H2dpa derivatives containing pentadentate ligands: An acyclic adjuvant potentiates meropenem activity in vitro and in vivo against metallo-β-lactamase-producing Enterobacterales

Article abstract of DOI:10.1016/j.ejmech.2021.113702

The emergence and dissemination of metallo-β-lactamases (MBLs) producing Enterobacterales is a great concern for public health due to the limited therapeutic options. No MBL inhibitors are currently available in clinical practice. Herein, we synthesized a series of H₂dpa derivatives containing pentadentate-chelating ligands and evaluated their inhibitory activity against MBLs. Related compounds inhibited clinically relevant MBLs (Imipenemase, New Delhi metallo-β-lactamase (NDM) and Verona integron-encoded metallo-β-lactamase) with IC₅₀ values of 1–4.9 μM. In vitro, the most promising compounds, 5b and 5c, which had a chiral methyl at the acid adjacent to 5a, demonstrated potent synergistic activity against engineered strains, with fractional inhibitory concentration index values as low as 0.07–0.18. The addition of 5b and 5c restored meropenem efficacy against 42 MBL-producing Enterobacterales and Pseudomonas aeruginosa to satisfactory clinical levels. In addition, safety tests revealed that 5b/5c showed no toxicity in red blood cells, cell lines or mouse model. Further studies demonstrated that compounds 5b and 5c were non-competitive MBL inhibitors. In vivo compounds 5b and 5c potentiated meropenem efficacy and increased the survival rate from 0 to at least 83% in mice with sepsis caused by an NDM-1-positive clinical strain. The activity of the compounds exhibited consistency at the molecular, cellular, and in vivo levels. These data indicated that H₂dpa derivatives 5b and 5c containing pentadentate-chelating ligands may be worthy of further study.

Full text of DOI:10.1016/j.ejmech.2021.113702

European Journal of Medicinal Chemistry 224 (2021) 113702
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
H2dpa derivatives containing pentadentate ligands: An acyclic
adjuvant potentiates meropenem activity in vitro and in vivo against
metallo-b-lactamase-producing Enterobacterales
,
,
Fangfang Chen a 1, Mengmeng Bai a 1, Wentian Liu a, Hongtao Kong a, Tingting Zhang a,
Hong Yao b, En Zhang a *, Juan Du a **, Shangshang Qin a
,
,
, ***
a School of Pharmaceutical Sciences, Institute of Drug Discovery and Development, Key Laboratory of Advanced Pharmaceutical Technology, Ministry of
Education of China, Zhengzhou University, Zhengzhou, 450001, PR China
b College of Veterinary Medicine, Henan Agricultural University, Zhengzhou, 450046, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
The emergence and dissemination of metallo-b-lactamases (MBLs) producing Enterobacterales is a great
Received 31 March 2021
Received in revised form
27 June 2021
Accepted 12 July 2021
Available online 13 July 2021
concern for public health due to the limited therapeutic options. No MBL inhibitors are currently
available in clinical practice. Herein, we synthesized series of H2dpa derivatives containing
pentadentate-chelating ligands and evaluated their inhibitory activity against MBLs. Related compounds
inhibited clinically relevant MBLs (Imipenemase, New Delhi metallo- -lactamase (NDM) and Verona
integron-encoded metallo- -lactamase) with IC50 values of 1e4.9 M. In vitro, the most promising
a
b
b
m
compounds, 5b and 5c, which had a chiral methyl at the acid adjacent to 5a, demonstrated potent
synergistic activity against engineered strains, with fractional inhibitory concentration index values as
low as 0.07e0.18. The addition of 5b and 5c restored meropenem efficacy against 42 MBL-producing
Enterobacterales and Pseudomonas aeruginosa to satisfactory clinical levels. In addition, safety tests
revealed that 5b/5c showed no toxicity in red blood cells, cell lines or mouse model. Further studies
demonstrated that compounds 5b and 5c were non-competitive MBL inhibitors. In vivo compounds 5b
and 5c potentiated meropenem efficacy and increased the survival rate from 0 to at least 83% in mice
with sepsis caused by an NDM-1-positive clinical strain. The activity of the compounds exhibited con-
sistency at the molecular, cellular, and in vivo levels. These data indicated that H2dpa derivatives 5b and
5c containing pentadentate-chelating ligands may be worthy of further study.
Keywords:
Antibiotic resistance
Metallo-b-lactamase inhibitors
Picolinic acid
Enterobacterales
© 2021 Elsevier Masson SAS. All rights reserved.
1. Introduction
of carbapenem-resistant Enterobacterales (CRE) in China revealed
that 85.7% of CRE produce a variety of carbapenemases, indicating
Among
b
-lactam antibiotics, carbapenems are atypical and have
that carbapenemase is the most important mechanism conferring
carbapenem resistance among clinical Enterobacterales isolates [3].
The emergence and spread of carbapenemase-producing Enter-
obacterales (CPE) presents a global threat to public health because
of the limited therapeutic options [4].
the broadest antibacterial spectrum and strongest antibacterial
activity, earning them a vital role in the treatment of multi-drug
resistant Gram-negative bacterial infections [1]. However, carba-
penems can be hydrolyzed by carbapenemase, a
has a variety of hydrolysis capabilities and the ability to decompose
almost all -lactam antibiotics [2]. Recent nationwide surveillance
b-lactamase that
Two types of carbapenemases, namely class A serine
mases (SBLs) of the KPC type and the zinc-dependent class B
metallo- -lactamases (MBLs), comprising Verona integron-
encoded metallo- -lactamases (VIMs), imipenemases (IMPs) and
New Delhi metallo- -lactamases (NDMs), are regarded as the most
b-lacta-
b
b
b
b
* Corresponding author.
** Corresponding author.
*** Corresponding author.
important carbapenemases clinically in both Enterobacterales and
non-fermenting Gram-negative bacilli such as Acinetobacter spp.
[1]. Since the discovery of the SBL inhibitors sulbactam and tazo-
E-mail addresses: zhangen@zzu.edu.cn (E. Zhang), dujuan2004@zzu.edu.cn
(J. Du), qinshangshang@126.com (S. Qin).
bactam [5,6], the novel and effective non-b-lactam b-lactamase
1
These authors have contributed equally to this work.
https://doi.org/10.1016/j.ejmech.2021.113702
0223-5234/© 2021 Elsevier Masson SAS. All rights reserved.

