Structure-based design of potent and ligand-efficient inhibitors of CTX-M class A β-lactamase

Article abstract of DOI:10.1021/jm2014138

The emergence of CTX-M class A extended-spectrum β-lactamases poses a serious health threat to the public. We have applied structure-based design to improve the potency of a novel noncovalent tetrazole-containing CTX-M inhibitor (K_i = 21 μM) mo

Full text of DOI:10.1021/jm2014138

Article
pubs.acs.org/jmc
Structure-Based Design of Potent and Ligand-Efficient Inhibitors of
CTX-M Class A β-Lactamase
Derek A. Nichols,† Priyadarshini Jaishankar,‡ Wayne Larson,‡,∥ Emmanuel Smith,† Guoqing Liu,†
Racha Beyrouthy,§,⊥ Richard Bonnet,§ Adam R. Renslo,*,‡ and Yu Chen*,†
†University of South Florida College of Medicine, Department of Molecular Medicine,12901 Bruce B. Downs Boulevard, MDC 3522,
Tampa, Florida 33612, United States
‡Department of Pharmaceutical Chemistry and Small Molecule Discovery Center, University of California San Francisco, 1700
Fourth Street, Byers Hall S504, San Francisco, California 94158, United States
§Clermont Universite,
́ ́ ́
Universite d’Auvergne, M2iSH, BP10448, INRA, USC 2018, CHU Clermont-Ferrand, Service de Bacteriologie,
F-63003 Clermont-Ferrand, France
ABSTRACT: The emergence of CTX-M class A extended-
spectrum β-lactamases poses a serious health threat to the
public. We have applied structure-based design to improve the
potency of a novel noncovalent tetrazole-containing CTX-M
inhibitor (Ki = 21 μM) more than 200-fold via structural
modifications targeting two binding hot spots, a hydrophobic
shelf formed by Pro167 and a polar site anchored by Asp240.
Functional groups contacting each binding hot spot
independently in initial designs were later combined to
produce analogues with submicromolar potencies, including 6-trifluoromethyl-3H-benzoimidazole-4-carboxylic acid [3-(1H-
tetrazol-5-yl)-phenyl]-amide, which had a Ki value of 89 nM and reduced the MIC of cefotaxime by 64-fold in CTX-M-9
expressing Escherichia coli. The in vitro potency gains were accompanied by improvements in ligand efficiency (from 0.30 to
0.39) and LipE (from 1.37 to 3.86). These new analogues represent the first nM-affinity noncovalent inhibitors of a class A β-
lactamase. Their complex crystal structures provide valuable information about ligand binding for future inhibitor design.
structurally novel (non-β-lactam) inhibitors of β-lacta-
mases.20−24
INTRODUCTION
■
β-Lactam compounds, such as penicillins, are the most widely
used antibiotics due to their effective inhibition of the
transpeptidases required for bacterial cell wall synthesis.1−4 β-
Lactamases catalyze β-lactam hydrolysis and are primary
mediators of bacterial resistance to these compounds.5−7
There are four β-lactamase families, classes A−D, among
which classes A and C are the most commonly observed in the
clinic.8,9 CTX-M is a new group of class A β-lactamases that is
particularly effective against the extended spectrum β-lactam
antibiotics such as cefotaxime,10−14 which itself was developed
to counter bacterial resistance to first-generation penicillins and
cephalosporins. The widespread emergence of extended
spectrum β-lactamase (ESBL) such as CTX-M will continue
to limit treatment options for bacterial infections. Since its
discovery in the 1990s, CTX-M has become the most
frequently observed ESBL in many regions of the world.
The use of a β-lactamase inhibitor in combination with a β-
lactam antibiotic is a well-established strategy to counter
resistance.15 Existing β-lactamase inhibitors (e.g., clavulanic
acid) generally contain a β-lactam ring, making them
susceptible to resistance stemming from up-regulation of β-
lactamase production, selection for new β-lactamases, and other
mechanisms evolved over millions of years of chemical warfare
between bacteria and β-lactam producing microorganisms.16−19
In principle, these problems can be overcome by developing
We recently described our application of fragment-based
molecular docking to identify a new class of noncovalent
inhibitors of CTX-M β-lactamase.25,26 The fragment-based
approach allowed us to overcome the limited chemical diversity
of lead/drug-sized compound libraries, which presents a
particularly challenging problem for antibiotic discovery. This
is due to the starkly different chemical features of antibiotics
when compared to drugs targeting human proteins such as
GPCRs, toward which most HTS screening libraries are
inherently biased.27−29 Thus, our virtual screening of fragment
libraries led to the discovery of a tetrazole-based inhibitor
chemotype, represented by compound 1, at the time the
highest affinity (Ki = 21 μM) noncovalent inhibitor of any class
A β-lactamase (Figure 1, Table 1). The tetrazole-based
chemotype appealed to us for several reasons. First, the
tetrazole group has excellent shape and electrostatic com-
plementarity with the active site subpocket of CTX-M that
usually binds the C(3)4′ carboxylate of traditional β-lactam
antibiotics (Figure 1). Second, the tetrazole ring is a well-
known binding bioisostere for carboxylate groups that often
possesses more favorable pharmacokinetic properties.
Received: October 19, 2011
Published: February 1, 2012
© 2012 American Chemical Society
2163
dx.doi.org/10.1021/jm2014138 | J. Med. Chem. 2012, 55, 2163−2172

Journal of Medicinal Chemistry
Article
bonding interaction or a salt bridge to Asp240 in a modified
ligand would greatly enhance binding affinity. In a previously
determined complex structure with a boronic acid inhibitor
bearing a ceftazidime side chain (PDB ID 1YLY),12 Asp240 has
been observed to form such a hydrogen bond with the ligand’s
distal amine group, which may partially account for the
improved β-lactamase activity of CTX-M against ceftazidime.
Meanwhile, the fluorine atom of compound 1 is 4−5 Å away
from a cluster of protein carbon atoms including Pro167Cβ,
Pro167Cγ, Pro167C, Thr168Cα, and Thr171Cγ, suggesting
that more favorable vdw contacts and hydrophobic interactions
could be formed between this site and analogues modified at
the three position of the aryl ring. Hence, on the basis of these
observations from the complex structure of 1, two series of
analogues independently targeting each of the two potential
binding hot spots were designed, docked, and synthesized.
Synthesis. The simple achiral structure of 1 is an attractive
feature of this chemotype compared to traditional β-lactamase
inhibitors. Designed analogues that docked well to CTX-M
computationally were synthesized in 1−3 synthetic steps or in
some cases could be purchased commercially. Thus, reaction of
commercially available 3-(1H-tetrazol-5-yl)aniline with various
commercial or synthesized carboxylic acids or acid chlorides
afforded the final analogues. Our initial designs focused
exclusively on the 3-fluoro aryl ring of 1 as this moiety is in
proximity to both of the putative binding hot spots we sought
to target. The aryl tetrazole moiety of 1 appears to be highly
complementary to its binding site already and was thus viewed
as a structural anchor that would be unlikely to bind much
differently in the new analogues. This prediction was borne out
when complex structures of new analogues were solved, as
detailed later.
Figure 1. Crystal structure of compound 1 in complex with CTX-M-
9.25 Compound 1 (Ki = 21 μM) carbon atoms are shown in yellow,
nitrogens in blue, oxygens in red, and fluorine in light blue. The
dashed lines represent hydrogen bonds with a sphere representing a
water molecule.
Close examination of the complex crystal structure with
compound 1 revealed two potential binding hot spots that
might be exploited to improve affinity (Figure 1). The first is a
relatively nonpolar binding surface surrounding Pro167, while
the second site, anchored by Asp240, suggested the possibility
of introducing a favorable electrostatic interaction. The design
and synthesis of compounds that focused on interacting with
these binding hot spots has resulted in several novel high
affinity noncovalent inhibitors. The binding of many new
inhibitors to the CTX-M active site was further investigated
using X-ray crystallography. Targeting both hot spots with
small substituents allowed us to identify more ligand-efficient
inhibitors, including the highest affinity (Ki = 89 nM)
noncovalent inhibitor of a class A β-lactamase reported to date.
