N2O6) C, H, N.
3:2). Yield: 40 g (76%). IR (neat): 1775, 1714 cm-1
. MS
(CI): 364 (M + NH4)+.
(c) 1:1 Mixtu r e of Ben zyl (1S,4S,5S)-6-(ter t-Bu tyld i-
m eth ylsila n yl)-4-[(R)- a n d (S)-tetr a h yd r op yr a n -2-yloxy]-
7-oxo-2,6-d ia za bicyclo[3.2.0]h ep ta n e-2-ca r boxyla te. A so-
lution of the lactam 8b (39.85 g, 115 mmol) in CH2Cl2 (0.8 L)
was treated with triethylamine (24.1 mL, 173 mmol) and tert-
butyldimethylchlorosilane (26.0 g, 173 mmol). After 24 h the
reaction mixture was worked up and chromatographed (EtOAc/
hexane, 3:2). Yield: 46.8 g (88%). IR (neat): 1757, 1711 cm-1
.
MS (CI): 478 (M + NH4)+.
Ben zyl (1S,5R)-7-Oxo-2,6-d ia za bicyclo[3.2.0]h ep ta n e-
2-ca r boxyla te (7). (a ) Ben zyl (1S,4S,5S)-6-(3,4-Dim eth -
oxyb en zyl)-7-oxo-4-[(t r iflu or om et h yl)su lfon yloxy]-2,6-
d ia za bicyclo[3.2.0]h ep ta n e-2-ca r boxyla te. A solution of
6 (206 g, 0.5 mol) and pyridine (48 mL, 0.6 mol) in CH2Cl2
(0.8 L) was reacted below 10 °C with (CF3SO2)2O (98 mL, 0.6
mol) for 2 h, then quenched with dilute NaHCO3 solution, and
worked up yielding 272 g (100%) of crude triflate. IR: 1771,
1712 cm-1. MS (EI): 544 (M+). NMR (CDCl3): 3.50 (br, 1H);
3.87 (s, 3H); 3.88 (s, 3H); 4.20 (br, 3H); 4.52 (br, 1H); 4.71 (br,
1H); 5.17 (br, 2H); 5.33 and 5.52 (br, 1H); 6.78 (s, 1H); 6.85
(m, s, 3H); 7.34 (m, 5H).
(b) Ben zyl (1S,5R)-6-(3,4-Dim eth oxyben zyl)-7-oxo-2,6-
d ia za bicyclo[3.2.0]h ep ta n e-2-ca r boxyla te. A solution of
the triflate 7a (272 g, 0.5 mol) in THF/DMF (1 L; 4:1) was
reacted at 0 °C overnight with NaBH4 (19 g, 0.5 mol), then
the mixture was adjusted to pH 4 with AcOH, and the solvents
were evaporated. The oily residue was poured into 1 N HCl,
extracted with AcOEt, worked up, chromatographed (AcOEt/
hexane, 1:1 to 4:1), and crystallized giving 120 g (60%) of
material, mp 76-78 °C (AcOEt/hexane). [R]D -187°(CHCl3, c
1). IR (film): 1755, 1705 cm-1. NMR (CDCl3): 1.46-1.79 (m,
2H); 3.24 (m, 1H); 3.87 (s, 3H); 3.88 (s, 3H); 4.0 (m, br, 1H);
4.18 (br, 1H); 4.32 (s, 2H); 5.06-5.31 (br, 3H); 6.8 (m, 3H);
7.35 (m, 5H). Anal. (C22H24N2O5) C, H, N.
(d ) Ben zyl (1S,4S,5S)-6-(ter t-Bu tyld im eth ylsiloxy)-4-
h yd r oxy-7-oxo-2,6-d ia za bicyclo[3.2.0]h ep ta n e-2-ca r boxy-
la te. The ether 8c (2.6 g, 5.64 mmol) was stirred with
pyridinium p-toluenesulfonate (140 mg, 0.56 mmol) in MeOH
(75 mL) at 50 °C for 24 h. The solvent was removed under
reduced pressure, and the residue was chromatographed
(EtOAc/hexane, 2:1) yielding 1.39 g (65%) of a colorless oil
which crystallized upon standing, mp 99.5-101 °C. Anal.
(C19H28N2O4Si) C, H, N.
(e) Ben zyl (1S,4S,5S)-6-(ter t-Bu tyld im eth ylsila n yl)-7-
oxo-4-[(t r iflu or om et h yl)su lfon yloxy]-2,6-d ia za b icyclo-
[3.2.0]h ep ta n e-2-ca r boxyla te. A solution of (CF3SO2)O (2.46
mL, 14.61 mmol) in CH2Cl2 (5 mL) was added dropwise at 0
°C to a solution of the alcohol 8d (5.0 g, 13.28 mmol) and
pyridine (1.18 mL, 4.61 mmol) in CH2Cl2 (50 mL). After 2 h
the reaction mixture was worked up with diluted NaHCO3
solution yielding 6.0 g (89%) of an orange oil. IR (neat): 1765,
1716 cm-1. MS (ISP): 526.4 (M + NH4+).