F. Chen, M. Bai, W. Liu et al.
European Journal of Medicinal Chemistry 224 (2021) 113702
inhibitors avibactam and vaborbactam [7,8], have recently been
introduced into clinical practice. Avibactam and vaborbactam
possess broad efficacy against SBLs, particularly KPC, and are as a
promising treatment for non-MBL-producing CPE. Unfortunately,
no effective inhibitor for MBLs has been approved for clinical use.
The development of new antibiotics is progressing at a slower rate
than the emergence and spread of CPEs. Therefore, the study of new
inhibitors to combat MBL-producing pathogens is vital [9,10]. Un-
like SBLs, MBLs use Zn(II) as the active site that catalyzes the hy-
imaging [37]. Given this finding, a series of H2dpa derivatives were
synthesized and their inhibitory activities against three types of
MBLs (including NDM, IMP and VIM) were systematically charac-
terized. Of note, the H2dpa derivatives 5b and 5c were demon-
strated to be non-competitive MBL inhibitors which were capable
of restoring the susceptibility of meropenem (MEM) against clinical
isolates producing NDMs and IMP. In addition, the combined use of
compounds 5b/5c and MEM improved the survival of septic mice
and reduced the bacterial load in mouse organs in vivo. Taken
together, compounds 5b and 5c showed great potential as anti-
biotic adjuvants and are worthy of further research.
drolysis of
b-lactam antibiotics [11]. Thus, most of the MBL
inhibitors interact via coordination to active site zinc ion(s).
To date, multiple structural types of promising MBL inhibitors
have been developed, most of which act via active site metal che-
lation. Examples include the natural product Aspergillomarasmine
A (AMA, Fig. 1), an inhibitor of NDM-1 and VIM-2 [12]. Some
compounds with inhibitory effects on NDM-1, VIM-2 and IMP-1
that have been investigated include the hypertension drug L-
captopril and its isomer D-captopril [13,14]. Others include dipi-
colinic acid (Fig. 1, 1) [15] and cyclic boronic acids with similar
2. Results and discussion
2.1. Synthesis and characterization
H3dpaa (5a) (Fig. 2), with a chiral methyl at the adjacent position
of acid analogues 5b and enantiomer 5c, and its phosphoric acid
analogue 5d were designed and synthesized. To test the effects of
the ethyl acid group of 5a, the second amine 5e was used as a
control compound. The general synthetic route is shown in
Schemes 1 and 2. Compound 2, which was obtained by selective
reduction of a single ester of dimethyl pyridine-2,6-dicarboxylate,
reacted with PBr3 to obtain compound 3. Compounds 4a-4e were
then generated from 3 via a double N-alkylation of the primary
amine synthon 1a-1e that corresponds to the desired compound
(Scheme 1), followed by deprotection in a strong acid to obtain the
final products 5a-5e (Scheme 1). The primary amine synthon 1d
required to make 5d is not commercially available. Accordingly, this
compound was synthesized in a single step under mild conditions
via an aqueous-mediated catalyst-free aza-Michael addition of
diethyl vinyl phosphonate with ammonium hydroxide (see Sup-
porting Information). Of note, the expected compound 5e′ was not
obtained, but compound 5e without the propionic acid structure
was achieved (Scheme 2). This may be due to the reverse reaction of
Michael addition in the acid reaction system. All final compounds
were characterized by 1H NMR, 13C NMR and HRMS. All final
compounds possessed a purity of ꢀ 95% according to HPLC.
structures to the
b-lactam tetrahedral transition state (Fig. 1, 2)
[16e18], etc [19e22]. Besides, nonmetal binding inhibitors also
have been investigated, and these have inhibitory effect on VIM-2
[23]. However, none of these compounds has been approved for
clinical use. Most inhibitors currently reported are only evaluated
for one type of MBL, with more inhibitors having been evaluated for
NDM than for IMP and VIM [24e26]. The existing inhibitors show a
high level of enzyme activity but no inhibitory activity at the
cellular level; or enzyme activity and inhibitory activity at the
cellular level are relatively high but toxicity is also high; or there are
no in vivo data available [27e29]. Moreover, one of the issues faced
in the development of such inhibitors is the biased selectivity to-
ward different MBLs [12,30]. Therefore, safe and effective com-
pounds with good enzyme activity and high cellular and in vivo
levels against different MBLs are urgently needed.
In recent years, our research group has been committed to
discovering new high-efficiency and low-toxicity small molecule
antimicrobial peptide mimics [31,32] and inhibitors against NDM-1.
In previous studies, we reported some metal chelating agents such
as dithiocarbamates that effectively inhibit the enzyme activity of
NDM-1 [33,34]. Furthermore, 2-picolinic acid and its derivatives
have been evaluated as potential MBL inhibitors. In particular,
dipicolinic acid derivatives with different chelating ligands have
been reported to exhibit significant inhibitory activity against some
MBLs [15,27,35,36]. These include dipicolinic acid with tridentate
ligands (DPA, 2,6-pyridine dicarboxylate), which can inhibit MBLs,
and H2dedpa (Fig. 2) and its derivatives, which were reported by
our research group to exhibit significant inhibitory activity against
NDM-1. The next step is to try dipicolinic acid derivatives with
other different chelating ligands to discover more potential MBL
inhibitors. H2dpa (Fig. 2) derivatives containing pentadentate li-
gands showed good performance in positron emission tomography
2.2. Inhibition assays on MBLs
Compounds 5a,5b,5c, 5d and 5e were assessed to determine
their ability to inhibit MBLs (NDM-1, VIM-2 and IMP-1) at the
molecular level. Firstly, proteins of NDM-1, VIM-2 and IMP-1 were
expressed and purified for detection of the half-maximum inhibi-
tory concentration (IC50). The IC50 values of compounds (Table 1)
revealed that each compound had effective inhibitory activity
against all three clinically relevant MBLs (NDM-1, VIM-2 and IMP-
1). The functional group on the nitrogen in 5a-5e had little effect
on the inhibitory activity against these three enzymes. However,
EDTA and AMA had weaker activity against IMP-1, and L-captopril
Fig. 1. Structures of selected New Delhi metallo-b-lactamase-1 (NDM-1) inhibitors.
2

F. Chen, M. Bai, W. Liu et al.
European Journal of Medicinal Chemistry 224 (2021) 113702
Fig. 2. Structure of the selected metal chelator containing different ligands.
a
Scheme 1. Synthesis of compounds 5ae5da. Reagents and conditions: (i) NaBH4, CH3OH, r.t.; (ii) PBr3, CHCl3, 0ꢁC; (iii) K2CO3, CH3CN, 1, 60ꢁC; (iv) 6 M HCl, reflux.
a
Scheme 2. Synthesis of compounds 5ea. Reagents and conditions: (i) 6 M HCl, reflux.
3

F. Chen, M. Bai, W. Liu et al.
European Journal of Medicinal Chemistry 224 (2021) 113702
exhibited less effective inhibitory activity against NDM-1. There-
fore, the test compounds demonstrated broader enzyme inhibitory
activity than the positive control drugs.
with FIC index values of 0.07-0.078 (Fig. S3A II, III). Thus, these two
compounds (5b and 5c) were selected for further evaluation.
2.3.3. Combination studies with MEM against clinical MBL-
producing isolates
2.3. Antibacterial activity
A total of 50 blaMBL (blaNDM, blaIMP, blaVIM) and blaSBL (blaOXA
,
blaKPC) positive Gram-negative clinical bacilli, including 7 Entero-
bacter cloacae, 21 Escherichia coli, 17 Klebsiella pneumoniae, 2
Klebsiella oxytoca, 1 Citrobacter freundii, 1 Enterobacter hormaechei
and 1 Pseudomonas aeruginosa strains were selected for evaluation
of the antimicrobial activity of 5b/5c plus MEM (Table S3). The
study and consent procedures were approved by the Ethical Com-
mittee of Zhengzhou University. The results showed that these
compounds did not exert an inhibitory effect on these bacteria even
2.3.1. Generation of engineered stains (E. coli-DH5
cell-based screening assay
a/pUC-MBL) for a
The recombinant pUC19-based plasmid carrying the blaNDM-1
blaIMP-4/blaVIM-1 gene (Table S1-S2) was introduced into wild-type
Escherichia coli competent cells DH5 to obtain NDM-1/IMP-4/
/
a
VIM-1-producing E. coli strains that carried no other resistance
determinants. Antimicrobial susceptibility testing results showed
the minimum inhibitory concentrations (MICs) of MEM against
at a dose of 64 mg/mL. The definition of MIC90 used in this study was
the concentration at which 90% of the bacterial isolates exhibited
no growth after 16e20 h of incubated antimicrobial exposure [40].
E. coli-DH5
DH5 /pUC-VIM-1 were 32, 8 and 4
significantly higher than that of E. coli-DH5
a
/pUC-NDM-1, E. coli-DH5
g/mL, respectively, which were
(0.03125 g/mL)
a/pUC-IMP-4 and E. coli-
a
m
a
m
Significantly, in the presence of 5b, 5c or EDTA (16
mg/mL) the MIC90
(P < 0.05). This result indicated that the NDM-1/IMP-4/VIM-1-
of MEM was 0.25 g/mL, 0.125 g/mL and 2 g/mL, respectively, for
m
m
m
encoding gene had been successfully expressed in the recipient
clinical strains carrying NDM types NDM-1, NDM-3 or NDM-5 (No.
E. coli-DH5a cells. These engineered MBL-producing E. coli strains
1e31, No.40-41 in Table S4). In the presence of 5b, 5c or EDTA
were used in a cell-based inhibitor screening assay.
(16 mg/mL) the MIC90 of MEM was 1 mg/mL, 0.5 mg/mL and 2 mg/mL,
respectively, for clinical strains carrying IMP type IMP-4 (No. 32e41
in Table S4). In the cellular experiments for isolates carrying either
NDM or IMP, the inhibitory efficiency of 5b or 5c was better than
that of EDTA at the same concentration (Table S4). Meanwhile,
compound 5b/5c showed a better inhibitory effect for NDM than
IMP producing clinical strains. Such trends were commonly
observed regardless of the Enterobacterales species, which is
consistent with the results of NDM/IMP producing engineered
strains (Table S4). For high-level resistant strains carrying NDM-1,
the addition of compound 5b/5c reduced the MIC values of MEM
against some strains (No.10 and No.11 in Table S4) by 2048 times or
more. However, it was not active against clinical strains that pro-
duce SBL (KPC, OXA) (Table S5). These data demonstrated the
capability of compounds 5b and 5c to effectively potentiate the
MEM effect against MBL-harboring Gram-negative bacilli was
attributable to their inhibition of MBLs activity.
2.3.2. Determination of the FIC values of compounds 5a-5e when
interacting with E. coli-DH5
The synergistic activities of compounds 5a-5e in combination
with MEM against E. coli-DH5 -pUC19-MBL strains were initially
a/pUC-MBL
a
evaluated by an agar disc diffusion assay. EDTA which can inactivate
MBL by chelating out the Zn2þ ions, served as a positive control
[39]. As shown in Fig. S1 and Fig. S2, all five compounds restored
the inhibitory activity of MEM against MBL-producing strains. Next,
we performed a standard checkerboard study of the compounds in
combination with MEM against E. coli-DH5
The MICs of these compounds (5a-5e) alone were 128
a
-pUC19-MBL strains.
g/mL,
m
showing that these inhibitors have no antibacterial activity against
the above MBL-producing E. coli strains. In addition, the FIC index
values, which quantified the degree of interaction between anti-
biotic and adjuvant, were calculated as described in the experi-
mental methods, with FIC index < 0.5 depicting synergistic
interactions. The FIC index values of these compounds on DH5a/
pUC-IMP-4 were between 0.14 and 0.38 (Fig. S3B I-VI, reflecting a
2.4. A potent synergistic effect revealed by time-kill assays
certain inhibitory effect. These compounds had similar inhibitory
effects on E. coli-DH5a/pUC-NDM-1 (Fig. S3A I-VI) and E. coli-DH5a/
pUC-VIM-1 (Fig. S3C I-VI), and the FIC index values of the five
compounds were between 0.07 and 0.14. Moreover, the FIC index
values of compounds 5b and 5c were lower than for other com-
pounds (Fig. S3A II, III, Fig. S3B II, III and Fig. S3C II, III). At low
The analysis of bactericidal kinetics clearly reflected the dy-
namics of MEM combined with inhibitors on the inhibition of
bacterial growth in real time. As shown in the time-kill curves in
Fig. 3 I-III, the combination of MEM with 5b/5c dynamically
inhibited the growth of bacteria producing MBL. The addition of
16 mg/mL of 5b/5c to the NDM-1, IMP-4 or VIM-1-producing clinical
concentrations of 5b and 5c (16
mg/mL), the MIC value of MEM for
isolates exposed to an 8 ꢂ MIC of MEM for 8 h reduced the popu-
lation by more than 10,000-fold (Fig. 3 I-II) or to the lowest
detection value (Fig. 3 III). After 24 h, the number of bacteria in the
clinical isolates producing NDM-1, IMP-4 and VIM-1 was signifi-
cantly reduced to the lowest detection value or 102 CFU/mL.
Accordingly, the broth from the MEM and compound combination
group was clear in appearance, compared with a turbid appearance
for the control group and the compound and MEM alone groups
(Fig.S4 I-III).
E. coli-DH5 /pUC-NDM-1 decreased from 32 to < 0.03125
(more than a 1000-fold reduction), showing apparent synergism
a
m
g/mL
Table 1
IC50 values of the compounds against NDM-1, VIM-2 and IMP-1.
Compound
IC50 (m
M)a
NDM-1
IMP-1
VIM-2
5a
5b
5c
5d
1.67 0.12
1.60 0.02
1.69 0.11
1.65 0.08
1.54 0.09
4.0 1.0b
1.53 0.11
1.56 0.02
1.54 0.05
2.03 0.04
1.05 0.03
>200
3.88 0.26
4.40 0.01
4.17 0.12
4.84 0.29
1.04 0.03
9.6 2.4b
2.5. Hemolytic activity, toxicity and safety
5e
To determine the toxicity of compounds 5b and 5c, their toxicity
against red blood cells (RBCs) was initially tested in blood. Lower
hemolysis rates of RBCs indicate lower toxicity of the compound.
That is, a higher the half value of hemolysis (HC50) was taken to
indicate lower toxicity of the compound. No hemolysis (0%) and full
hemolysis (100%) were observed in the presence of PBS and Triton
AMA
EDTA
L-Captopril
1.75 0.16
>200
12.5 0.12
157.4 0.3c
23.3 1.3c
4.4 0.8c
a
Values reported as the mean SD of at least three independent experiments.
[12].
[38].
b
c
4