RESULTS
■
Enzymology and Binding Affinities. To investigate the
effectiveness of the new analogues, we employed CTX-M-9 and
a UV-absorbance based biochemical assay to obtain binding
affinities. A series of analogues with modifications at the three
position were evaluated with the expectation that these
compounds would form more favorable nonpolar contacts
with Pro167 (Table 1). Compounds 2−11 possess 3-
substituents of roughly increasing size and with generally
lipophilic character, though not exclusively so. Interestingly, all
of these modified analogues proved superior to 1 in terms of Ki,
Structure-Based Design. Compound 1 and other less
potent inhibitors were identified by a molecular docking screen
of the ZINC lead-like database using the program
DOCK.25,30−32 In the complex crystal structure of 1 with
CTX-M, the fluoro-benzene ring is in close proximity to
Asp240 and the binding surface surrounding Pro167. Two
carbon atoms on the fluoro-benzene ring are in vdw contact
(3.43 Å) with the Oδ1 atom of Asp240. Although there are
favorable electrostatic interactions between the ring hydrogen
atoms of 1 and Asp240 Oδ1, we hypothesized that a hydrogen
Table 1. Analogues Designed to Target Hydrophobic Shelf Formed by Pro167
a
b
compd
R =
Ki (μM)
LE
LipE
1
F
21
0.30
0.34
0.35
0.36
0.33
0.30
0.30
0.32
0.29
0.32
0.26
1.37
1.63
1.84
1.74
1.74
1.95
2.51
2.45
2.09
1.73
1.57
2
Cl
Me
Br
6.2
4.5
3.0
3.1
17.1
10.2
3.8
6.9
2.4
9.7
3
4
5
cyclopropyl
OMe
6
7
Ac
8
NO2
9
CO2Me
CF3
10
11
2-pyrimidyl
a
b
LE, ligand efficiency; ΔGbind(kcal)/(no. of heavy atoms). LipE = pKi − clogP (clogP calculated using MarvinSketch 5.5.0.1).
2164
dx.doi.org/10.1021/jm2014138 | J. Med. Chem. 2012, 55, 2163−2172

Journal of Medicinal Chemistry
Article
but the most ligand-efficient analogues were those possessing
spheroid hydrophobes (e.g., Me, Br, CF3). The highest affinity
compound from the series targeting Pro167 was compound 10,
the 3-trifluoromethyl variant, with a Ki of 2.4 μM (Table 1).
Analogues 6−8, bearing less hydrophobic and nonspheroid
substituents, had similar ligand efficiency as 1 but superior LipE
values (1.37 for 1 vs 2.51 for 7). The parameter LipE (defined
as logKi − clogP)33 provides a measure of binding affinity
improvement achieved while retaining favorable physiochemical
properties. A surprising result was that even large substituents
(2-pyrimidyl, compound 11) could be tolerated at the 3-
position, although ligand efficiency suffers. A complex crystal
structure of 11 confirmed a similar binding pose as 1 (see
below), so the reduced ligand efficiency of this analogue
perhaps reflects a steric clash and/or unfavorable desolvation
energy associated with the burial of a pyrimidine ring nitrogen
atom.
From these initial two libraries, we concluded that the
independent targeting of each binding hot spot (Asp240 and
Pro167) could indeed be leveraged to produce more potent
and ligand-efficient inhibitors. As described below, the design
principles and predicted binding poses of these initial analogues
were validated by the solution of complex crystal structures for
representative examples. Having identified more favorable
binding elements for both hot spots, the logical next step was
to combine these to produce inhibitors that targeted both sites
simultaneously (Table 3). Indeed, the combination of a
Table 3. Analogues Designed to Target Both Pro167 and
Asp240
To target the second hot spot comprising the area around
Asp240, we designed analogues bearing hydrogen bond donors
and/or charged side chains at various positions on the aryl ring.
These various designs were docked to CTX-M and some of the
best-scoring analogues were synthesized and tested in the
biochemical assay (Table 2). Our attempt to form a salt bridge
Table 2. Analogues Designed to Target Asp240
a
b
LE, ligand efficiency; ΔGbind(kcal)/(no. of heavy atoms). LipE = pKi
− clogP (clogP calculated using MarvinSketch 5.5.0.1).
benzimidazole ring as in 16 with a trifluoromethyl substituent
as in 10, afforded analogue 19, the most potent analogue yet
identified (Ki = 89 nM; LE = 0.36; LipE = 3.86). We expected
that the benzimidazole ring in 19 might contribute an
important hydrogen bond to Asp240 and therefore explored
whether simple hydroxyl or amino substituents in this position
could function similarly (analogues 20−22). These analogues
were indeed more potent than the direct comparators 4 and 10
which lack a hydrogen bond donor but 20−22 were not as
potent as 19. As detailed later, the solution of a complex
structure of benzimidazole analogue 16 revealed additional
contacts that may explain the improved potency of
benzimidazole 19 as compared to 20 and 22. Unexpectedly,
the fluoro benzimidazole analogue 23 was only equipotent to
des-fluoro comparator 16, the flouro substituent apparently not
providing any additional affinity via putative interaction with
Pro167. Perhaps tighter association of the benzimidazole ring
with Asp240 draws the ligand slightly away from Pro167, thus
requiring a larger substituent (such as trifluoromethyl in 19) to
productively contact Pro167.
a
b
LE, ligand efficiency; ΔGbind(kcal)/(no. of heavy atoms). LipE = pKi
− clogP (clogP calculated using MarvinSketch 5.5.0.1).
to Asp240 by the introduction of a basic dimethylamino side
chain (compound 12) was unsuccessful, the analogue
possessing only modest affinity (Ki = 76 μM). The
regioisomeric aryl nitriles 13 and 14 had very different
affinities, with meta-substitution as in 14 preferred (Ki = 7.2
μM). By far the most interesting analogues from this series (Ki
∼ 1 μM) were the heterocyclic analogues 16−18, each of which
possesses a potential hydrogen bond donor in a position
(pseudo meta or para) predicted by docking to be in close
proximity to Asp240. Moreover, compounds 16−18 exhibited
improved ligand efficiency (0.34−0.35) and LipE values (2.59−
3.58) as compared to 1 (0.31 and 1.37).
X-ray Crystallographic Structure Determination. The
structural details of the interactions between CTX-M-9 and
several of the new analogues were investigated in order to gain
an understanding of the molecular basis for the binding affinity
improvement and to facilitate future inhibitor development.
Complex crystal structures with CTX-M-9 were determined to
2165
dx.doi.org/10.1021/jm2014138 | J. Med. Chem. 2012, 55, 2163−2172

Journal of Medicinal Chemistry
Article
Table 4. X-Ray Data Collection and Refinement Statistics
compounds
4
10
11
12
16
18
Data Collection
space group
cell dimensions
a (Å)
P21
P21
P21
P21
P21
P21
45.174
107.188
47.487
90
45.014
107.105
47.509
90
45.245
106.941
47.681
90
45.125
106.922
47.481
90
45.192
106.565
47.669
90
45.166
106.567
47.825
90
b (Å)
c (Å)
α (deg)
β (deg)
100.38
90
102.105
90
100.972
90
101.469
90
101.717
90
101.857
90
γ (deg)
resolution (Å)
no. reflns
50−1.44
75901
6.3
50−1.36
92455
6.0
50−1.36
87996
5.6
50−1.40
86200
7.0
50−1.12
158174
5.7
50−1.26
118432
4.2
Rmerge (%)
I/σI
a
12.9(2.2)
95.0
17.9(2.1)
98.9
18.2(2.1)
92.5
14.8(2.1)
99.7
14.7(3.0)
93.3
21.2(4.0)
99.4
completeness (%)
redundancy
Refinement
resolution (Å)
2.5
3.7
3.7
3.7
3.6
3.7
50−1.44
50−1.36
50−1.36
50−1.40
50−1.12
50−1.26
R
work/Rfree (%)
16.3/20.6
17.1/20.5
15.6/19.9
15.4/19.4
13.4/16.6
12.3/16.3
no. heavy atoms
protein
4017
238
3941
300
3990
134
4023
104
4085
50
4090
76
ligand/ion
water
480
519
493
546
734
752
B factors (Å2)
protein
11.70
24.84
27.64
12.40
22.90
28.08
8.414
19.73
24.47
10.76
14.42
26.58
7.028
6.883
21.29
8.856
14.70
23.78
ligand/ion
water
rms deviations
bond lengths (Å)
bond angles (deg)
0.013
1.555
0.015
1.656
0.011
1.380
0.007
1.149
0.012
1.482
0.011
1.437
Ramachandran plot
most favored region (%)
additionally allowed (%)
generously allowed (%)
96.9
1.9
97.2
1.8
96.7
2.1
96.6
2.3
96.7
2.0
96.7
2.0
1.2
1.0
1.3
1.1
1.3
1.3
a
Values in parentheses represent highest resolution shells.