(f) Ben zyl (1S,4S,5S)-4-(1-Meth yl-1H-tetr a zol-5-ylsu l-
fa n y l)-7-o x o -2,6-d ia za b ic y c lo [3.2.0]h e p t a n e c a r b o x -
yla te. A solution of the triflate 8e (1.0 g, 1.96 mmol) in DMF
(20 mL) was treated under argon with sodium 5-mercapto-1-
methyltetrazole (hydrate; 600 mg, 2.16 mmol) at 80 °C for 48
h. The solvent was removed under pressure and the residue
taken up in EtOAc, worked up, and chromatographed (EtOAc).
Yield: 150 mg (21%). IR (film): 1769, 1708 cm-1. MS (ISP):
383.2 (M + Na)+.
(c) Ben zyl (1S,5R)-7-Oxo-2,6-diazabicyclo[3.2.0]h eptan e-
2-ca r boxyla te. The following procedure is representative for
the removal of the 3,4-dimethoxybenzyl protecting group. A
solution of 7b (117 g, 0.3 mol) in MeCN/H2O (3 L; 2:1) was
(1S,4R,5S)-2-(Ben zyloxyca r bon yl)-4-(1-m eth yl-1H-tet-
r a zol-5-ylsu lfa n yl)-7-oxo-2,6-d ia za bicyclo[3.2.0]h ep ta n e-

Design of â-Lactamase Inhibitors. 1
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 21 3969
6-su lfon ic Acid Sod iu m Sa lt (8). The following procedure
is representative for the sulfonation. A solution of 8f (1.7 g,
4.7 mmol) in DMF (75 mL) was cooled to 0 °C and reacted
with DMF‚SO3 complex (1.0 g, 7.1 mmol). After 4 h, the
solvent was evaporated under high vacuum. The residue was
taken up in water (3 mL), the pH was adjusted to pH 6 with
NaHCO3, and the solution was chromatographed by gel
filtration. Yield: 10%.
under reduced pressure and the residue taken up in EtOAc,
worked up, and chromatographed (EtOAc) yielding 700 mg
(92%) of a yellowish oil. IR (neat): 1764, 1708 cm-1
. MS
(ISP): 437.2 (M + Na)+.
(c) Ben zyl (1S,4S,5S)-4-flu or o-7-oxo-2,6-d ia za bicyclo-
[3.2.0]h ep ta n e-2-ca r boxyla te was prepared in 69% yield
from 12b according to the procedure described for the prepara-
tion of 7c. IR (neat): 1772, 1707 cm-1. MS (ISN): 263.1 (M
- H)-.
ter t-Bu tyl (1S,5R)-7-Oxo-2,6-diazabicyclo[3.2.0]h eptan e-
2-ca r boxyla te (15; Meth od C). A solution of 7 (11.2 g, 0.1
mol), (BOC)2O (22.9 g, 0.105 mol), and DMAP (1.2 g, 10 mmol)
in THF (0.25 L) was stirred at room temperature for 3 h. The
solvent was evaporated, and the residue was chromatographed
(AcOEt) and crystallized leaving 16.4 g (77%) of 15, mp 125-
126 °C (ether/hexane). NMR (DMSO-d6): 1.40 (s, 9H); 1.60
(m, 1H); 1.80 (m, 1H); 3.20 (m, 1H); 3.90 (m, 1H); 4.15 (t, J )
5, 1H); 4.90 (br, 1H); 8.10 (br, 1H). Anal. (C10H16N2O3) C, H,
N.
(1S,5R)-7-Oxo-2,6-d ia za bicyclo[3.2.0]h ep ta n e-6-su lfon -
ic Acid (13). A solution of 15 (18 g, 85 mmol) in anhydrous
DMF (0.2 L) was reacted below 5 °C with DMF‚SO3 (14.3 g,
93.5 mmol) for 3 h. The solvent was evaporated, and the
residue was treated with water (0.1 L), stirred overnight, and
then concentrated to about 15 mL. The crystals (9.4 g, 57%)
were collected by filtration. IR: 1778, 1614 cm-1. MS (ISN):
191 (M - H)-. NMR (DMSO-d6): 1.74 (m, 1H); 2.46 (dd, J )
6 and 14, 1H); 3.05 (m, 1H); 3.61 (dd, J ) 7.5 and 14, 1H);
4.43 (t, J ) 5, 1H); 5.21 (d, J ) 5, 1H); 9.50 (br, 3H). Anal.
(C5H8N2O4S) C, H, N.
(1S ,5R )-2-[D -2-(4-H yd r oxyp h e n yl)glycyl]-7-oxo-2,6-
d ia za bicyclo[3.2.0]h ep ta n e-6-su lfon ic Acid (21; Meth od
A). (a ) ter t-Bu t yl [(R)-4-H yd r oxy-r-[[(1S,5R)-7-oxo-6-
su lfo-2,6-d ia za bicyclo[3.2.0]h ep t-2-yl]ca r bon yl]ben zyl]-
ca r ba m a te Tr ieth yla m m on iu m Sa lt. A solution of 13 (0.8
g, 2.6 mmol) in DMF (30 mL) was reacted with TEA (0.53 g,
5.2 mmol) and (tert-butyloxycarbonyl)-D-(4-hydroxyphenyl)-
glycine hydroxysuccinimide (1.05 g, 2.85 mmol) for 4 h. The
solvent was evaporated under reduced pressure, and the
residue was purified by gel filtration giving 0.41 g (30%) of
material. IR: 3413, 1770, 1708, 1651 cm-1. MS (FAB): 439.8
(M)-.