F. Chen, M. Bai, W. Liu et al.
European Journal of Medicinal Chemistry 224 (2021) 113702
Fig. 3. Time-dependent killing of bacterial by MEM combined with compounds 5b and 5c. (I) No.14 (blaNDM-1) strain (Table S4) was challenged with MEM (0.25
compounds 5b and 5c (16 g/mL) alone or in combination with MEM. (II) No.34 (blaIMP-4) strain (Table S4) was challenged with MEM (0.25 g/mL) or compounds 5b and 5c (16
mL) alone or in combination with MEM. (III) No.42 (blaVIM-1) (Table S4) strain was challenged with MEM (1 g/mL) or compounds 5b and 5c (16 g/mL) alone or in combination
mg/mL) or
m
m
mg/
m
m
with MEM. Microbial concentration: mean SD (n ¼ 3). Data represent the mean SD from three independent experiments. The symbol “” indicates the detection limit (10 CFU/
mL).
X-100, respectively [41,42]. The HC50 values of both compounds, 5b
and 5c, to RBCs compared with Triton X-100 treatment (HC50) were
kidney function. These results confirmed that compounds 5b and
5c were not toxic and were therefore deemed safe in vivo.
greater than 1024
mg/mL, which indicated that compounds 5b and
5c were non-hemolytic. Even at a value of 1024
portion of hemolysis induced by 5b/5c among RBCs was less than
7% (Fig. 4A).
m
g/mL, the pro-
2.6. Determination of Ki values
To further identify the inhibition model, the inhibition constants
Ki of these inhibitors against the NDM-1 enzyme were determined.
Next, a steady-state kinetic study was performed on NDM-1, IMP-1
and VIM-2 to explore the inhibition mode of compounds 5b and 5c
(Table 2). As shown in the Lineweaver-Burk diagram of MBL-
catalyzed substrate hydrolysis in Fig. 5, as the concentration of
compounds 5b (Fig. 5A I-III) and 5c (Fig. 5B I-III) gradually
increased, the addition of the compound resulted in a significant
decrease in Vmax, indicating that the compound exhibited a non-
competitive inhibitory effect. The inference of the Ki value is
similar to that of EDTA, AMA and other reported compounds [20].
Subsequently, we evaluated the toxicity of compounds 5b and
5c against mammalian cells. A LIVE/DEAD staining method was
employed in which green fluorescence represented live cells and
red fluorescence represented dead cells. The treated human gastric
epithelial (GES) cells exhibited almost exclusively green fluores-
cence at 32 mg/mL (5b and 5c) (Fig. 4B III-VI), similar to the negative
control (Fig. 4B I-II). However, cells treated with the positive control
(Triton-X) displayed almost exclusively red fluorescence (Fig. 4B
VII-VIII). These results indicated that the MBL inhibitors 5b and 5c
exhibited low toxicity against GES cells.
The acute toxicity of compounds 5b and 5c in vivo was deter-
mined after intraperitoneal (i.p.) administration of a single dose to
BALB/c mice. Different concentrations (50, 30, 20 and 10 mg/kg) of
compounds 5b and 5c were i.p. injected into the abdominal cavity
of BALB/c mice to observe the status and survival rate of the mice as
a preliminary examination of the toxicity of the compounds in vivo.
After injection of compounds 5b and 5c, the survival rates of the
mice at each concentration remained at 100%, consistent with the
normal saline group, within 72 h. This demonstrated that the
compounds were less toxic at a concentration of 50 mg/kg.
Next, after i.p. injection of the BALB/c mice with 50 mg/kg of
compounds 5b and 5c, we obtained blood samples for routine
testing and organ (liver, spleen and kidney) tissue sections for he-
matoxylin and eosin (HE) staining. As shown in Fig. 4C and 24 h
after injection, the liver, spleen, and kidneyshowed no obvious
pathological changes in the medication group compared with the
control group.
2.7. Testing of zinc homeostasis
To initially explore the mechanism of action of the compound,
we performed a zinc ion dependence experiment. When deter-
mining the IC50 values of compounds 5b and 5c on NDM-1, IMP-1
and VIM-2, we noticed that under different concentrations of zinc
sulfate (ranging from 0.1 to 100 mM), the IC50 values of the com-
pounds also differed (Fig. 6A). This zinc-dependent effect was
similar to that of EDTA. By contrast, the inhibitory activity of
captopril was not affected by changing the concentration of exog-
enous zinc added to the assay medium. These findings supported a
mechanism of action of compounds 5b and 5c based on zinc
chelation.
To explore the effects of zinc ions on enzyme activity further, we
carried out a zinc inhibition sensitivity experiment using com-
pounds 5b and 5c (8 mg/mL) and bacteria containing NDM-1, IMP-4
Routine blood analysis tested the white blood cell count (WBC),
red blood cell count (RBC), hemoglobin (HGB), hematocrit (HCT),
mean corpuscular volume (MCV) and platelet count (PLT). The
statistical results showed that compared with the normal saline
control group, no statistically significant changes in the blood
analysis data were detected following the i.p. injection of com-
pounds 5b and 5c into BALB/c mice (Fig. 4D). We also collected
blood samples for serum biochemical testing and analysis,
including albumin (ALB), urea (UREA) and creatinine (CREA).
Compared with the normal saline group (Fig. 4D I-IX), when
compounds 5b and 5c were i.p. injected into BALB/c mice, there was
no significant difference in the corresponding indexes of liver and
and VIM-1. Zinc ions were provided by the addition of zinc sulfate.
As shown in Fig. 6B I-III, when compounds 5b, 5c and MEM were
tested alone, the growth rate of bacteria was about 95%, and no
antibacterial activity was detected; however, when MEM was used
in combination with 5b or 5c, the growth of bacteria was reduced to
less than 10%. When the amount of zinc ions ranged from 1
10 M, the growth of bacteria was inhibited, whereas with a zinc
ion concentration of 100 M, almost 100% bacterial growth was
mM to
m
m
recorded. Furthermore, when the zinc ion concentration was higher
than that of compounds 5b or 5c, the zinc effect was observed.
These data indicated that the activity of compounds 5b and 5c was
affected by the presence of additional zinc ions.
5