a resolution in the range of 1.2−1.4 Å, where the ligand binding
pose can be determined unambiguously. In all of these
structures, the inhibitor adopts a single pose, as shown by the
unbiased Fo − Fc electron densities calculated prior to the
fitting of the ligand (Table 4).
functional groups on 4 and 10 are evident. The bromine atom
on the ring structure of 4 is 4−5 Å away from the cluster of
protein carbon atoms including Pro167Cβ, Pro167Cγ,
Pro167C, Thr168Cα, and Thr171Cγ. Likewise, the three
branched fluorine atoms of compound 10 are in close vdw
contact with these protein carbon atoms, which are
approximately 3.4−3.8 Å away. The binding of compound
11, on the other hand, differs slightly from compounds 1, 4,
and 10 in these regions (Figure 2c). The pyrimidine ring forms
a water-mediated interaction with Asp240 and induces Asp240
to adopt a new conformation (Figure 2). This water-mediated
contact exists in the complex structure with only partial
occupancy, as suggested by the relatively weak electron density
of the water (2σ) and the presence of two Asp240
conformations, including the one previously observed in apo
and other complex structures. There is vdw contact observed
between the carbon atoms of the pyrimidine ring in 11 and
Pro167Cγ, Thr168Cγ, and Thr171Cγ, which are 3.3−3.6 Å
away. Despite the water-mediated hydrogen bond and vdw
contacts between the ring carbon atoms and Pro167 and
Thr168, the affinity of compound 11 is less than that of 4 or
10; this may be due to unfavorable burial of a polar pyrimidine
ring nitrogen atom and electrostatic repulsion between this
nitrogen and Pro167O. Additionally, the vdw contacts
Figure 2 shows the X-ray crystal structures of compounds 4,
10, and 11 in the active site of CTX-M-9; these compounds
were designed to make significant nonpolar interactions with
Pro167. The size increase in the bulkier side substituents such
as trifluoromethyl is demonstrated in their larger electron
density volumes compared with that of the fluorine atom in 1.
In the larger sense, the atoms of the new ligands make similar
contacts with the surrounding active site atoms, as does 1. For
instance, compounds 4, 10, and 11 (Figure 2a−c) all form
hydrogen bonds between the tetrazole ring and Thr235,
Ser237, and Ser130 from the protein, which is similar to
compound 1 (Figure 1, Table 1). They also share the
characteristic water-mediated interaction between the amide
linkage and Ser237, as well as two hydrogen bonds with Asn132
or Asn104. The contacts between the distal benzene ring and
Asp240, as observed in compound 1, are maintained in
compounds 4 and 10, with two ring carbon atoms in vdw
contact and approximately 3.2−3.3 Å away from the Oδ1 atom
of Asp240. The favorable contacts between Pro167 and the
2166
dx.doi.org/10.1021/jm2014138 | J. Med. Chem. 2012, 55, 2163−2172

Journal of Medicinal Chemistry
Article
Figure 3. Crystal complex structures with compounds targeting
Asp240. (a) Compound 18. (b) Compound 16. (c) Compound 12, in
comparison to the designed pose in cyan. The red dashed lines
represent hydrogen bonds between the ligand and CTX-M-9. The
carbon atoms of the protein are colored in green along with oxygens in
red and nitrogens in blue. The ligand carbon atoms are colored yellow.
Resolution for the structures ranges from 1.2 to 1.4 Å. Unbiased Fo −
Fc densities are shown in blue at 3σ.
Figure 2. Crystal complex structures with compounds targeting
Pro167. (a) Compound 4. (b) Compound 10. (c) Compound 11. The
red dashed lines represent hydrogen bonds between the ligand and
CTX-M-9. The carbon atoms of the protein are colored in green along
with oxygens in red and nitrogens in blue. The ligand carbon atoms
are colored yellow. Resolution for the structures ranges from 1.2 to 1.4
Å. Unbiased Fo − Fc densities are shown in blue at 3σ.
described above may be slightly too close for the optimal
carbon−carbon distance in vdw interaction (∼4 Å) and thus
suggest a possible minor steric clash.
significantly. The modest additional affinity of compound 16,
in comparison to 15, may instead originate from an
intramolecular hydrogen bond between N-3 and the proximal
amide N−H, an interaction that would stabilize the
conformation conducive to hydrogen bond formation with
Asp240.
Figure 3c shows the discrepancy between the designed
interaction of compound 12 with Asp240 (in cyan) and its
actual interactions observed in the crystal structure (in yellow).
Initially, we designed compound 12 to form a salt bridge with
Asp240. However, the X-ray crystal structure reveals the actual
binding pose in which a water-mediated hydrogen bond is
formed between the positively charged side chain and Asp240.
The new side chain is cradled in the small pocket surrounding
Pro167, once again underscoring the potential of this binding
surface in establishing new interactions with future inhibitors.
Antimicrobial Activity. The activity of compound 19 was
investigated in clinical bacterial isolates that exhibit high levels
of resistance against third-generation cephalosporins (e.g.,
cefotaxime) via expression of CTX-M β-lactamases. When the
compound was administrated alone in a disk diffusion assay, it
Crystal structures were also obtained for compounds
designed to establish polar interactions with Asp240, including
compounds 12, 16, and 18. Again, the core structure of these
compounds, including the tetrazole ring and the amide bond,
establishes contacts with Ser130, Thr235, Ser237, Asn104, and
Asn132, similar to compound 1 (Figure 3a−c). Both
compounds 16 and 18 form a direct hydrogen bond with
Asp240 as designed. In addition, compound 16 has a favorable
contact between N-1/C-2/N-3 of the benzimidazole ring and
the main chain atoms around Gly238, while compound 18
establishes more vdw interactions with Pro167 and Thr168.
The second ring nitrogen (N-3) in benzimidazole 16, which is a
carbon in indole 15, appears to form a water-mediated
hydrogen bonding contact with Ser237 (Figure 3b). However,
the electron density for the water molecule contacting N-3 in
16 is weaker (2.4σ) than other structural waters in the active
site, suggesting that this water-mediated interaction is relatively
unstable and may not contribute to binding affinity
2167
dx.doi.org/10.1021/jm2014138 | J. Med. Chem. 2012, 55, 2163−2172

Journal of Medicinal Chemistry
Article
had little or no detectable effect on bacteria growth, as expected
(Figure 4). However, in combination with cefotaxime, the
Both Pro167 and Asp240 have been observed to interact
with β-lactam substrates or covalent inhibitors in complex
structures with CTX-M-9. In a recent crystal structure of CTX-
M-9 S70G mutant and cefotaxime (PDB ID 3HLW),11 the
amino group on the aminothiazole ring of cefotaxime forms a
hydrogen bond with Asp240 while the methoxyimino group
nestles comfortably in the subpocket around Pro167.
Compared with the apo structure, such interactions cause
small shifts in atom positions for residues in this area (e.g., ∼0.5
Å for Asp240Cα); a conformational change not observed in
complex structures with smaller substrates such as benzylpe-
nicillin. Similar hydrogen bonds with Asp240 have also been
found in previous complex structures with boronic acid
inhibitors.11,12 Additionally, Pro167 and Asp240 are conserved
in other CTX-M type enzymes such as Toho-1.35−37 In the
acyl-enzyme complex structure of Toho-1 Glu166A mutant
with cefotaxime (PDB ID, 1IYO),37 the aminothiazole ring
makes both a direct and a water-mediated hydrogen bond with
Asp240 while establishing vdw interactions with Pro167.
Together with our studies described herein, these observations
suggest both Asp240 and Pro167 are binding hot spots useful
for inhibitor design against CTX-M β-lactamases.
Figure 4. Disk diffusion plate assay showing the antimicrobial activity
of compound 19. The compound was administered alone or in
combination with cefotaxime against an E. coli strain producing CTX-
M-9 β-lactamase.