(b) (1S,5R)-2-[D-2-(4-Hyd r oxyp h en yl)glycyl]-7-oxo-2,6-
d ia za bicyclo[3.2.0]h ep ta n e-6-su lfon ic Acid (21). The es-
ter 21a (0.4 g, 0.74 mmol) was reacted with TFA (9 mL) at
-10 °C for 1 h. The solvent was evaporated, and the residue
was purified by gel filtration giving 0.15 g (60%) of 21. IR:
1766, 1660, 1614, 1517 cm-1. MS (FAB): 340 (M - Na)-.
(1S,5R)-2-(4-Ca r b a m oylp h en ylca r b a m oyl)-7-oxo-2,6-
d ia za bicyclo[3.2.0]h ep ta n e-6-su lfon ic Acid Sod iu m Sa lt
(23; Meth od B). The described methods are general for the
preparation of compounds 22-27 as well as their required
activated side chains.
(a ) 4-Ca r ba m oylp h en ylca r ba m ic Acid 2,5-Dioxop yr -
r olid in -1-yl Ester . A solution of 4-aminobenzamide (2.72 g,
2 mmol) and N,N′-disuccinimidylcarbonate (5.6 g, 2.2 mol) in
MeCN (0.25 L) was stirred for 5 h. The crystals were collected
by filtration leaving 4.83 g (87%) of colorless material. NMR
(DMSO-d6): 2.84 (s, 4H); 7.30 (b, 1H); 7.50 (d, J ) 9, 1H); 7.88
(d, J ) 9, 1H); 7.90 (b, 1H). Anal. (C12H11N3O5) C, H, N.
(b ) (1S,5R)-2-(4-Ca r b a m oylp h en ylca r b a m oyl)-7-oxo-
2,6-d ia za bicyclo[3.2.0]h ep ta n e-6-su lfon ic Acid Sod iu m
Sa lt (23). A solution of 13 (0.15 g, 0.78 mmol) in H2O/MeCN
(1:1; 10 mL) was reacted with NaHCO3 (66 mg, 0.78 mmol)
and the above-described side chain 23a (216 mg, 0.78 mmol)
for 5 h. The organic solvent was evaporated, and the crude
aqueous phase was purifed by gel filtration giving 230 mg
(78%) of 23. NMR (DMSO-d6): 1.75 (m, 1H); 3.33 (dd, J ) 6
and 15, 1H); 3.12 (m, 1H); 4.00 (dd, J ) 9 and 12, 1H); 4.43 (t,
J ) 5, 1H); 5.28 (d, J ) 5, 1H); 7.10 (br, 1H); 7.56 and 7.77
(2d, J ) 9, 4H); 7.8 (b, 1H); 8.80 (s, 1H). Anal. (C13H13N4O6-
SNa) C, H, N.
Sta r tin g Ma ter ia l for th e P r ep a r a tion of (1S,5R)-(E)-
2-(Ben zyloxyca r b on yl)-4-(ca r b a m oylm et h ylen e)-7-oxo-
2,6-d ia za bicyclo[3.2.0]h ep ta n esu lfon ic Acid Sod iu m Sa lt
(10). (a ) Ben zyl (1S,5S)-6-(ter t-Bu tyld im eth ylsila n yl)-4,7-
d ioxo-2,6-d ia za bicyclo[3.2.0]h ep ta n e-2-ca r boxyla te (9). A
solution of DMSO (2.08 g, 26.56 mmol) in CH2Cl2 (40 mL) was
treated at -78 °C with (CF3CO)2O (4.18 g, 19.92 mmol). After
15 min, a solution of 8d (5.0 g, 13.28 mmol) in CH2Cl2 (45 mL)
was added dropwise and the mixture was stirred for another
1.5 h. N-Ethyldiisopropylamine (5.12 g, 39.64 mmol) was
added, and the mixture was allowed to reach room tempera-
ture and was worked up with 0.5 N HCl and saturated
NaHCO3 solution yielding quantitatively the product as a
yellowish oil. IR (neat): 1760, 1712 cm-1
.
(b) 1:1 Mixtu r e of (E)- a n d (Z)-Ben zyl (1S,5R)-6-(ter t-
Bu t yld im et h ylsila n yl)-4-(ca r b a m oylm et h ylen e)-7-oxo-
2,6-d ia za bicyclo[3.2.0]h ep ta n e-2-ca r boxyla te. The ketone
10a (2.43 g, 6.5 mmol) and H2NCOCHdP(Ph)3 (2.28 g, 7.14
mmol) were refluxed in C6H6 (60 mL) for 30 min. The solvent
was evaporated and the residue was purified by chromatog-
raphy (EtOAc) yielding quantitatively an E/ Z (1:1) mixture
of the product. IR (neat): 1745, 1692, 1658 cm-1. MS (EI):
358 (M - tBu).