F. Chen, M. Bai, W. Liu et al.
European Journal of Medicinal Chemistry 224 (2021) 113702
Fig. 4. Evaluation of the safety of the compounds in vivo and in vitro. (A) Hemolysis of red blood cells treated with different concentrations of compounds 5b and 5c. Mock: treated
with phosphate-buffered saline (PBS); Triton X: treated with 0.5% Triton X-100 detergent. Hemolysis rate was the mean SD (n ¼ 3). (B) Fluorescence microscopy images of GES
cells after treatment with compounds 5b and 5c for 24 h and staining with calcein AM and propidium iodide. (I) Non-treated live cells (negative control); (II) non-treated dead cells
(negative control); (III) live cells treated with compound 5b (32
mg/mL); (IV) dead cells treated with compound 5b (32 mg/mL); (V) live cells treated with compound 5c (32 mg/mL);
6

F. Chen, M. Bai, W. Liu et al.
European Journal of Medicinal Chemistry 224 (2021) 113702
Table 2
Ki determination of the test compounds.
2.8. Docking analysis
To verify the possible binding patterns between compounds 5b
and 5c and NDM-1 enzyme, MOE 2019 software was used to
analyze the interface between the inhibitors and the NDM-1 crystal
structure (PDB code 4EYL). The virtual docking results are shown in
Fig. 7. The dock scores (binding energy) of compounds 5b and 5c
with NDM-1 were -7.8031 and -7.9433 respectively. In compound
5b, the carbonyl of one carboxylic acid was bound to two Zn2þ ions
of NDM-1 via a coordinate bond, and the hydrogen on the other
carboxylic acid formed hydrogen bond with Gln123. One of the
Compound
Ki (m
M)a
NDM-1
VIM-2
IMP-1
5b
5c
1.59 0.03
1.70 0.13
6.16 0.25
5.83 0.52
1.57 0.01
1.54 0.06
a
Values reported as the mean SD of at least three independent experiments.
of the experiment,
7 days following infection. It is worth
mentioning that the MEM plus compound 5b combination treat-
ment resulted in 100% mouse survival. Effective coadministration of
MEM plus compound 5b/5c in mice therefore was translated into
in vivo efficacy. Furthermore, there are no safety concerns associ-
ated with this dual therapy under conditions in which in vivo effi-
cacy has been demonstrated.
pyridine groups formed as
p-p stacking interaction with His-250
(Fig. 7 I and II). In compound 5c, the hydroxyl of one of the car-
boxylic acids bound to the Zn2þ ions (ZN302) of NDM-1 via a co-
ordinate bond, and the other carboxylic acid formed two hydrogen
bonds with Gln-123 and Glu-152 (Fig. 7 III and IV). These results
supported the hypothesis that compounds 5b and 5c inhibit NDM-
1 by binding to the Zn2þ ion. Similar results were obtained in the
docking analysis of compounds 5b and 5c with IMP-1 (Fig. S5) and
VIM-2 (Fig. S6).
3. Experimental
3.1. General information
2.9. In vivo activities of compounds 5b and 5c
Chemicals were commercially available and were used without
further purification unless otherwise stated. All solvents were dried
according to standard procedures. The 1H NMR and 13C NMR
spectra were recorded at 25ꢁC in CDCl3, D2O, DMSO‑d6 or CD3OD at
600, 400 and 100 MHz, respectively, with TMS as the internal
Compound 5b/5c enhanced the efficacy of MEM on NDM-1
positive clinical CRE activity, indicating that compound 5b/5c may
restore the clinical efficacy of this antibiotic. To investigate this
further, a mouse sepsis model infected with NDM-1-positive
extensively drug resistant (XDR) No.12 strain (Table S4) [43] was
used to evaluate the potential clinical benefits of compounds 5b
and 5c combined with MEM against lethal infection in vivo (for
further data on No.12 strain please see Table S3). As shown in
Fig. 8A I-III, BALB/c mice were i.p. infected with a sub-lethal dose of
No.12 strain, and combination therapy with compound 5b/5c and
MEM showed a significant reduction of the bacterial load detected
in the liver, spleen and kidney compared with mice treated with
either agent separately. Interestingly, the combined administration
group showed improved treatment effects compared with mice
treated with either agent or physiological saline separately,
demonstrating that a combination of compound 5b/5c and MEM
has the therapeutic potential to combat the clinical challenge of
NDM-1-producing CRE.
The analysis of HE-stained sections revealed swollen and
degenerated liver cells and localized focal necrosis in the control
group (Fig. 8B). The central vein was dilated and congested, and
inflammatory cell infiltration was occasionally detected, with the
red pulp of the spleen mainly being infiltrated with neutrophils. In
the MEM (4 mg/kg) and compound 5b/5c (10 mg/kg) single-use
groups, the symptoms were similar to those in control group. In
the MEM (4 mg/kg) plus compound 5b/5c (10 mg/kg) combined
treatment group, a complete liver lobule structure was observed,
hepatic cords were arranged radially, the central veins were clearly
visible, and only a small amount of neutrophil infiltration was seen
in the red pulp of the spleen, with the liver and spleen showing
significant improvement.
standard. Chemical shifts (d) are expressed in ppm, and coupling
constants J are given in hertz. High-resolution mass spectra (HRMS)
data were obtained by ESI on a TOF mass analyzer. Analytical thin
layer chromatography (TLC) was performed on glass plates pre-
coated with silica gel to monitor the reactions. Column chroma-
tography was performed on silica gel (300e400 mesh). The purity
of each final compound was determined by analytical HPLC on a
Water 2695 unit with a C18 column (5
mm, 4.6 ꢂ 100 mm).
3.2. Synthesis
3.2.1. Methyl 6-(hydroxymethyl)picolinate (2) [37]
NaBH4 (7.75 g, 204.95 mmol) was slowly added to the solution of
dimethyl pyridine-2,6-dicarboxylate (10.00 g, 51.2 mmol) in MeOH
(250 mL) at 0ꢁC in 3 times. The reaction mixture was stirred for 7 h
at 0ꢁC. When the reaction was completed, saturated NaHCO3 so-
lution (50 mL) was added and extracted with dichloromethane
(200 mL ꢂ 3). The organic phase was washed with saturated
aqueous NaCl solution (10 mL ꢂ 3). The organic phases were dried
over anhydrous MgSO4, and evaporated to dryness to give white
solid 2 (7.01 g) which was used without purification. Yield: 82%.1H
NMR (400 MHz, DMSO‑d6)
1H), 7.74e7.69 (m, 1H), 5.57 (t, J ¼ 5.9 Hz, 1H), 4.62 (d, J ¼ 5.8 Hz,
2H), 3.87 (s, 3H). 13C NMR (100 MHz, DMSO‑d6)
165.3,162.5,146.4,
d
7.99 (t, J ¼ 7.7 Hz, 1H), 7.96e7.88 (m,
d
138.0, 123.8, 123.0, 64.0, 52.3. The spectral data are in accordance
with those reported in the literature.
After i.p. injection of the mice with a lethal dose of No.12 strain,
MEM and compound 5b/5c monotherapy and combination therapy
were initiated. Compound 5b/5c (10 mg/kg) or MEM (4 mg/kg)
monotherapy failed to rescue mice within 48 h (Fig. 8C). However,
MEM (4 mg/kg) and compound 5b/5c (10 mg/kg) combination
therapy rescued at least 83% of mice, who survived to the endpoint
3.2.2. Methyl 6-(bromomethyl)picolinate (3) [37]
PBr3 (11.08 g, 40.92 mmol) was added dropwise to the solution
of 2 (5.70 g, 34.10 mmol) in chloroform (60 mL) at 0ꢁC. The solution
was stirred at room temperature for 4 h. After the reaction was over,
saturated sodium bicarbonate was added to adjust pH to 8,
extracted with dichloromethane (200 mL ꢂ 4). The organic phases
(VI) dead cells treated with compound 5c (32 mg/mL); (VII) live cells treated with 0.05% Triton X-100 (positive control); (VIII) dead cells treated with 0.05% Triton X-100 (positive
control). Scale bar is 200
injection. Scale bar is 100
(n ¼ 3 for each group).
m
m
m. (C) A single dose of compound 5b (50 mg/kg), compound 5c (50 mg/kg) or physiological saline was administered to BALB/c mice by intraperitoneal
m. (D) Compound blood routine testing and blood biochemical indexes in mice. The data represent the standard errors of three independent experiments
7