Even more significantly, it is possible to consider targeting
similar hot spots in other class A β-lactamases. In narrow-
spectrum β-lactamases, such as TEM-1 and SHV-2, residue 240
is a glutamate. Although it has been hypothesized that the
substitution of Glu240 for Asp may enlarge the active site and
allow ESBLs, such as CTX-M, to accommodate the bulkier side
chains of cefotaxime and other third-generation cephalosporins,
both Glu240 and Asp240 present similar features in the protein
binding pocket, including the net negative charge and the
nearly identical positioning of one oxygen atom from the
carboxylate group. Comparing the complex structures between
a ceftazidime-like boronic acid inhibitor and CTX-M-9 (PDB
ID, 1YLY) to that of the same compound with TEM-1 (PDB
ID 1M40) shows that the aminothiazole ring of the inhibitor is
placed in similar positions and forms a hydrogen bond with
residue 240 in both structures.12,38 Additionally, comparing the
affinity of compound 19 with those of compounds 21 and 22
suggests that interactions with the main chain atoms around
Gly238, the residue immediately preceding Asp240 (note the
numbering gap due to convention), may also contribute
significantly to binding. Gly238 is highly conserved in CTX-M,
TEM, and SHV enzymes. Meanwhile, the nonpolar binding
surface around residue 167 is also largely conserved in these β-
lactamases. Like CTX-M, TEM-1 has a proline in this position.
Although it is replaced by a threonine in SHV enzymes, most of
the carbon atoms, like Cα and Cβ atoms of residues 167 and
168, are in similar positions and thus form a binding subpocket
with features comparable to that in CTX-M, albeit with some
new features such as Thr167Oγ. In the crystal structure
between cefoperazone transition state analog and SHV-1, the
carbon atoms of the compound’s piperazine ring are in vdw
contacts with the Cβ and C atoms of Thr167.39
compound produced a large inhibition halo surrounding the
disk. The size of the inhibition zone showed improved
inhibition of bacterial growth compared to cefotaxime alone
and increased with the concentration of the compound. This
dose-responsive inhibition of bacterial growth revealed clear
synergy between cefotaxime and 19.
Antimicrobial activity was investigated quantitatively to
determine the minimum inhibitory concentrations (MICs) of
the β-lactam/inhibitor combination necessary to inhibit
bacterial growth. The MICs of cefotaxime alone against the
two Escherichia coli strains, expressing CTX-M-14 and CTX-M-
9, respectively, corresponded to a high level of resistance
(MICs, 32 and 64 μg/mL) according to Clinical and
Laboratory Standards Institute (CLSI) standards.34 Compound
19 had no measurable antibiotic activity when used alone (MIC
≥ 64 μg/mL). In combination with cefotaxime, compound 19
improved MIC values by 64-fold and restored susceptibility to
cefotaxime in these resistant bacterial strains (MICs, 0.5 and 1
μg/mL). These results provide evidence that compound 19 is
able to cross bacterial outer membrane to access its intended
target.
DISCUSSION
■
The identification of novel noncovalent inhibitors of class A β-
lactamases is a promising approach to maintain the
effectiveness of β-lactam antibiotics. The purpose of this initial
study was to rapidly identify regions of the active site that could
be more productively engaged with designed ligands, thus
enabling further optimization of tetrazole-based inhibitors of
CTX-M β-lactamase. The targeted introduction of new
functional groups in the distal ring of 1 succeeded in producing
improved analogues that make both nonpolar and polar
contacts with CTX-M β-lactamase, improving affinity ∼200-
fold while retaining good lead-like properties (reflected in
notably improved LipE values). The results confirm the
importance of Pro167 and Asp240 as binding hot spots in
CTX-M β-lactamase and demonstrate the tractability of this
novel inhibitor chemotype.
In addition to revealing the importance of Pro167 and
Asp240 in ligand binding, the rapid evolution of compound 1
into nanomolar inhibitors like 19 demonstrates the tractability
of the tetrazole chemotype as a lead scaffold. The five-member
tetrazole ring displays both good shape and electrostatic
complementarity with a subpocket usually occupied by the
C(3)4′ carboxylate group of β-lactam compounds, forming
three hydrogen bonds with Ser130, Thr235, and Ser237 while
being stacked against the peptide bond between Thr235 and
2168
dx.doi.org/10.1021/jm2014138 | J. Med. Chem. 2012, 55, 2163−2172

Journal of Medicinal Chemistry
Article
All synthesized analogues tested against CTX-M were judged to be
of 95% or higher purity based on analytical LC/MS analysis. LC/MS
analyses were performed on a Waters Micromass ZQ/Waters 2795
separation module/Waters 2996 photodiode array detector system
controlled by MassLynx 4.0 software. Separations were carried out on
an XTerra MS C18 5 μm 4.6 mm × 50 mm column at ambient
temperature using a mobile phase of water−acetonitrile containing
0.05% trifluoroacetic acid. Gradient elution was employed wherein the
acetonitrile−water ratio was increased linearly from 5% to 95%
acetonitrile over 2.5 min, then maintained at 95% acetonitrile for 1.5
min, and then decreased to 5% acetonitrile over 0.5 min and
maintained at 5% acetonitrile for 0.5 min. Compound purity was
determined by integrating peak areas of the liquid chromatogram,
monitored at 254 nm.
General Procedure A. An oven-dried vial or flask is charged with 3-
(1H-tetrazol-5-yl)aniline (1 equiv), the appropriate carboxylic acid (1
equiv), 1-ethyl-3-[3-dimethylaminopropyl]carbodiimide hydrochloride
(1.5 equiv), 1-hydroxybenzotriazole (1.5 equiv), and N,N′-diisopropy-
lethylamine (2 equiv) and stirred in DMF (0.5 mL) at room
temperature for 24 h or until judged complete by LC/MS analysis.
The reaction mixture is diluted with water (2 mL), and after adjusting
the pH to ∼2 with 1N HCl, the mixture is extracted with ethyl acetate.
The organic extracts are washed with brine, dried over magnesium
sulfate, and concentrated under reduced pressure. The crude material
thus obtained is purified by reverse phase HPLC to afford the desired
product.
Gly236. Several key features of this binding subpocket are also
present in the active site of AmpC class C β-lactamase. For
example, Thr235 and Gly236 are conserved in AmpC (Thr316
and Gly317). Tyr150, a key catalytic residue in AmpC, places
its hydroxyl group in a position similar to that of Ser130 in
CTX-M. Other common features shared by the active sites of
class A and C enzymes may further allow the design of
inhibitors with broader spectra. For instance, existing covalent
inhibitors against both classes of enzymes almost invariantly
place an oxygen atom in the oxyanion hole formed by two
backbone amide groups. A water molecule occupies the
oxyanion hole in the complex structures of our tetrazole-
based inhibitors. The identification of tetrazole-based inhibitors
that favorably displace this water molecule may improve
binding affinity and expand the utility of this chemotype to
target a wider range of β-lactamases.
CONCLUSIONS
■
Structurally guided optimization of a novel-class of CTX-M β-
lactamase inhibitors has confirmed two binding hot spots that
can be targeted to produce high affinity inhibitors. Importantly,
these hot spots are shared by other therapeutically important
groups of β-lactamases, suggesting the potential for tetrazole-
class inhibitors with an expanded spectrum of β-lactamase
activity. More generally, the approach we have employed to
identify and optimize novel noncovalent inhibitors of CTX-M
can be applied to develop novel inhibitors of other important β-
lactamases. The nanomolar potency of 19 distinguishes this
compound as the highest affinity noncovalent inhibitor yet
identified for a class A β-lactamase. Its antimicrobial activity
also demonstrates the possible utility of the tetrazole scaffold in
countering bacterial resistance. Current efforts are focused on
further elaborating the tetrazole chemotype, with the goal of
producing a novel class of compounds effective against a wide
range of clinically relevant β-lactamases.
General Procedure B. An oven-dried vial or flask is charged with 3-
(1H-tetrazol-5-yl)aniline (1 equiv), the appropriate acid chloride (1.05
equiv), and N,N′-diisopropylethylamine (2 equiv) and stirred in
dichloromethane (5 mL) at room temperature for 30 min. The
reaction mixture is diluted with dichloromethane and washed with
water. After adjusting the pH to ∼2 with 1N HCl, the mixture is
extracted with ethyl acetate. The organic extracts are washed with
brine, dried over magnesium sulfate, and concentrated under reduced
pressure. The crude material thus obtained is purified by flash column
chromatography (5−20% methanol/dichloromethane).