(c) Ben zyl (E)-(1S,5R)-4-(Ca r ba m oylm eth ylen e)-7-oxo-
2,6-d ia za bicyclo[3.2.0]h ep ta n e-2-ca r boxyla te. The lactam
10b (1.5 g, 3.6 mmol) was deprotected by reaction with NBu4F
(133 mg, 3.6 mmol) in MeOH (25 mL) for 20 min. The solvent
was removed under reduced pressure, and the residue taken
up in EtOAc was worked up and chromatographed (EtOAc/
EtOH, 3:2). Yield: 900 mg (83%). IR: 1720, 1653 cm-1. Anal.
(C15H15N3O4) C, H, N.
Ben zyl (1S,4R,5S)-6-(3,4-Dim eth oxyben zyl)-4-h yd r oxy-
7-oxo-2,6-d ia za bicyclo[3.2.0]h ep ta n e-2-ca r boxyla te (11).
(a ) Ben zyl (1S,5S)-6-(3,4-Dim eth oxyben zyl)-4,7-d ioxo-2,6-
d ia za bicyclo[3.2.0]h ep ta n e-2-ca r boxyla te. A solution of
DMSO (3.35 mL, 47 mmol) in CH2Cl2 (50 mL) was treated at
-78 °C with (CF3CO)2O (4.9 mL, 35.3 mmol). After 15 min, a
solution of 6 (9.7 g, 23.5 mmol) in CH2Cl2 (50 mL) was added
dropwise, and the mixture was stirred for another 1.5 h.
N-Ethyldiisopropylamine (12.1 mL, 70.5 mmol) was added, and
the mixture was allowed to reach room temperature and was
worked up with 0.5 N HCl and saturated NaHCO3 solution
yielding quantitatively the product as a yellowish oil. IR
(neat): 1760, 1709 cm-1. MS (EI): 410 (M).
(b) Ben zyl (1S,4R,5S)-6-(3,4-Dim eth oxyben zyl)-4-h y-
d r oxy-7-oxo-2,6-d ia za b icyclo[3.2.0]h ep t a n e-2-ca r b oxy-
la te (11). A solution of 11a (1.8 g, 4.4 mmol) in EtOH (50
mL) was treated with NaBH4 (166 mg, 4.4 mmol) at 0 °C. The
reaction mixture was stirred overnight at room temperature
and then acidified with AcOH. After removal of the solvent,
the residue was taken up in CH2Cl2, worked up, and chro-
matographed (AcOEt) yielding 1.2 g (66%) of 11, mp 118-120
°C. Anal. (C22H24N2O6) C, H, N.
Sta r tin g Ma ter ia l for th e P r ep a r a tion of Ben zyl (1S,-
4S ,5S )-4-F lu or o-7-oxo-6-su lfo-2,6-d ia za b icyclo[3.2.0]-
h ep ta n e-2-ca r boxyla te Sod iu m Sa lt (12). (a ) Ben zyl
(1S,4R,5S)-6-(3,4-d im eth oxyben zyl)-7-oxo-4-[(tr iflu or om -
eth yl)su lfon yloxy]-2,6-d ia za bicyclo[3.2.0]h ep ta n e-2-ca r -
boxyla te was prepared from 11 in 58% yield according to the
procedure described for the derivative 8e. IR (KBr): 1772,
1713 cm-1. MS (EI): 544 (M+).
(b) Ben zyl (1S,4S,5S)-4-F lu or o-6-(3,4-d im eth oxyben -
zyl)-7-oxo-2,6-d ia za b icyclo[3.2.0]h e p t a n e -2-ca r b oxy-
la te. A solution of the triflate 12a (1.1 g, 2.02 mmol) in THF
(10 mL) was treated at room temperature with NBu4F‚3H2O
(704 mg, 2.23 mmol) for 10 min. The solvent was removed
(1S,5R)-7-Oxo-2-(p ip er id in -4-ylca r b a m oyl)-2,6-d ia za -
bicyclo[3.2.0]h ep ta n e-6-su lfon ic Acid Sod iu m Sa lt (28).

3970 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 21
Heinze-Krauss et al.
(a ) (1S,5R)-2-(1-Ben zylp ip er id in -4-ylca r ba m oyl)-7-oxo-
2,6-d ia za bicyclo[3.2.0]h ep ta n e-6-su lfon ic a cid sod iu m
sa lt was prepared in 84% yield from 13 and 1-benzylpiperidin-
4-ylcarbamic acid 2,5-dioxopyrrolidin-1-yl ester. IR: 1759,
1638 cm-1. MS (ISN): 407 (M - Na)-.
(b) A solution of 28a (90 mg, 0.21 mmol) in water (10 mL)
was hydrogenated over Pd/C in the presence of 1 N HCl (1
mL) at 50 °C. The catalyst was removed by filtration, and
the pH of the solution was adjusted to pH 6 with saturated
NaHCO3 solution. The resulting crude aqueous phase was
purifed by gel filtration giving 50 mg (70%) of 28.
3H); 3.61 (m, 1H); 3.88 (s, 6H); 4.13 (m, 2H); 4.16 and 4.55
(2d, J ) 16, 2H); 5.0 (m, 1H); 5.24 (d, J ) 8, 1H); 6.75-6.80
(m, 3H).