F. Chen, M. Bai, W. Liu et al.
European Journal of Medicinal Chemistry 224 (2021) 113702
Fig. 5. (A) LineweavereBurk plots of inhibition of MBL hydrolysis activity by compound 5b. (B) LineweavereBurk plots of inhibition of MBL hydrolysis activity by compound 5c.
were washed with brine (20 mL ꢂ 3). The organic phases were
dried over anhydrous MgSO4, and evaporated to dryness to give
light yellow solid 3 (7.61 g). Yield: 97%. 1H NMR (400 MHz,
mixture was stirred for 12 h at 60ꢁC. After the reaction was over,
sodium carbonate was removed by filtration, and the crude reac-
tion mixture was concentrated by rotary evaporation. The crude oil
was purified by column chromatography (A: dichloromethane, B:
methanol, 100% A to 30% B gradient) to afford compound 4a-4e.
Yield: 39%e50%
DMSO‑d6)
d
8.1e8.0 (m, 2H), 7.8 (dd, J ¼ 7.4, 1.4 Hz, 1H), 4.8 (s, 2H),
3.9 (s, 3H). The spectral data are in accordance with those reported
in the literature.
3.2.3. General procedure for the synthesis compound 4a-4e [44]
K2CO3 (0.72 g, 5.21 mmol) and 1 (1.30 mmol) were added to the
solution of 3 (0.60 g, 2.60 mmol) in MeCN (6 mL). The reaction
3.2.3.1. Dimethyl
methylene))dipicolinate (4a) [37]. 4a: White solid, 1H NMR
(400 MHz, CDCl3)
8.0 (dd, J ¼ 7.5, 1.3 Hz, 2H), 7.9e7.8 (m, 4H), 4.2
6,6'-(((2-ethoxy-2-oxoethyl)azanediyl)bis(-
d
8

F. Chen, M. Bai, W. Liu et al.
European Journal of Medicinal Chemistry 224 (2021) 113702
Fig. 6. (A) The effect of the zinc ion content on the half inhibitory activity of compounds 5b and 5c on NDM-1, IMP-1 and VIM-2. (B) Metal suppression experiments: control,
without drugs; Zn2þ was provided by ZnSO4. a: compound treated with 1 M ZnSO4; b: compound treated with 10 M ZnSO4; c: compound treated with 100
M ZnSO4; E. coli-MBL
refer to E. coli DH5 /pUC-NDM-1, E. coli DH5 /pUC-IMP-4 and E. coli DH5 /pUC-VIM-1 respectively.
m
m
m
a
a
a
(q, J ¼ 7.1 Hz, 2H), 4.1 (s, 4H), 4.0 (s, 6H), 1.3 (t, J ¼ 7.1 Hz, 3H). 13
C
3.2.3.5. Dimethyl
methylene)) dipicolinate (4e) [37]. 4e: Light yellow oil, 1H NMR
(400 MHz, CDCl3)
6,6'-(((3-ethoxy-3-oxopropyl)azanediyl)bis(-
NMR (100 MHz, CDCl3) d 171.1,165.8,159.9,147.4, 137.5,126.3,123.8,
60.7, 60.0, 55.3, 52.9, 14.2. Yield: 41%. The spectral data are in
accordance with those reported in the literature.
d
8.0 (dd, J ¼ 7.5, 1.3 Hz, 2H), 7.8 (t, J ¼ 7.7 Hz, 2H),
7.8 (dd, J ¼ 7.7, 1.3 Hz, 2H), 4.1 (q, J ¼ 7.1 Hz, 2H), 4.0 (s, 6H), 4.0 (s,
4H), 3.0 (t, J ¼ 7.0 Hz, 2H), 2.6 (t, J ¼ 7.0 Hz, 2H), 1.2 (t, J ¼ 7.1 Hz, 3H).
13C NMR (100 MHz, CDCl3)
d 172.2, 165.8, 160.1, 147.3, 137.4, 126.0,
3.2.3.2. Dimethyl 6,6'-(((1-ethoxy-1-oxopropan-2-yl)azanediyl)bis(-
methylene)) (R)-dipicolinate (4b). 4b: White solid, 1H NMR
123.7, 60.4, 60.0, 52.9, 50.1, 32.8, 14.2. Yield: 47%. The spectral data
are in accordance with those reported in the literature.
(400 MHz, CDCl3)
d
8.0 (d, J ¼ 7.3 Hz, 2H), 7.9 (d, J ¼ 7.3 Hz, 2H), 7.8
(t, J ¼ 7.7 Hz, 2H), 4.2 (q, J ¼ 7.1 Hz, 3H), 4.1 (s, 4H), 4.0 (s, 6H), 3.6 (q,
J ¼ 7.1 Hz, 1H), 1.4 (d, J ¼ 7.1 Hz, 3H), 1.3 (t, J ¼ 7.1 Hz, 4H). 13C NMR
3.2.4. General procedure for the synthesis compound 5a-5e. [37]
(100 MHz, CDCl3) d 173.5,165.8,160.8,147.3,137.4,125.8,123.6, 60.7,
Compound 4 (500
mmol) was added 6 M hydrochloric acid
58.7, 57.1, 52.9, 14.3, 14.0. Yield: 45%. The spectral data are in
accordance with those reported in the literature.
(8 mL). The reaction mixture was refluxed for 4 h and evaporate
under vacuum to obtain product. The crude product was dissolved
in ethanol solvent, acetone was added until no solids precipitate,
and a white solid was obtained by filtration. Yield: 53%e73%.
3.2.3.3. Dimethyl 6,6'-(((1-ethoxy-1-oxopropan-2-yl)azanediyl)bis(-
methylene)) (S)-dipicolinate (4c) [45]. 4c: White solid, 1H NMR
(400 MHz, CDCl3)
d
8.0 (dd, J ¼ 7.5, 1.3 Hz, 2H), 7.9 (dd, J ¼ 7.9,1.3 Hz,
3.2.4.1. 6,6'-(((carboxymethyl)azanediyl)bis(methylene))dipicolinic
2H), 7.8 (t, J ¼ 7.7 Hz, 2H), 4.2 (q, J ¼ 7.1 Hz, 2H), 4.1 (s, 4H), 4.0 (s,
acid (5a) [44]. 5a: White solid, mp: 229e230ꢁC. 1H NMR (600 MHz,
6H), 3.6 (q, J ¼ 7.0 Hz,1H),1.4 (d, J ¼ 7.1 Hz, 3H),1.3 (t, J ¼ 7.1 Hz, 4H).
13C NMR (100 MHz, CDCl3)
d 173.5, 165.8, 160.8, 147.3, 137.4, 125.8,
D2O)
d
8.05e7.99 (m, 4H), 7.70 (dd, J ¼ 7.0, 1.8 Hz, 2H), 4.75 (s, 4H),
4.31 (s, 2H). 13C NMR (100 MHz, D2O)
d
169.6, 166.0, 150.5, 145.8,
123.6, 60.7, 58.7, 57.1, 52.9, 14.3, 14.0. Yield: 50%. The spectral data
are in accordance with those reported in the literature.
141.2, 128.5, 125.4, 58.5, 55.7. HRMS(ESIþ): calcd for C16H16N3O6
(M þ H)þ 346.1034, found 346.1040. Yield: 64%. The spectral data
are in accordance with those reported in the literature.
3.2.3.4. Dimethyl6,6'-(((2-(diethoxyphosphoryl)ethyl)azanediyl)bis(-
methylene)) dipicolinate (4d) [37]. 4d: Light gray solid, 1H NMR
(400 MHz, CDCl3
d
7.9 (d, J ¼ 8.8 Hz, 2H), 7.8 (d, J ¼ 4.9 Hz, 4H),
3.2.4.2. (R)-6,6'-(((1-carboxyethyl)azanediyl)bis(methylene))dipico-
4.0e3.9 (m, 4H), 3.9 (s, 6H), 3.9 (s, 4H), 2.9 (q, J ¼ 8.5 Hz, 2H),
linic acid (5b). 5b: White solid, mp: 231e232ꢁC. 1H NMR (600 MHz,
2.1e1.9 (m, 2H), 1.2 (t, J ¼ 7.0 Hz, 6H).13C NMR (100 MHz, CDCl3)
D2O)
d
7.9e7.8 (m, 4H), 7.6 (d, J ¼ 7.7 Hz, 2H), 4.6 (q, J ¼ 15.0 Hz, 4H),
d
165.7,159.9,147.4,137.5,126.1,123.8, 61.6 (d, J ¼ 6.5 Hz), 59.5, 52.9,
4.4 (q, J ¼ 7.2 Hz,1H),1.7 (d, J ¼ 7.2 Hz, 3H). 13C NMR (100 MHz, D2O)
47.8, 23.2 (d, J ¼ 137.8 Hz), 16.4 (d, J ¼ 6.1 Hz). Yield: 39%. The
spectral data are in accordance with those reported in the
literature.
d 173.1, 165.9, 151.2, 145.8, 141.2, 128.1, 125.0, 63.2, 56.0, 11.6.
HRMS(ESIþ): calcd for C17H18N3O6 (M þ H)þ 360.1190, found
360.1198. Yield: 73%.
9