3-Cyclopropyl-N-[3-(1H-tetrazol-5-yl)-phenyl]-benzamide (5). 3-
(1H-Tetrazol-5-yl)aniline was reacted with commercially available 3-
cyclopropylbenzoic acid according to general procedure A to afford
1
the title compound in 63% yield. H NMR (DMSO-d6) δ 10.42 (s,
1H), 8.55 (s, 1H), 7.93 (d, J = 8 Hz, 1H), 7.72 (d, J = 8 Hz, 2H), 7.65
(s, 1H), 7.56 (t, J = 8 Hz, 1H), 7.39 (t, J = 8 Hz, 1H), 7.29 (d, J = 8
Hz, 1H), 1.97−2.03 (m, 1H), 0.97−1.01 (m, 2H), 0.74−0.78 (m, 2H).
LCMS (ESI) m/z 306 (MH+).
EXPERIMENTAL METHODS
■
Compound Docking. Molecular docking was used as previously
described25 to evaluate newly designed compounds or existing ones
from the ZINC small-molecule database with the program DOCK
3.5.54.25,30−32
3-Acetyl-N-[3-(1H-tetrazol-5-yl)-phenyl]-benzamide (7). 3-(1H-
Tetrazol-5-yl)aniline was reacted with commercially available 3-
acetylbenzoic acid according to general procedure A to afford the
Synthesis. General Methods. 1H NMR spectra were
recorded on a Varian INOVA-400 400 MHz spectrometer.
Chemical shifts are reported in δ units (ppm) relative to TMS
as an internal standard. Coupling constants (J) are reported in
hertz (Hz). The known compounds 1, 2, 3, 4, 6, 8, 10, and 1340
were prepared according the general procedures and/or were
obtained from commercial sources (Ryan Scientific, TimTec).
All other reagents and solvents were purchased from Aldrich
Chemical, Acros Organics, Enamine, Alfa Aesar, and Apollo
Scientific and used as received. Air and/or moisture sensitive
reactions were carried out under an argon atmosphere in oven-
dried glassware using anhydrous solvents from commercial
suppliers. Air and/or moisture sensitive reagents were trans-
ferred via syringe or cannula and were introduced into reaction
vessels through rubber septa. Solvent removal was accom-
plished with a rotary evaporator at ca. 10−50 Torr. Column
chromatography was carried out using a Biotage SP1 flash
chromatography system and silica gel cartridges from Biotage.
Analytical TLC plates from EM Science (Silica Gel 60 F254)
were employed for TLC analyses. Microwave heating was
accomplished using a CEM reaction microwave. Hydrogenation
reactions were carried out with a ThalesNano H-Cube
hydrogenator.
1
title compound in 33% yield. H NMR (DMSO-d6) δ 10.67 (s, 1H),
8.57 (s, 1H), 8.52 (s, 1H), 8.22 (d, J = 8 Hz, 1H), 8.16 (d, J = 8 Hz,
1H) 7.96 (d, J = 8 Hz, 1H), 7.75 (d, J = 8 Hz, 1H), 7.70 (t, J = 8 Hz,
1H), 7.59 (t, J = 8 Hz, 1H), 2.65 (s, 3H). LCMS (ESI) m/z 308 (MH
+).
N-[3-(1H-Tetrazol-5-yl)-phenyl]-isophthalamic Acid Methyl Ester
(9). 3-(1H-Tetrazol-5-yl)aniline was reacted with commercially
available mono-methylisophthalate according to general procedure A
1
to afford the title compound in 22% yield. H NMR (DMSO-d6) δ
10.70 (s, 1H), 8.56 (s, 2H), 8.25 (d, J = 8 Hz, 1H), 8.16 (d, J = 8 Hz,
1H) 7.96 (d, J = 8 Hz, 1H), 7.68−7.75 (m, 2H), 7.58 (t, J = 8 Hz, 1H),
3.90 (s, 3H). LCMS (ESI) m/z 324 (MH+).
3-Pyrimidin-2-yl-N-[3-(1H-tetrazol-5-yl)-phenyl]-benzamide (11).
3-(1H-Tetrazol-5-yl)aniline was reacted with commercially available 3-
pyrimidin-2-yl-benzoic acid according to general procedure A to afford
1
the title compound in 11% yield. H NMR (DMSO-d6) δ 10.71 (s,
1H), 8.99 (s, 1H), 8.96 (d, J = 4 Hz, 2H), 8.58−8.61 (m, 2H), 8.14 (d,
J = 8 Hz, 1H), 7.99 (d, J = 8 Hz, 1H), 7.68−7.76 (m, 2H), 7.58 (t, J =
8 Hz, 1H), 7.49 (d, J = 4 Hz, 1H). 13C NMR (100 MHz, DMSO-d6) δ
166.2, 163.3, 158.5, 140.7, 138.1, 135.9, 131.4, 130.7, 130.4, 129.6,
127.7, 123.6, 122.8, 121.0, 119.3. LCMS (ESI) m/z 344 (MH+).
3-Dimethylaminomethyl-N-[3-(1H-tetrazol-5-yl)-phenyl]-benza-
mide (12). 3-(1H-Tetrazol-5-yl)aniline was reacted with commercially
2169
dx.doi.org/10.1021/jm2014138 | J. Med. Chem. 2012, 55, 2163−2172

Journal of Medicinal Chemistry
Article
available 3-dimethylaminomethyl-benzoic acid according to general
procedure A to afford the title compound in 28% yield. H NMR
Hz, 1H), 7.72 (d, J = 4 Hz, 1H), 7.60 (s, 1H), 7.56 (d, J = 8 Hz, 1H),
7.34 (s, 1H), 7.15 (s, 1H). LCMS (ESI) m/z 361 (MH+).
1
(DMSO-d6) δ 10.60 (s, 1H), 8.57 (s, 1H), 8.08−8.10 (m, 2H), 7.95
(d, J = 8 Hz, 1H), 7.70−7.76 (m, 2H), 7.58−7.64 (m, 2H), 4.37 (s,
2H), 2.75 (s, 6H). 13C NMR (100 MHz, DMSO-d6) δ 165.9, 140.5,
135.9, 134.8, 131.4, 131.3, 130.5, 129.7, 129.3, 123.5, 122.9, 119.4,
60.0, 42.5. LCMS (ESI) m/z 323 (MH+)
3-Hydroxy-N-[3-(1H-tetrazol-5-yl)-phenyl]-5-trifluoromethyl-ben-
zamide (21). 3-(1H-Tetrazol-5-yl)aniline was reacted with commer-
cially available 3-hydroxy-5-trifluoromethyl carboxylic acid according
1
to general procedure A to afford the title compound in 24% yield. H
NMR (DMSO-d6) δ 10.60 (s, 1H), 10.51 (s, 1H), 8.55 (s, 1H), 7.94
(d, J = 8 Hz, 1H), 7.74 (d, J = 12 Hz, 2H), 7.64 (s, 1H), 7.58 (d, J = 8
Hz, 1H), 7.24 (s, 1H). LCMS (ESI) m/z 350 (MH+).
3-Cyano-N-[3-(1H-tetrazol-5-yl)-phenyl]-benzamide (14). 3-(1H-
Tetrazol-5-yl)aniline was reacted with commercially available 3-
cyanobenzoic acid according to general procedure A and purified by
flash column chromatography (5−15% methanol/dichloromethane)
3-Amino-N-[3-(1H-tetrazol-5-yl)-phenyl]-5-trifluoromethyl-ben-
zamide (22). 3-(1H-Tetrazol-5-yl)aniline (75 mg, 0.47 mmol), 3-
amino-5-trifluoromethyl-benzoic acid (96 mg, 0.47 mmol), 1-ethyl-3-
[3-dimethylaminopropyl]carbodiimide hydrochloride (135 mg, 0.7
mmol), 1-hydroxybenzotriazole (95 mg, 0.7 mmol), and N,N′-
diisopropylethylamine (0.16 mL, 0.94 mmol) were stirred in DMF
(0.5 mL) at room temperature for 18 h. The reaction mixture was
filtered and purified by reverse phase HPLC to afford the title
1
to afford the title compound in 66% yield. H NMR (DMSO-d6) δ
10.63 (s, 1H), 8.53 (s, 1H), 8.43 (s, 1H), 8.25 (d, J = 8 Hz, 1H), 8.06
(d, J = 8 Hz, 1H) 7.92 (d, J = 8 Hz, 1H), 7.75 (d, J = 8 Hz, 2H), 7.56
(t, J = 8 Hz, 1H). LCMS (ESI) m/z 291 (MH+).