(b )
(1S ,6R )-7-(3,4-D im e t h o x y b e n zy l)-8-o x o -2,7-
d ia za b icyclo[4.2.0]oct a n e-2-ca r b oxylic Acid ter t-Bu t yl
Ester . A solution of the mesylate 42a (1.0 g, 2.1 mmol) in
THF (20 mL) was reacted with NaH (0.1 g, 2.1 mmol; 50%
dispersed in oil) for 1 h. The reaction mixture was quenched
with saturated NH4Cl solution and extracted with AcOEt. The
organic solution was worked up and chromatographed (AcOEt/
hexane, 1:1) leaving 0.7 g (87%) of material. MS (EI): 321
(M - (CHdCMe2)). NMR (CDCl3): 1.46 and 1.47 (2s, 9H);
1.59-1.80 (m, 4H); 3.3-3.55 (m, 2H); 3.88 (s, 6H); 3.95 (m,
1H); 4.12 and 4.45 (2d, J ) 16, 2H); 5.12 and 5.38 (d, J ) 8,
1H); 6.80 (s, 3H).
(c) ter t-Bu tyl (1S,6R)-8-oxo-2,7-d ia za bicyclo[4.2.0]-
octa n e-2-ca r boxyla te was prepared in a similar way to 7c
in 55% yield from 42b, mp 146-48 °C (ether). IR: 1755, 1663
cm-1. MS (EI): 183 (M - HNCO). Anal. (C11H18N2O3) C, H,
N.
(1S,5R)-2-(Br om oa cetyl)-2,6-d ia za bicyclo[3.2.0]h ep ta n -
7-on e (16; Meth od D). A solution of 7 (6.0 g, 53.5 mmol) and
N-methyl-N-(trimethylsilyl)trifluoroacetamide (19.8 mL, 107
mmol) in CH2Cl2 (0.2 L) was cooled to 0 °C and reacted with
pyridine (4.3 mL, 54 mmol) and bromoacetyl bromide (5.15 mL,
59 mmol) for 2 h. The reaction mixture was worked up and
chromatographed leaving 9.3 g (75%) of 16. NMR (DMSO-
d6): 1.73 (m, 1H); 1.89 (m, 1H); 3.21 and 3.40 (2 m, 1H); 3.95-
4.40 (m, 4H); 5.21 and 5.26 (br, 1H).
(d ) ter t-Bu tyl (1S,6R)-8-oxo-7-su lfo-2,7-d ia za bicyclo-
[4.2.0]octa n e-2-ca r boxyla te sod iu m sa lt (42) was prepared
in a similar way to 13 in 85% yield after gel filtration. IR:
1757, 1688, 1408 cm-1. MS (FAB): 304.9 (M - Na)-. NMR
(DMSO-d6): 1.40 (s, 9H); 1.40-1.80 (m, 4H); 2.23 (db, J ) 12,
1H); 3.10-3.25 (m, 1H); 4.17 (b, 1H); 4.91 and 5.02 (2d, J ) 6,
1H). Anal. (C11H17N2O6SNa) C, H, N.
In Vitr o An tiba cter ia l Activity. The minimal inhibitory
concentrations (MICs) of the test compound were determined
according to a standard method37 by a serial 2-fold dilution
method using Mueller-Hilton broth (Difco Laboratories, De-
troit, MI). The inocculum size was approximately 105 cfu/mL.
The MIC of a compound or a combination was defined as the
lowest concentration that prevented visible growth of bacteria
after incubation at 37 °C for 18 h.
Sta r tin g Ma ter ia l for th e P r ep a r a tion of (1S,5R)-2-[[(1-
M e t h y l -1 H -t e t r a z o l -5 -y l ) t h i o ] a c e t y l ] -7 -o x o -2 ,6 -
d ia za bicyclo[3.2.0]h ep ta n e-6-su lfon ic Acid Sod iu m Sa lt
(32; Meth od E). (1S,5R)-2-[[(1-Meth yl-1H-tetr a zol-5-yl)-
th io]a cetyl]-2,6-d ia za bicyclo[3.2.0]h ep ta n -7-on e (17). A
solution of 1-[(methyl-1H-tetrazol-5-yl)thio]acetic acid (940 mg,
5.35 mmol) and TEA (0.75 mL, 5.35 mmol) in DMF (30 mL)
was sequentially treated with HBTU (2.0 g, 5.35 mmol) at 20
°C and 7 (500 mg, 4.46 mmol). The mixture was stirred for 5
h. The solvent was evaporated under reduced pressure, and
the residue was chromatographed (EtOAc/MeOH, 10:1 to 5:1).
Yield: 840 mg (70%). IR: 1770, 1633 cm-1; MS (EI): 225 (M
- CONH). This method is representative for the synthesis of
compounds 18-20. Yields and physical properties are given
in Table 1.