F. Chen, M. Bai, W. Liu et al.
European Journal of Medicinal Chemistry 224 (2021) 113702
Fig. 7. Low energy conformations of compounds 5b and 5c docked into NDM-1 (PDB code 4EYL). Locally-enlarged stereoscopic view of compounds 5b (I) and 5c (II) bound to NDM-
1. The space represented by the enzyme is colored and the NDM-1 selected residues are displayed by different colored sticks (C, cyan; N, blue; O, red; H, gray). The compounds 5b
and 5c are shown as sticks colored by atoms (C, orange; N, blue; O, red; H, gray). The enzyme backbone is displayed as a cyan ribbon. Associations with protein residues are indicated
by dashed lines. 2D binding models of compounds 5b (III) and 5c (IV) in the active site.
3.2.4.3. (S)-6,6'-(((1-carboxyethyl)azanediyl)bis(methylene))dipico-
C
16H19N3O7P (M þ H)þ 396.0955, found 396.0963. Yield: 53%. The
linic acid (5c) [45]. 5c: White solid, mp: 233e234ꢁC. 1H NMR
spectral data are in accordance with those reported in the
literature.
(600 MHz, D2O)
d
7.9e7.8 (m, 4H), 7.6 (d, J ¼ 7.7 Hz, 2H), 4.6 (q,
J ¼ 15.0 Hz, 4H), 4.4 (q, J ¼ 7.2 Hz,1H),1.7 (d, J ¼ 7.2 Hz, 3H). 13C NMR
(100 MHz, D2O)
d 172.9, 165.9, 151.1, 145.8, 141.1, 128.1, 125.0, 63.2,
56.1, 11.6. HRMS(ESIþ): calcd for C17H18N3O6 (M þ H)þ 360.1190,
found 360.1199. Yield: 69%. The spectral data are in accordance with
those reported in the literature.
3.2.4.5. 6,6'-(azanediylbis(methylene))dipicolinic acid (5e) [37].
5e: Light gray solid, mp: 269e270ꢁC. 1H NMR (600 MHz, DMSO‑d6)
3.2.4.4. 6,6'-(((2-phosphonoethyl)azanediyl)bis(methylene))dipico-
linic acid (5d) [37]. 5d: Light yellow, mp: 226e227ꢁC. 1H NMR
d
13.1 (s, 2H), 10.1 (s, 2H), 8.2e8.0 (m, 4H), 7.8 (dd, J ¼ 7.5, 1.3 Hz,
(600 MHz, DMSO‑d6)
d
8.0 (dt, J ¼ 13.6, 7.6 Hz, 4H), 7.8 (d, J ¼ 7.4 Hz,
2H), 4.6 (s, 4H). 13C NMR (100 MHz, DMSO‑d6)
d 165.8, 152.4, 147.5,
2H), 4.8 (s, 4H), 3.5 (q, J ¼ 8.9, 8.0 Hz, 2H), 2.3 (dt, J ¼ 18.0, 7.3 Hz,
139.5, 127.4, 124.5, 50.5. HRMS(ESIþ): calcd for C14H14N3O4
(M þ H)þ 288.0979, found 288.0984. Yield: 70%. The spectral data
are in accordance with those reported in the literature.
2H). 13C NMR (100 MHz, DMSO‑d6)
d 165.7, 151.3, 147.9, 139.4, 128.5,
124.7, 56.9, 50.7, 23.5 (d, J ¼ 132.7 Hz). HRMS(ESIþ): calcd for
10

F. Chen, M. Bai, W. Liu et al.
European Journal of Medicinal Chemistry 224 (2021) 113702
Fig. 8. Compounds 5b and 5c rescue the activity of MEM in vivo. BALB/c mice were administered the NDM-1-positive No.12 strain (MEM MIC 32 mg/mL) by i.p. injection. Each group
of mice received a single dose of MEM (4 mg/kg) plus compound 5b or 5c (10 mg/kg) or saline. (A) For BALB/c mice injected with a sub-lethal dose (1 ꢂ 107 CFU) of No.12 strain, the
bacterial loads in the liver (I), spleen (II) and kidney (III) were detected. A single dose of MEM, compound 5b, compound 5c or MEM plus compound 5b and compound 5c, was
injected into the abdominal cavity for treatment. Data represent the mean and standard error from two separate experiments (n ¼ 6 per group). (B) HE staining results from BALB/c
mice injected with a sub-lethal dose (1 ꢂ 107 CFU) of No.12 strain. A single dose of MEM, compound 5b, compound 5c, or MEM plus compound 5b and compound 5c, was injected
into the abdominal cavity for treatment. (C) In the survival experiment, BALB/c mice were injected with a lethal dose (5 ꢂ 107 CFU) of No.12 strain. A single dose of MEM, MEM plus
compound 5b and compound 5c alone, or saline was injected for treatment. (n ¼ 6 per group) ****P < 0.0001. P values < 0.05 were considered statistically significant.
3.3. Biological activity
Spectramax reader (Molecular Devices) at 30ꢁC. The assay was
performed in 10 mM HEPES pH 7.5, 200 mL final volume.
3.3.1. IC50 enzyme inhibition assays
Enzyme (NDM-1, 50 nM; VIM-2, 500 pM; IMP-1, 50 nM) was
3.3.2. Expression and purification of MBLs enzyme
The pET-28a-NDM-1, VIM-2, IMP-1 plasmids were introduced
into the expression strains E. coli BL21(DE3), inoculated in 15 mL of
mixed with substrate (120
mM MEM for NDM-1; 20 mM Nitrocefin
(NIT) for VIM-2; 120 M Imipenem (IPM) for IMP-1) after a
m
5e10 min preincubation with compounds and linear portions of
curves were used to analyze data. Assays were read in 96-well
LB liquid medium containing 35 mg/mL kanamycin, and cultured at
37ꢁC with shaking for 12 h. Expand culture in 800 mL of LB con-
taining 35 mg/mL kanamycin at 37ꢁC. When OD600 reached 1.0, add
microplate format at 300 nm (486 nm for NIT) using
a
11