1H-Indole-4-carboxylic Acid [3-(1H-Tetrazol-5-yl)-phenyl]-amide
(15). 3-(1H-Tetrazol-5-yl)aniline was reacted with commercially
available indole-4-carboxylic acid according to general procedure A
1
compound in 52% yield. H NMR (DMSO-d6) δ 10.50 (s, 1H), 8.53
1
(s, 1H), 7.92 (d, J = 8 Hz, 1H), 7.72 (d, J = 8 Hz, 1H), 7.56 (t, J = 8
Hz, 1H), 7.37 (d, J = 8 Hz, 2H), 7.24 (s, 1H). LCMS (ESI) m/z 349
(MH+).
6-Fluoro-3H-benzoimidazole-4-carboxylic Acid [3-(1H-Tetrazol-
5-yl)-phenyl]-amide (23). 3-(1H-Tetrazol-5-yl)aniline was reacted
with commercially available 6-fluoro-benzimidazole-4-carboxylic acid
according to the general procedure A to afford the title compound in
17% yield. 1H NMR (DMSO-d6) δ 8.63 (s, 1H), 8.53 (s, 1H), 7.99 (d,
J = 8 Hz, 1H), 7.71−7.80 (m, 4H), 7.63 (t, J = 8 Hz, 1H). LCMS
(ESI) m/z 324 (MH+).
to afford the title compound in 10% yield. H NMR (DMSO-d6) δ
11.35 (s, 1H), 10.41 (s, 1H), 8.65 (s, 1H), 7.93 (d, J = 8 Hz, 1H), 7.70
(d, J = 4 Hz, 1H), 7.53−7.62 (m, 3H), 7.47 (s, 1H), 7.20 (t, J = 8 Hz,
1H), 6.85 (s, 1H). LCMS (ESI) m/z 305 (MH+).
3H-Benzoimidazole-4-carboxylic Acid [3-(1H-Tetrazol-5-yl)-phe-
nyl]-amide (16). 3-(1H-Tetrazol-5-yl)aniline (75 mg, 0.47 mmol),
1H-benzimidazole-4-carboxylic acid (76 mg, 0.47 mmol), 1-ethyl-3-[3-
dimethylaminopropyl]carbodiimide hydrochloride (135 mg, 0.7
mmol), 1-hydroxybenzotriazole (95 mg, 0.7 mmol), and N,N′-
diisopropylethylamine (0.17 mL, 0.94 mmol) were stirred in DMF
(0.5 mL) at room temperature for 4 h. The reaction mixture was
diluted with water (1 mL), and after adjusting the pH to ∼4 with 1N
HCl, the mixture was filtered. The filtered precipitate was purified by
reverse phase HPLC to afford the product as a trifluoroacetic acid salt
in 21% yield. 1H NMR (DMSO-d6) δ 11.69 (s, 1H), 9.00 (s, 1H), 8.61
(s, 1H), 8.12 (d, J = 8 Hz, 1H), 7.93−7.99 (m, 2H), 7.79 (d, J = 8 Hz,
1H), 7.52−7.63 (m, 2H). 13C NMR (100 MHz, DMSO-d6) δ 164.1,
159.2, 148.4, 143.4, 140.2, 134.3, 130.8, 125.6, 124.3, 123.1, 122.9,
121.9, 118.9, 118.4. LCMS (ESI) m/z 306 (MH+).
3-Bromo-5-cyano-N-[3-(1H-tetrazol-5-yl)-phenyl]-benzamide
(24). 3-(1H-Tetrazol-5-yl)aniline was reacted with commercially
available 3-cyano-5-bromobenzoic acid according to the general
1
procedure A to afford the title compound in 24% yield. H NMR
(DMSO-d6) δ 10.70 (s, 1H), 8.53 (s, 1H), 8.45 (s, 1H), 8.41 (d, J = 8
Hz, 2H), 7.94 (d, J = 8 Hz, 1H), 7.76 (d, J = 8 Hz, 1H), 7.59 (t, J = 8
Hz, 1H). LCMS (ESI) m/z 370 (MH+).
2-Amino 3-Nitro-5-trifluoromethylbenzoic Acid (25). Commer-
cially available 2-chloro-3-nitro-5-trifluoromethylbenzoic acid (0.10 g,
0.37 mmol) and aqueous ammonium hydroxide (2 mL) were heated
in a sealed tube in a CEM microwave at 120 °C for an hour. After
cooling, the pH was adjusted to 2 with 1N HCl. The precipitate was
filtered and dried to obtain 2-amino 3-nitro-5-trifluoromethylbenzoic
acid as a yellow solid (80 mg). This material was used in the next step
without further purification. 1H NMR (DMSO-d6) δ 8.49 (s, 1H), 8.32
(s, 1H).
1H-Indole-5-carboxylic Acid [3-(1H-Tetrazol-5-yl)-phenyl]-amide
(17). 3-(1H-Tetrazol-5-yl)aniline was reacted with commercially
available indole-5-carboxylic acid according to the general procedure
1
A to afford the title compound in 9% yield. H NMR (DMSO-d6) δ
11.39 (s, 1H), 10.36 (s, 1H), 8.62 (s, 1H), 8.30 (s, 1H), 7.96 (d, J = 8
Hz, 1H), 7.75 (d, J = 8 Hz, 1H), 7.69 (d, J = 8 Hz, 1H), 7.45−7.57 (m,
3H), 6.58 (s, 1H). LCMS (ESI) m/z 305 (MH+).
2,3-Diamino-5-trifluoromethylbenzoic Acid (26). A solution of 2-
amino-3-nitro-5-trifluoromethylbenzoic acid (25, 75 mg, 0.3 mmol) in
methanol was passed through a Pd/C cartridge (10 wt %) at a flow
rate of 1 mL/min using the H-Cube hydrogenation system. The
solution was concentrated under reduced pressure and dried to obtain
the title compound (62 mg). This material was used without further
1H-Indazole-5-carboxylic Acid [3-(1H-Tetrazol-5-yl)-phenyl]-
amide (18). 3-(1H-Tetrazol-5-yl)aniline was reacted with commer-
cially available indazole-5-carboxylic acid according to general
procedure A to afford the title compound in 6% yield. 1H NMR
(DMSO-d6) δ 10.50 (s, 1H), 8.61 (s, 1H), 8.51 (s, 1H), 8.26 (s, 1H),
7.95−7.98 (m, 3H), 7.71 (d, J = 4 Hz, 1H), 7.64 (d, J = 8 Hz, 1H),
7.57 (t, J = 8 Hz, 1H). LCMS (ESI) m/z 306 (MH+).
6-Trifluoromethyl-3H-benzoimidazole-4-carboxylic Acid [3-(1H-
Tetrazol-5-yl)-phenyl]-amide (19). 3-(1H-Tetrazol-5-yl)aniline (15
mg, 0.09 mmol), 6-trifluoromethyl-benzimidazole-4-carboxylic acid
(25 mg, 0.09 mmol), 1-ethyl-3-[3-dimethylaminopropyl]carbodiimide
hydrochloride (26 mg, 0.135 mmol), 1-hydroxybenzotriazole (18 mg,
0.135 mmol), and N,N′-diisopropylethylamine (0.047 mL, 0.27 mmol)
were stirred in DMF (0.2 mL) at room temperature for 24 h. The
reaction mixture was filtered and purified by reverse phase HPLC to
afford the title compound as a trifluoroacetic acid salt in 30% yield. 1H
NMR (DMSO-d6) δ 8.78 (s, 1H), 8.54 (s, 1H), 8.24 (d, J = 12 Hz,
2H), 8.02 (d, J = 8 Hz, 1H), 7.80 (d, J = 8 Hz, 1H), 7.64 (t, J = 8 Hz,
1H). 13C NMR (100 MHz, DMSO-d6) δ 163.1, 155.1, 146.9, 140.0,
130.9, 126.5, 125.6, 123.8, 123.5, 123.1, 122.3, 119.6, 118.8, 94.6.