Str u ctu r e Deter m in a tion . The inhibitor complex was
obtained by soaking crystals of the C. freundii enzyme15 for
1.5 h in 3 mM inhibitor solution in 50% saturated potassium
phosphate buffer, pH 8.0. Data were collected at room
temperature with a Nicolet/Siemens area detector using Cu
KR radiation from an Elliott GX21 rotating anode X-ray
generator operated at 40 kV, 80 mA with a 0.3-mm focal spot
and a graphite monochromator. Rotation frames of 0.15° were
for two different crystal orientations. The frames were
processed and scaled using the XDS software;33 100 963
observations in the resolution range 20-2.1 Å yielded 39 500
unique reflections (89% completeness) with a scaling R-factor
of 3.8% on intensities. The refined cell constants (space group
P212121) are 97.96, 84.63, and 89.92 Å for a, b, and c. Analysis
and refinement of the structure were performed with X-Plor
version 3.134 and Moloc35 on Indigo 2 workstations. Refine-
ment started from the refined model determined for the crystal
structure of the uncomplexed enzyme15 with water molecules
omitted and all B-values set to 25 Å2. Rigid body refinement
of the two molecules of the noncrystallographic dimer using
data from 10-2.5 Å resolution reduced the R-factor from an
initial 32.2% to 30.6% (free R from 33.0% to 31.7%). In the
first round of positional refinement using data from 20.0-2.2
Å (36 289 reflections, 94% complete) the R-factor was de-
creased from an initial 31.4% to 25.5% (free R from 32.0% to
29.4%). A subsequent electron density difference map showed
well-defined density for the inhibitor moieties in both mol-
ecules, and they were built and added to the model. Several
rounds of positional and thermal parameter (restrained,
isotropic) refinement and manual model rebuilding reduced
the R-factor to 16.4% (free R 21.2%). A bulk solvent correction
as available in X-Plor was applied throughout the refinement.
During manual model building, the conformations of some side
chains were changed and 252 well-defined solvent sites were
gradually added to the model. The final model contained 2807
non-hydrogen atoms for each of the two complete protein
chains (residues 1-361, average B-factors of 26.6 and 23.7 Å2
for chains A and B, respectively), 22 atoms for each inhibitor
moiety (average B-factor of 25.3 and 30.5 Å2 for molecules A
(1S,5R)-2-[[(Ca r b a m oylp yr id in yl)t h io]a cet yl]-7-oxo-
2,6-d ia za bicyclo[3.2.0]h ep ta n e-6-su lfon a te (36; Meth od
F ). A solution of nicotinamide (122 mg, 1 mmol) and 30 (335
mg, 1 mmol) in DMF (4 mL) was stirred for 3 days. The
solvent was evaporated under reduced pressure, and the
residue was purified by gel filtration leaving 153 mg (43%) of
a white powder.
(3S,4S)-[1-(3,4-Dim eth oxyben zyl)-2-oxo-4-(3-oxop r op e-
n yl)a zetid in -3-yl]ca r ba m ic Acid Ben zyl Ester (40).
A
solution of the aldehyde 3922 (18.6 g, 46.6 mmol) in THF (0.2
L) was reacted with P(Ph)3CHCHO (15.6 g, 51.2 mmol) for 20
h. The solvent was evaporated, and the residue was taken
up in AcOEt, worked up, chromatographed (AcOEt/hexane,
1:1), and crystallized (AcOEt/hexane/ether) leaving 13.8 g
(70%) of 40. IR: 1771, 1685 cm-1
.
(3S,4S)-[1-(3,4-Dim eth oxyben zyl)-2-(3-h yd r oxyp r op yl)-
4-oxoa zetid in -3-yl]ca r ba m ic Acid ter t-Bu tyl Ester (41).
A solution of 40 (4.24 g, 10 mmol) in EtOH (0.12 L) was
reduced with NaBH4 (135 mg, 3.6 mmol) for 1 h and then
hydrogenated over 10% Pd/C. At the end of the reaction the
catalyst was filtered off and washed with EtOH. The filtrate
was reacted with (BOC)2O (2.18 g, 10 mmol) in the presence
of DMAP. The solvent was evaporated, and the residue was
taken up in AcOEt, worked up, and chromatographed (AcOEt)
giving 1.1 g (28%) of 41. MS (EI): 395.1 (M + H)+. NMR
(CDCl3): 1.43 (s, 9H); 1.46-1.80 (m, 4H); 3.57 (t br, 2H); 3.65
(m, 1H); 3.87 (s, 6H); 4.16 and 4.55 (2d, J ) 16, 2H); 5.0 (m,
1H); 5.25 (d, J ) 8, 1H); 6.75-6.80 (m, 3H).
ter t-Bu tyl (1S,6R)-8-Oxo-7-su lfo-2,7-d ia za bicyclo[4.2.0]-
octa n e-2-ca r boxyla te Sod iu m Sa lt (42). (a ) (3S,4S)-[1-
(3,4-Dim eth oxyben zyl)-2-[3-(m eth ylsu lfon yloxy)p r op yl]-
4-oxoa zetid in -3-yl]ca r ba m ic Acid ter t-Bu tyl Ester .
A
solution of 41 (4.24 g, 10 mmol) in CH2Cl2 (50 mL) was reacted
at room temperature with MeSO2Cl (0.26 mL, 3.3 mmol) and
TEA (0.5 mL, 3.6 mmol) for 2 h. The reaction mixture was
worked up and chromatographed (AcOEt/hexane, 1:1 to 8:2)
leaving 1.0 g (76%) of intermediate mesylate. MS (EI): 473
(M + H)+. NMR (CDCl3): 1.43 (s, 9H); 1.66 (m, 4H); 2.97 (s,

Design of â-Lactamase Inhibitors. 1
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 21 3971
(18) Dubus, A.; Normark, S.; Kania, M.; Page, M. G. P. The Role of
Tyrosine 150 in Catalysis of â-Lactam Hydrolysis by AmpC
â-Lactamase from Escherichia coli Investigated by Site-Directed
Mutagenesis. Biochemistry 1994, 33, 8577-8586.