F. Chen, M. Bai, W. Liu et al.
European Journal of Medicinal Chemistry 224 (2021) 113702
0.5 mM per 100 mL transfer cultures of IPTG, 20ꢁC to cultivate the
16 h. After induction, centrifuged at 4600 g at 16ꢁC for 8 min, lysate
buffer (20 mM Tris pH ¼ 7.5, 300 mM NaCl) washing twice. Added
about 100 mL of lysate buffer to the bacterial pellet and blow evenly
biology Technology Co., Ltd, excellent grade GR), was suspended
with 1 ꢂ PBS and centrifuged (220 g, 10 min) to remove excess
impurities. The treated red blood cells were diluted and tested with
the resulting 5% erythrocyte suspension. DMSO or water as the
solvent (final concentration of no more than 0.5%) in each experi-
ment were hemolytic determination to prevent appearance of false
positive results. The Triton-X detergent at 0.5% was used as the
positive control. The 5% erythrocyte suspension was placed in a 96-
to a sterile mass. At the same time, 1/1000 of b-ME (protected di-
sulfide bond) and PMSF (inhibited protease) were added. Crushed
the cells and centrifuge for 50e60 min at 13500 g at 4ꢁC. Divided
the filtered liquid and grate into centrifuge tubes containing HisPur
Ni-NTA resin, and shook for 1 h at the lowest speed at 4ꢁC. The
mixture was allowed to stand for 30 min at 4ꢁC. Rinse the grate
with lysis solution (pH ¼ 7.5). Washed the contaminated protein
with washing buffer (lysate containing 10 mM imidazole), and then
purified the protein with elution buffer (containing 250 mM
imidazole). Run SDS-PAGE to detect the concentration and content
of protein. Then used desalting column for replacing MBLs,
measured the protein content, and store the enzyme at -80ꢁC.
well plate of 150
mL/well. The drug was diluted in a sterile EP tube
using 1 ꢂ PBS, 50
mL/well was added and each sample was tested in
parallel using three duplicate wells under the same concentration.
After incubation at 37ꢁC for 1 h, the plates were placed in a 4ꢁC
centrifuge (1150 g, 5 min).100 mL of the supernatant in each hole
was transferred to a new 96-well plate. Eventually, the results were
measured at 540 nm on the microplate reader. Percentage of he-
molytic was determined as (A - A0)/(Atotal - A0) ꢂ 100, and A was the
test well, A0 was the negative control, and Atotal was the positive.
3.3.3. Standard agar disc diffusion assay
The MIC was determined by the microdilution method accord-
ing to the Clinical and Laboratory Standards Institute (CLSI)
guidelines [46]. In short, spread 1 ꢂ 105 CFU of bacteria on the MH
plate, waited for the semi-dry sticker, and then dropped the mer-
3.3.7. Fluorescence microscopy
GES cells with good growth were tested. The cells were homo-
geneously cultured in 24-well plates, and the cells were incubated
for 24 h and then the test compound was added at the different
concentration. According to the mild destruction of the cell mem-
brane by 0.05% Triton-X, use the PI to detection the dead cells. The
reason is when the cells treated with the 0.05% Triton-X, PI can be
easily combined with the DNA in the nucleus. The PI-DNA complex
will release red fluorescence when the excitation and emission
wavelengths are 535 nm and 615 nm respectively. However, in the
living cells, the non-fluorescence of Calcium-AM is converted to
green fluorescence of calcium after the esterase of the cells is hy-
drolyzed by hydroxymethyl ester. There is no such process for dead
cells. Cells treated with 0.1% Triton-X was the positive control. After
openem (10 mg) and tested the compound (0.1 mM, 10 mL) on the
respective paper. The test plate was incubated at 37ꢁC for 16e20 h.
According to CLSI standards, when MEM was used as an antimi-
crobial compound, a single inhibition zone < 14 mm was resistant
and > 18 mm was sensitive.
3.3.4. Susceptibility testing
MICs were determined by the microdilution method according
to the CLSI guidelines [46]. In short, 1 ꢂ 105 CFU bacterial cells were
inoculated in 100
containing
mL of cation-adjusted Mueller Hinton broth
-lactam and appropriately diluted inhibitor. The final
24 h, the solution with 1 ꢂ PBS and stained with 2
mM Calcein-AM
b
(Sigma-Aldrich) or 4.5 M propidium iodide (PI, Aladdin) was
m
concentration of DMSO was kept below 1% in the measurement.
The tested plates were then incubated at 37ꢁC for 16e20 h. The MIC
value was determined as the lowest concentration of lactam tested
in the absence of visible bacterial growth. Use a microplate reader
to read the optical density at 600 nm to draw the heat maps. The FIC
index value was calculated using the following formula:
added and incubated for 15 min in a sterile environment. In the
end, the solution contains Calcein-AM: PI ¼ 1: 1. At last,1 ꢂ PBS was
used to wash the plate with the dead cells in the collection and
were added to the corresponding wells. Images were captured with
20 ꢂ objective in Nikon 901 fluorescence microscope.
MIC of
b
"TT5843c571"
MIC of "TT5843c571"
MIC of inhibitor with "TT5843c571"
MIC of inhibitor alone
"
ADlactam with inhibitor
ADlactam alone
ADlactam
3.3.8. Zinc Dependency assay
FIC index ¼
b
"
We added MBL enzyme (NDM-1, 50 nM; VIM-2, 500 pM; IMP-1,
50 nM) together with the dissolved compound to a 96-well plate
and incubated for 5 min according to the method as previously
b
"
þ
described [27]. After that, substrates (120
20 M NIT for VIM-2; 120 M IPM for IMP-1) containing different
final concentrations of zinc sulfate (0.1, 1, 10, and 100 M) were
mM MEM for NDM-1;
If the MIC of the individual inhibitor was > 128
concentration of “MIC of the individual inhibitor” was set to 128
mg/mL, the
m
m
m
g/
m
mL to determine the FIC index value. Unless otherwise stated, the
FIC index was defined as the lowest value among all obtained
values. Synergy was defined as the effect observed when the FIC
index value was ꢃ 0.5.
added to the wells, and the IC50 of compounds 5b and 5c were
evaluated.
The test of zinc homeostasis experiments were had done ac-
cording to the experiment of the influence of zinc ion content on
enzyme activity [47]. We carried out the zinc inhibition sensitivity
experiment of compounds 5b and 5c to bacteria containing NDM-1,
IMP-4 and VIM-1. Similarly, zinc ions were provided by adding zinc
sulfate. The experimental operation was the same except that
different amounts of ZnSO4 was added respectively. Eventually, the
results were measured at the OD600 nm of the microplate reader.
The test results were plotted as the mean SD of the three samples.
3.3.5. Time-dependent killing
A time-kill assay was performed to further determine the syn-
ergistic effect of compounds 5b and 5c with
according to the method as previously described [34]. Either MEM
(4 MIC) alone or 5b and 5c (16 g/mL) alone or a combination of the
b-lactam antibiotics
m
two was added to the bacterial solution, and then incubation was
performed at 37ꢁC. At 0, 2, 4, 6, 8, and 24 h after adding the agents,
an aliquot of bacterial culture was removed, diluted, and spotted on
MH agar plates to count the viable bacterial cells.
3.3.9. Ki value determination
Compounds 5b and 5c were used for in vitro inhibition deter-
mination of MBL. Mix MBL enzyme (NDM-1, 50 nM; VIM-2, 500
pM; IMP-1, 50 nM) with different concentrations of inhibitors 5b
and 5c, and incubate at 30ꢁC for 5 min. Add the corresponding
substrate (MEM corresponds to NDM-1 to make the concentration
3.3.6. Hemolytic activity
According to the method as previously described [34], fresh
sterile sheep blood (provided by the China Nanjing Bian Zhen
12

F. Chen, M. Bai, W. Liu et al.
European Journal of Medicinal Chemistry 224 (2021) 113702
reach 120
reach 120
centration reach 20
m
m
M, IPM corresponds to IMP-1 to make the concentration
M, and NIT corresponds to VIM-2 to make the con-
M) and monitor at 300 nm The hydrolysis rate
ligands were prepared and found to have inhibitory activity against
MBLs (NDM, VIM and IMP) at the molecular, cellular and in vivo
level. In vitro, compounds 5b and 5c had low FIC values against
m
of each substrate (NIT was 486 nm). All reactions were performed
in 10 mM HEPES (pH 7.5) buffer.
engineered strains (E. coli-DH5
a/pUC-NDM-1, E. coli-DH5a/pUC-
IMP-4 and E. coli-DH5 /pUC-VIM-1), and restored the activity of
a
MEM against clinical strains containing MBLs. In addition, com-
pounds 5b and its enantiomer 5c exhibited a low level of RBC he-
molysis and cytotoxicity, and were considered safe in mice. It was
speculated that the compound exerted its activity in a non-
competitive manner by chelating zinc ions. Compared with the
MEM group in the sepsis model, combination therapy with MEM
and 5b/5c was successful in reducing bacterial loads in the organs
and improving the survival rate in mice infected with NDM-1-
producing K. pneumoniae clinical isolates. Compound 5b com-
bined with MEM was particularly successful, resulting in a survival
rate of 100%. The results herein suggest that compounds 5b and 5c
warrant further investigation as a novel combination treatment.
3.3.10. Molecular docking studies
The protein complexes used for this study were obtained from
protein data bank (PDB code: 4EYL), and all water molecules were
eliminated. Hydrogen and partial charges were added by the pro-
tonate 3D program of MOE 2019; Energy minimization of the ligand
was carried out using energy minimize program of MOE. Default
parameter settings generated by the program of MOE were used for
docking.
3.3.11. In vivo safety animal experiment
Female BALB/c mice weighing 15e17 g, age 7e9 weeks, pur-
chased from Beijing Weitong Lihua Laboratory. Material Technology
Co., Ltd. (SCXK (Beijing) 2016-0006, Beijing, China).
Statement
The drug safety evaluation experiment was to determine the
acute toxicity of compound 5b and 5c in vivo by single-dose
intraperitoneal injection of mice, and tissue staining to check the
safety of the compounds [20]. In the acute toxicity experiment,
mice were divided into 9 groups, each with 6 mice, and different
concentrations of compounds were injected intraperitoneally to
observe the 72-h survival rate and survival status of the mice.
The tissue staining experiment was to divide the mice into 3
groups, 3 in each group, control (normal saline, intraperitoneal
injection), 5b (50 mg/kg, intraperitoneal injection) and compound
5c (50 mg/kg, intraperitoneal injection). 24 h after injection, these
mice were euthanized and dissected. The organs (liver, spleen and
kidney) of the mice were sectioned and stained with HE.
The blood routine experiment is to divide the mice into 3
groups, 3 in each group, control (normal saline, intraperitoneal
injection), 5b (50 mg/kg, intraperitoneal injection) and compound
5c (50 mg/kg, intraperitoneal injection). 24 h after the injection,
blood was taken for routine testing in each group.
The studies involving human participants were reviewed and
approved by the Ethical Committee of Zhengzhou University.
Written informed consent to participate in this study was provided
by the participants' legal guardian/next of kin. All animal studies
were approved by the Life Science Ethics Review Committee of
Zhengzhou University, and were conducted in compliance with
European guidelines for the care and use of laboratory animals, all
experiments were performed in a manner that minimized animal
suffering.
Declaration of competing interest
The authors declare that they have no known competing
financial interests or personal relationships that could have
appeared to influence the work reported in this paper.
Acknowledgments
3.3.12. In vivo sepsis animal experiment
This research was financially supported by the National Natural
Female BALB/c mice weighing 15e17 g, age 7e9 weeks, pur-
chased from Beijing Weitong Lihua Laboratory. Material Technology
Co., Ltd. (SCXK (Beijing) 2016-0006, Beijing, China).
Science Foundation of China (Nos. 21702190).
Appendix A. Supplementary data
We established a mouse sepsis model to evaluate the thera-
peutic effects of compounds 5b and 5c [12]. For this organ loading
experiment, the mice were randomly divided into 9 mice in each
group. A sublethal dose of K. pneumoniae (1 ꢂ 107 CFU) was given to
BALB/c mice by intraperitoneal injection. Thirty minutes after the
injection, a single dose of MEM (4 mg/kg), a single dose of com-
pound (10 mg/kg) and a combination of MEM (4 mg/kg) plus
compound (10 mg/kg) were treated, and the state of the mice was
observed for 2 days, And dissected the mice to take out the liver,
spleen and kidney, and counted the homogenized organs of 6 mice.
The other 3 mice were sent tissue sections for HE staining.
For this survival rate experiment, we randomly divided the mice
Supplementary data related to this article can be found at
https://doi.org/10.1016/j.ejmech.2021.113702.
References
[1] D. Hammoudi Halat, C. Ayoub Moubareck, The current burden of carbapene-
mases: Review of significant properties and dissemination among gram-
negative bacteria, Antibiotics 9 (2020) 186, 186.
[2] L.Y. Hsu, A. Apisarnthanarak, E. Khan, N. Suwantarat, A. Ghafur, P. Tambyah,
Carbapenem-resistant Acinetobacter baumannii and enterobacteriaceae in
south and southeast asia, Clin. Microbiol. Rev. 30 (2017) 1e22.
[3] Q. Wang, X. Wang, J. Wang, P. Ouyang, C. Jin, R. Wang, Y. Zhang, L. Jin, H. Chen,
Z. Wang, F. Zhang, B. Cao, L. Xie, K. Liao, B. Gu, C. Yang, Z. Liu, X. Ma, L. Jin,
X. Zhang, S. Man, W. Li, F. Pei, X. Xu, Y. Jin, P. Ji, H. Wang, Phenotypic and
genotypic characterization of carbapenem-resistant enterobacteriaceae: data
from a longitudinal large-scale CRE study in China (2012-2016), Clin. Infect.
Dis. 67 (2018) S196es205.
[4] R. Han, Q. Shi, S. Wu, D. Yin, M. Peng, D. Dong, Y. Zheng, Y. Guo, R. Zhang, F. Hu,
Dissemination of carbapenemases (KPC, NDM, OXA-48, IMP, and VIM) among
carbapenem-resistant enterobacteriaceae isolated from adult and children
patients in China, Front. Cell. Infect. Microbiol. 10 (2020) 314.
into
6 cages per group. The lethal dose of K. pneumoniae
(5 ꢂ 107 CFU) was given to BALB/c mice by intraperitoneal injection.
After 30 min of injection, a single dose of MEM (4 mg/kg), a single
dose of compound (10 mg/kg), and a combination of MEM (4 mg/
kg) plus compound (10 mg/kg) were treated, and the survival of
mice was observed for 7 days state and calculate the survival rate to
draw a survival curve.
[5] M. Akova, Sulbactam-containing beta-lactamase inhibitor combinations, Clin.
Microbiol. Infect. 14 (2008) 185e188.
[6] R.C. Xamplas, G.S. Itokazu, R.C. Glowacki, A.E. Grasso, C. Caquelin,
D.N. Schwartz, Implementation of an extended-infusion piperacillin-tazo-
bactam program at an urban teaching hospital, Am. J. Health Syst. Pharm. 67
(2010) 622e628.
4. Conclusions
A series of H2dpa derivatives containing pentadentate chelating
13