LCMS (ESI) m/z 374 (MH+).
1
purification in the next reaction. H NMR (CDCl3) δ 7.80 (s, 1H),
7.05 (s, 1H). LCMS (ESI) m/z 221 (MH+).
6-Trifluoromethyl-benzimidazole-4-carboxylic Acid (27). Formic
acid (0.34 mmol, 3 equiv) was added to intermediate 26 (0.11 mmol,
1 equiv) in aqueous 4 M HCl (0.35 mL) and the reaction mixture
heated to 100 °C for 2 h. The reaction mixture was concentrated
under reduced pressure and dried to obtain the title compound (35
mg) as a hydrochloride salt. This material was used without further
purification in the next reaction. 1H NMR (DMSO-d6) δ 8.74 (s, 1H),
8.33 (s, 1H), 8.07 (s, 1H). LCMS (ESI) m/z 231 (MH+).
Protein Purification, Crystallization and Structure Determi-
nation. CTX-M-9, a class A β-lactamase that we have previously
studied, was used to represent the CTX-M family. The protein was
purified as previously described10 and crystallized in 1.2−1.6 M
potassium phosphate buffer (pH 8.3) from hanging drops at 20 °C.
The final concentration of the protein in the drop ranged from 6.5 mg
mL−1 to 9 mg mL−1. The complex crystals were obtained through
soaking methods. On the basis of the variability in terms of solubility
and affinity, compound soaking times varied considerably, from 1 to 24
h. Diffraction was measured at two beamlines: X6A at National
3-Bromo-5-hydroxy-N-[3-(1H-tetrazol-5-yl)-phenyl]-benzamide
(20). 3-(1H-Tetrazol-5-yl)aniline was reacted with commercially
available 3-bromo-5-hydroxybenzoic acid according to general
1
procedure A to afford the title compound in 22% yield. H NMR
(DMSO-d6) δ 10.48 (s, 1H), 10.27 (s, 1H), 8.55 (s, 1H), 7.91 (d, J = 8
2170
dx.doi.org/10.1021/jm2014138 | J. Med. Chem. 2012, 55, 2163−2172

Journal of Medicinal Chemistry
Article
Synchrotron Light Source, Brookhaven, New York, and 23-ID-B of
GM/CA CAT at Advanced Photon Source (APS), Argonne, Illinois.
Data were processed with HKL2000.41 The models for refinement
were first obtained through using a rigid-body refinement by Refmac
in CCP442 with an apo CTX-M-9 structure. CCP4 and Coot43 were
used to complete the model rebuilding and refinement.
ABBREVIATIONS USED
■
ESBL, extended spectrum β-lactamase; CTX-M, cefotaximase;
GPCR, G-protein coupled receptor; HTS, high-throughput
screening; vdw, van der Waals
Inhibition Assays. The hydrolysis reaction of CTX-M activity was
measured using the β-lactam substrate nitrocefin in 100 mM Tris-HCl
(pH 7.0, with 0.01% v/v Triton X-100) and monitored using a Biotek
Synergy Mx monochromator-based multimode microplate reader at
480 nM wavelength. Nitrocefin was 50 μM in the inhibition assays.
The Km of nitrocefin for CTX-M was determined to be 24 μM. The
compounds were synthesized as previously described or purchased
from the company Chembridge and assayed without further
purification. The highest concentrations at which the compounds
were tested were up to 1−3 mM (depending on their solubility) in the
IC50 experiment. The reaction was initiated by adding protein to the
reaction buffer last.
Microbiology. Susceptibility testing was performed and inter-
preted following the guidelines of CLSI.34 The compounds were
dissolved in DMSO, and dilutions were performed using Muller−
Hinton medium. An adequate final concentration was obtained to
determine the minimum inhibitory concentrations (MICs). The
concentration of DMSO was maintained below 10%. The inhibitors
were tested for synergy with the third-generation β-lactam cefotaxime
against two clinical bacteria. The ratio of β-lactam to inhibitor was 1:1.
Each value reported reflects the average of three independent
experiments. The bacteria belonged to the species Escherichia coli
and exhibited high level of resistance to β-lactams because of CTX-M-
9 and CTX-M-14 β-lactamases. For disk diffusion plate assays,
bacterial strains were diluted in sterile water to a turbidity equivalent
to 0.5 McFarland turbidity standards. After a subsequent 10-fold
dilution, the bacterial suspensions were inoculated on Mueller Hinton
agar. The plates were dried for 10 min before applying the disks
containing cefotaxime antibiotic (30 μg) or the inhibitor (30 μg) or
both (cefotaxime:inhibitor: 30 μg:15 μg, 30 μg:30 μg) or the solvent.
After overnight incubation at 37 °C, the zones of bacterial-growth
inhibition were measured.
REFERENCES
■
(1) Chen, Y.; Zhang, W.; Shi, Q.; Hesek, D.; Lee, M.; Mobashery, S.;
Shoichet, B. K. Crystal structures of penicillin-binding protein 6 from
Escherichia coli. J. Am. Chem. Soc. 2009, 131, 14345−14354.
(2) Llarrull, L. I.; Testero, S. A.; Fisher, J. F.; Mobashery, S. The
future of the beta-lactams. Curr. Opin. Microbiol. 2010, 13, 551−557.
(3) Tipper, D. J.; Strominger, J. L. Mechanism of action of penicillins:
a proposal based on their structural similarity to acyl-D-alanyl-D-
alanine. Proc. Natl. Acad. Sci. U.S.A. 1965, 54, 1133−1141.
(4) Silvaggi, N. R.; Anderson, J. W.; Brinsmade, S. R.; Pratt, R. F.;
Kelly, J. A. The crystal structure of phosphonate-inhibited D-Ala-D-Ala
peptidase reveals an analogue of a tetrahedral transition state.
Biochemistry 2003, 42, 1199−1208.
(5) Frere, J. M. Beta-lactamases and bacterial resistance to antibiotics.
Mol. Microbiol. 1995, 16, 385−395.
(6) Fisher, J. F.; Meroueh, S. O.; Mobashery, S. Bacterial resistance to
beta-lactam antibiotics: compelling opportunism, compelling oppor-
tunity. Chem. Rev. 2005, 105, 395−424.
(7) Taubes, G. The bacteria fight back. Science 2008, 321, 356−361.
(8) Bush, K.; Jacoby, G. A.; Medeiros, A. A. A functional classification
scheme for beta-lactamases and its correlation with molecular
structure. Antimicrob. Agents Chemother. 1995, 39, 1211−1233.
(9) Livermore, D. M. beta-Lactamases in laboratory and clinical
resistance. Clin. Microbiol. Rev. 1995, 8, 557−584.
(10) Chen, Y.; Delmas, J.; Sirot, J.; Shoichet, B.; Bonnet, R. Atomic
resolution structures of CTX-M beta-lactamases: extended spectrum
activities from increased mobility and decreased stability. J. Mol. Biol.
2005, 348, 349−362.
(11) Delmas, J.; Leyssene, D.; Dubois, D.; Birck, C.; Vazeille, E.;
Robin, F.; Bonnet, R. Structural insights into substrate recognition and
product expulsion in CTX-M enzymes. J. Mol. Biol. 2010, 400, 108−
120.
(12) Chen, Y.; Shoichet, B.; Bonnet, R. Structure, function, and
inhibition along the reaction coordinate of CTX-M beta-lactamases. J.
Am. Chem. Soc. 2005, 127, 5423−5434.
(13) Bonnet, R. Growing group of extended-spectrum beta-
lactamases: the CTX-M enzymes. Antimicrob. Agents Chemother.
2004, 48, 1−14.
(14) Bradford, P. A. Extended-spectrum beta-lactamases in the 21st
century: characterization, epidemiology, and detection of this
important resistance threat. Clin. Microbiol. Rev. 2001, 14, 933−951.
(15) Drawz, S. M.; Bonomo, R. A. Three decades of beta-lactamase
inhibitors. Clin. Microbiol. Rev. 2010, 23, 160−201.
(16) Bennett, P. M.; Chopra, I. Molecular basis of beta-lactamase
induction in bacteria. Antimicrob. Agents Chemother. 1993, 37, 153−
158.