(19) Dubus, A.; Normark, S.; Kania, M.; Page, M. G. P. Role of
Asparagine 152 in Catalysis of â-Lactam hydrolysis by Escheri-
chia coli AmpC â-Lactamase Studied by Site-Directed Mutagen-
esis. Biochemistry 1995, 34, 7757-7764.
(20) Matsuda, K.; Sanada, M.; Nakagawa, S.; Inoue, M.; Mitashashi,
S. Preferential Hydrolysis of cis Configuration Compounds at
the 3,4 Position of Monobactams by â-Lactamase from Mor-
ganella morganii. Antimicrob. Agents Chemother. 1991, 35, 458-
461.
(21) Hubschwerlen, C.; Schmid, G. An Enantioselective â-Lactam
Synthesis Starting from L-(S)-Glyceraldehyde Acetonide. Helv.
Chim. Acta 1983, 66, 2206-2209.
and B, respectively), and 252 solvent sites (average B-factor
31.0 Å2). Deviations from stereochemical ideality are 0.007 Å
for bond lengths and 1.70° for bond angles using reference
values.36
Ack n ow led gm en t. The authors would like to ex-
press their thanks to Drs. Arnold, W. Huber, and S.
Mueller, Mr. W. Meister, and A. Bubendorf for spectral
data and microanalysis. We also wish to thank M. P.
Imhoff, D. Granjean, J . Nell, I. Pfister, G. Schnider, and
H. Straub for their valuable technical assistance.
Refer en ces
(22) Yoshino, T.; Inaba, S.; Komura, H.; Ishido, Y. Synthetic Studies
by the Use of Carbonates. I. The Nucleophilic Substitution
Reaction of Ethylene Carbonate with Amine Hydrohalides. Bull.
Chem. Soc. J pn. 1974, 47, 405-409.
(23) Komura, H.; Yoshino, T.; Ishido, Y. Synthetic Studies by the Use
of Carbonates. VII. New Method for the Synthesis of Aldityl
Derivatives of Heterocyclic Compounds. Carbohydr. Res. 1973,
31, 154-156.
(24) Mancuso, A. J .; Swern, D. Activated Dimethyl Sulfoxide: Useful
Reagents for Synthesis. Synthesis 1981, 165-185.
(25) Page, M. G. P. The Kinetics of Nonstoichiometric Bursts of
â-Lactam Hydrolysis Catalyzed by Class C â-Lactamases. Bio-
chem. J . 1993, 295, 295-304.
(26) Waley, S. G. The Kinetics of Substrate-induced Inactivation.
Biochem. J . 1991, 279, 87-94.
(27) Moews, P.; Knox, J . R.; Dideberg, O.; Charlier, P.; Frere, J .-M.
â-Lactamase of Bacillus licheniformis 744/C at 2 Å Resolution.
Proteins: Struct. Func. Genet. 1990, 7, 156-171.
(1) J arvis, W. R.; Martone, W. J . Predominant pathogens in hospital
infections. J . Antimicrob. Chemother. 1992, 29A, 19-24.
(2) J ones, R. N.; Kehrberg, E. N.; Erwin, M. E.; Anderson, S. C.
Prevalence of Important Pathogens and Antimicrobial Activity
of Parenteral Drugs at Numerous Medical Centers in the United
States. I. Study on the Threat of Emerging Resistances: Real
of Perceived? Diagn. Microbiol. Infect. Dis. 1994, 19, 203-215.
(3) Murray, B. E. New Aspects of Antimicrobial Resistance and the
Resulting Therapeutic Dilemmas. J . Infect. Dis. 1991, 163,
1185-1194.
(4) Travis, J . Reviving the Antibiotic Miracle? Science 1994, 264,
360-362.
(5) Sanders, C. C.; Sanders, W. E. J . â-Lactam Resistance in Gram-
Negative Bacteria: Global Trends and Clinical Impact. Clin.
Infect. Dis. 1992, 15, 824-839.
(6) Davies, J . Inactivation of Antibiotics and the Dissemination of
Resistance Genes. Science 1994, 264, 375-382.
(7) Poernuli, K. J .; Rodigo, G.; Dornbusch, K. Production of a
Plasmid Mediated AmpC-like â-Lactamase by
a Klebsiella
(28) Herzberg, O.; Moult, J . Bacterial Resistance to â-Lactam Anti-
biotics: Crystal Structure of â-Lactamase from Staphylococcus
aureus PC1 at 2.5 Å Resolution. Science 1987, 236, 694-701.
(29) Strynadka, N. C. J .; Adachi, H.; J ensen, S. E.; J ohns, K.; Sielecki,
A.; Betzel, C.; Sutoh, K.; J ames, M. N. G. Molecular Structure
of the Acyl-Enzyme Intermediate in â-Lactam Hydrolysis at 1.7
Å Resolution. Nature 1992, 359, 700-705.
pneumoniae Septicaemia Isolate. J . Antimicrob. Chemother.
1994, 34, 943-954.
(8) Livermore, D. M. Determinants of the Activity of â-Lactamase
Inhibitor Combinations. J . Antimicrob. Chemother. 1993, 21
(Suppl. A), 9-21.
(9) Fisher, J .; Charnas, R. L.; Knowles, J . R. Kinetic Studies of the
Inactivation on the Inactivation of Escherichia coli RTEM
â-Lactamase by Clavulanic Acid. Biochemistry 1978, 17, 2180-
2184.
(30) J elsch, C.; Mourey, L.; Masson, J .-M.; Samama, J .-P. Crystal
structure of Escherichia coli TEM1 â-Lactamase at 1.8
Å
(10) Charnas, R. L.; Fisher, J .; Knowles, J . R. Chemical Studies on
the Inactivation of Escherichia coli RTEM â-Lactamase by
Clavulanic Acid. Biochemistry 1978, 17, 2185-2189.
(11) Chen, C. C. H.; Herzberg, O. Inhibition of â-Lactamase by
Clavulanate. Trapped Intermediates in Cryocrystallographic
Studies. J . Mol. Biol. 1992, 224, 1103-1113.
(12) Brenner, D. G.; Knowles, J . R. Penicillanic Acid Sulfone: An
Unexpected Isotope Effect in the Interaction of 6R- and 6â-
Monodeuterio and of 5,6-Dideuterio Derivatives with RTEM
â-Lactamase from Escherichia coli. Biochemistry 1981, 20,
3680-3687.
(13) Kemal, C.; Knowles, J . R. Penicillanic Acid Sulfone: Interaction
with RTEM â-Lactamase from Escherichia coli at Different pH
Values. Biochemistry 1981, 20, 3688-3695.
(14) Richter, H. G. F.; Angehrn, P.; Hubschwerlen, C.; Kania, M.;
Page, M. G. P.; Specklin, J .-L.; Winkler, F. K. Design, Synthesis,
and Evaluation of 2â-Alkenyl Penam Sulfone Acids as Inhibitors
of â-Lactamases. J . Med. Chem. 1996, 39, 3712-3722.
(15) Oefner, C.; D’Arcy, A.; Daly, J . J .; Gubernator, K.; Charnas, R.
L.; Heinze, I.; Hubschwerlen, C.; Winkler, F. K. Refined Crystal
Structure of â-Lactamase from Citrobacter freundii Indicates a
Mechanism for â-Lactam Hydrolysis. Nature 1990, 343, 284-
288.
(16) Ujiie, K.; Nishida, N.; Sawai, T. Role of Lysine 67 in the Active
Site of Class C â-Lactamase from Citrobacter freundii GN346.
Eur. J . Biochem. 1990, 188, 15-22.
(17) Tsukamato, K.; Nishida, N.; Tsuruoka, M.; Sawai, T. Function
of the Conserved Triad Residues in the Class C â-Lactamase
from Citrobacter freundii GN346. FEBS Lett. 1990, 271, 243-
246.
Resolution. Proteins: Struct. Funct. Genet. 1993, 16, 364-383.
(31) Li, X.-Z.; Ma, D.; Livermore, D. M.; Nikaido, H. Role of Efflux
Pumps in Intrinsic Resistance of Pseudomonas aeruginosa:
Active Efflux as a Contributing Factor to â-Lactam Resistance.
Antimicrob. Agents Chemother. 1994, 38, 1742-1752.
(32) Livermore, D. M.; Chen, H. Y. Potentiation of â-Lactams against
Pseudomonas aeruginosa strains by Ro 48-1256, a bridged
monobactam inhibitor of AmpC â-lactamases. J . Antimicrob.
Chemother. 1997, 40, 335-343.
(33) Kabsch, W. Evaluation of Single-Crystal X-ray Diffraction Data
from a Position-sensitive Detector. J . Appl. Crystallogr. 1988,
21, 916-924.
(34) Bruenger, A. T.; Kuriyan, J .; Karplus, M. Crystallographic R
Factor Refinement by Molecular Dynamics. Science 1987, 235,
458-460.
(35) Gerber, P. R.; Mu¨ller, K. MAB: a Generally Applicable Molecular
Force Field for Structure Modelling in Medicinal Chemistry. J .
Comput.-Aided Mol. Des. 1995, 9, 251-268.
(36) Engh, R. A.; Huber, R. Accurate bond and angle parameters for
X-ray protein structure refinement. Acta Crystallogr. 1991, A47,
392-400.
(37) Heinze-Krauss, I.; Angehrn, P.; Guerry, P.; Hebeisen, P.; Hub-
schwerlen, C.; Kompis, I.; Page, M. G. P.; Richter, H. G. F.;
Runtz, V.; Stalder, H.; Weiss, U.; Wei, C.-C. Synthesis and
Structure-Activity Relationship of (Lactamylvinyl)cephalospor-
ins Exhibiting Activity against Staphylococci, Pneumococci, and
Enterococci. J . Med. Chem. 1996, 39, 1864-1871.
J M980023C