ASSOCIATED CONTENT
■
Accession Codes
The coordinates and structure factors have been deposited in
the Protein Data Bank with the accession codes 4DE3, 4DDY,
4DDS, 4DE2, 4DE0, and 4DE1 for CTX-M-9 in complex with
4, 10, 11, 12, 16, and 18, respectively.
AUTHOR INFORMATION
■
Corresponding Author
*For A.R.R.: phone, (415) 514-9698; fax, (415) 482-1972; E-
mail, adam.renslo@ucsf.edu. For Y.C.: phone, (813) 974-7809;
fax, (813) 974-7357; E-mail, ychen1@health.usf.edu.
Present Addresses
(17) Jacobs, C.; Frere, J. M.; Normark, S. Cytosolic intermediates for
cell wall biosynthesis and degradation control inducible beta-lactam
resistance in gram-negative bacteria. Cell 1997, 88, 823−832.
(18) Petrosino, J.; Cantu, C. III; Palzkill, T. beta-Lactamases: protein
evolution in real time. Trends Microbiol. 1998, 6, 323−327.
(19) Pages, J. M.; Lavigne, J. P.; Leflon-Guibout, V.; Marcon, E.; Bert,
F.; Noussair, L.; Nicolas-Chanoine, M. H. Efflux pump, the masked
side of beta-lactam resistance in Klebsiella pneumoniae clinical isolates.
PLoS ONE 2009, 4, e4817.
∥Diablo Valley College, Department of Chemistry, Pleasant
Hill, California 94523, United States.
⊥Lebanese University, EDST, AZM Center for Biotechnology
Research and Applications, Tripoli, Lebanon.
Notes
The authors declare no competing financial interest.
(20) Bonnefoy, A.; Dupuis-Hamelin, C.; Steier, V.; Delachaume, C.;
Seys, C.; Stachyra, T.; Fairley, M.; Guitton, M.; Lampilas, M. In vitro
activity of AVE1330A, an innovative broad-spectrum non-beta-lactam
beta-lactamase inhibitor. J. Antimicrob. Chemother. 2004, 54, 410−417.
(21) Majumdar, S.; Pratt, R. F. Inhibition of class A and C beta-
lactamases by diaroyl phosphates. Biochemistry 2009, 48, 8285−8292.
ACKNOWLEDGMENTS
■
We thank Dr. Rongshi Li for insightful discussions and Eric
Lewandowski for reading the manuscript. This work was
supported by USF Start-up Fund (to Y.C.) and the Sandler
Foundation (to A.R.R.).
2171
dx.doi.org/10.1021/jm2014138 | J. Med. Chem. 2012, 55, 2163−2172

Journal of Medicinal Chemistry
Article
(22) Morandi, F.; Caselli, E.; Morandi, S.; Focia, P. J.; Blazquez, J.;
Shoichet, B. K.; Prati, F. Nanomolar inhibitors of AmpC beta-
lactamase. J. Am. Chem. Soc. 2003, 125, 685−695.
(42) Collaborative Computational Project, N. The CCP4 suite:
programs for protein crystallography. Acta Crystallogr., Sect. D: Biol.
Crystallogr. 1994, 50, 760−763.
(43) Emsley, P.; Cowtan, K. Coot: model-building tools for
molecular graphics. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2004,
60, 2126−2132.
(23) Strynadka, N. C.; Martin, R.; Jensen, S. E.; Gold, M.; Jones, J. B.
Structure-based design of a potent transition state analogue for TEM-1
beta-lactamase. Nature Struct. Biol. 1996, 3, 688−695.
(24) Tondi, D.; Morandi, F.; Bonnet, R.; Costi, M. P.; Shoichet, B. K.
Structure-based optimization of a non-beta-lactam lead results in
inhibitors that do not up-regulate beta-lactamase expression in cell
culture. J. Am. Chem. Soc. 2005, 127, 4632−4639.
(25) Chen, Y.; Shoichet, B. K. Molecular docking and ligand
specificity in fragment-based inhibitor discovery. Nature Chem. Biol.
2009, 5, 358−364.
(26) Chen, Y.; Pohlhaus, D. T. In silico docking and scoring of
fragments. Drug Discovery Today: Technol. 2010, 7, 149−156.
(27) Hert, J.; Irwin, J. J.; Laggner, C.; Keiser, M. J.; Shoichet, B. K.
Quantifying biogenic bias in screening libraries. Nature Chem. Biol.
2009, 5, 479−483.
(28) O’Shea, R.; Moser, H. E. Physicochemical properties of
antibacterial compounds: implications for drug discovery. J. Med.
Chem. 2008, 51, 2871−2878.
(29) Payne, D. J.; Gwynn, M. N.; Holmes, D. J.; Pompliano, D. L.
Drugs for bad bugs: confronting the challenges of antibacterial
discovery. Nature Rev. Drug Discovery 2007, 6, 29−40.
(30) Chen, Y.; Bonnet, R.; Shoichet, B. K. The acylation mechanism
of CTX-M beta-lactamase at 0.88 a resolution. J. Am. Chem. Soc. 2007,
129, 5378−5380.
(31) Lorber, D. M.; Shoichet, B. K. Hierarchical docking of databases
of multiple ligand conformations. Curr. Top. Med. Chem. 2005, 5,
739−749.
(32) Irwin, J. J.; Shoichet, B. K. ZINCa free database of
commercially available compounds for virtual screening. J. Chem. Inf.
Model. 2005, 45, 177−182.
(33) Ryckmans, T.; Edwards, M. P.; Horne, V. A.; Correia, A. M.;
Owen, D. R.; Thompson, L. R.; Tran, I.; Tutt, M. F.; Young, T. Rapid
assessment of a novel series of selective CB2 agonists using parallel
synthesis protocols: a lipophilic efficiency (LipE) analysis. Bioorg. Med.
Chem. Lett. 2009, 19, 4406−4409.
(34) Wayne, P. Performance Standards for Antimicrobial Suscept-
ibility Testing; 20th Informational Supplement. Clin. Lab. Stand. Inst.
2010, M100−S20, 1−23.
(35) Ibuka, A. S.; Ishii, Y.; Galleni, M.; Ishiguro, M.; Yamaguchi, K.;
Frere, J. M.; Matsuzawa, H.; Sakai, H. Crystal structure of extended-
spectrum beta-lactamase Toho-1: insights into the molecular
mechanism for catalytic reaction and substrate specificity expansion.
Biochemistry 2003, 42, 10634−10643.
(36) Tomanicek, S. J.; Blakeley, M. P.; Cooper, J.; Chen, Y.; Afonine,
P. V.; Coates, L. Neutron diffraction studies of a class A beta-lactamase
Toho-1 E166A/R274N/R276N triple mutant. J. Mol. Biol. 2010, 396,
1070−1080.
(37) Shimamura, T.; Ibuka, A.; Fushinobu, S.; Wakagi, T.; Ishiguro,
M.; Ishii, Y.; Matsuzawa, H. Acyl-intermediate structures of the
extended-spectrum class A beta-lactamase, Toho-1, in complex with
cefotaxime, cephalothin, and benzylpenicillin. J. Biol. Chem. 2002, 277,
46601−46608.
(38) Minasov, G.; Wang, X.; Shoichet, B. K. An ultrahigh resolution
structure of TEM-1 beta-lactamase suggests a role for Glu166 as the
general base in acylation. J. Am. Chem. Soc. 2002, 124, 5333−5340.
(39) Ke, W.; Sampson, J. M.; Ori, C.; Prati, F.; Drawz, S. M.; Bethel,
C. R.; Bonomo, R. A.; van den Akker, F. Novel insights into the mode
of inhibition of class A SHV-1 beta-lactamases revealed by boronic acid
transition state inhibitors. Antimicrob. Agents Chemother. 2011, 55,
174−183.
(40) Makovec, F.; Peris, W.; Revel, L.; Giovanetti, R.; Redaelli, D.;
Rovati, L. C. Antiallergic and cytoprotective activity of new N-
phenylbenzamido acid derivatives. J. Med. Chem. 1992, 35, 3633−
3640.
(41) Otwinowski, Z.; Minor, W. Processing of X-ray diffraction data
collected in oscillation mode. Methods Enzymol. 1997, 276, 307−326.
2172
dx.doi.org/10.1021/jm2014138 | J. Med. Chem. 2012, 55, 2163−2172