F. Chen, M. Bai, W. Liu et al.
European Journal of Medicinal Chemistry 224 (2021) 113702
[7] D. van Duin, R.A. Bonomo, Ceftazidime/avibactam and ceftolozane/tazo-
metallo-beta-lactamases, Br. J. Pharmacol. 176 (2019) 4548e4557.
[27] K. Tehrani, N.C. Bruchle, N. Wade, V. Mashayekhi, D. Pesce, M.J. van Haren,
N.I. Martin, Small molecule carboxylates inhibit metallo-beta-lactamases and
resensitize carbapenem-resistant bacteria to meropenem, ACS Infect. Dis. 6
(2020) 1366e1371.
bactam: second-generation
b-Lactam/b-Lactamase inhibitor combinations,
Clin. Infect. Dis. 63 (2016) 234e241.
[8] G.G. Zhanel, C.K. Lawrence, H. Adam, F. Schweizer, S. Zelenitsky, M. Zhanel,
ꢀ
P.R.S. Lagace-Wiens, A. Walkty, A. Denisuik, A. Golden, A.S. Gin, D.J. Hoban,
J.P. Lynch 3rd, J.A. Karlowsky, Imipenem-Relebactam and meropenem-
vaborbactam: two novel carbapenem-b-lactamase inhibitor combinations,
Drugs 78 (2018) 65e98.
[28] P.W. Kang, J.P. Su, L.Y. Sun, H. Gao, K.W. Yang, 3-Bromopyruvate as a potent
covalently reversible inhibitor of New Delhi metallo-beta-lactamase-1 (NDM-
1), Eur. J. Pharmaceut. Sci. 142 (2020) 105161.
[9] K. Bush, P.A. Bradford, Interplay between
inhibitors, Nat. Rev. Microbiol. 17 (2019) 295e306.
[10] Y.H. Yan, G. Li, G.B. Li, Principles and current strategies targeting metallo-
b
-lactamases and new
b
-lactamase
[29] C. Schnaars, G. Kildahl-Andersen, A. Prandina, R. Popal, S. Radix, M. Le Borgne,
T. Gjoen, A.M.S. Andresen, A. Heikal, O.A. Okstad, C. Frohlich, O. Samuelsen,
S. Lauksund, L.P. Jordheim, P. Rongved, O.A.H. Astrand, Synthesis and pre-
clinical evaluation of TPA-based zinc chelators as metallo-beta-lactamase in-
hibitors, ACS Infect. Dis. 4 (2018) 1407e1422.
b-
lactamase mediated antibacterial resistance, Med. Res. Rev. 40 (2020)
1558e1592.
[11] S.E. Boyd, D.M. Livermore, D.C. Hooper, W.W. Hope, Metallo-
b-Lactamases:
[30] J.-i. Wachino, W. Jin, K. Kimura, H. Kurosaki, A. Sato, Y. Arakawa, Sulfamoyl
structure, function, epidemiology, treatment options, and the development
pipeline, Antimicrob. Agents Chemother. 64 (2020) e00397-00320.
heteroarylcarboxylic acids as promising metallo-b-lactamase inhibitors for
controlling bacterial carbapenem resistance, mBio 11 (2020) e03144-03119.
[31] W. Chu, Y. Yang, J. Cai, H. Kong, M. Bai, X. Fu, S. Qin, E. Zhang, Synthesis and
bioactivities of new membrane-active agents with aromatic linker: high
selectivity and broad-spectrum antibacterial activity, ACS Infect. Dis. 5 (2019)
1535e1545.
[32] W. Chu, Y. Yang, S. Qin, J. Cai, M. Bai, H. Kong, E. Zhang, Low-toxicity
amphiphilic molecules linked by an aromatic nucleus show broad-spectrum
antibacterial activity and low drug resistance, Chem. Commun. 55 (2019)
4307e4310.
[12] A.M. King, S.A. Reid-Yu, W. Wang, D.T. King, G. De Pascale, N.C. Strynadka,
T.R. Walsh, B.K. Coombes, G.D. Wright, Aspergillomarasmine A overcomes
metallo-beta-lactamase antibiotic resistance, Nature 510 (2014) 503e506.
[13] F.-M. Klingler, T.A. Wichelhaus, D. Frank, J. Cuesta-Bernal, J. El-Delik,
€
H.F. Müller, H. Sjuts, S. Gottig, A. Koenigs, K.M. Pos, D. Pogoryelov, E. Proschak,
Approved drugs containing thiols as inhibitors of metallo-b-lactamases:
strategy to combat multidrug-resistant bacteria, J. Med. Chem. 58 (2015)
3626e3630.
[14] Y. Yusof, D.T.C. Tan, O.K. Arjomandi, G. Schenk, R.P. McGeary, Captopril ana-
[33] M.M. Wang, W.C. Chu, Y. Yang, Q.Q. Yang, S.S. Qin, E. Zhang, Di-
logues as metallo-
b
-lactamase inhibitors, Bioorg. Med. Chem. Lett 26 (2016)
thiocarbamates: efficient metallo-b-lactamase inhibitors with good antibac-
1589e1593.
terial activity when combined with meropenem, Bioorg. Med. Chem. Lett 28
(2018) 3436e3440.
[15] A.Y. Chen, P.W. Thomas, A.C. Stewart, A. Bergstrom, Z. Cheng, C. Miller,
C.R. Bethel, S.H. Marshall, C.V. Credille, C.L. Riley, R.C. Page, R.A. Bonomo,
M.W. Crowder, D.L. Tierney, W. Fast, S.M. Cohen, Dipicolinic acid derivatives
as inhibitors of New Delhi metallo-
7267e7283.
[16] J. Brem, R. Cain, S. Cahill, M.A. McDonough, I.J. Clifton, J.-C. Jimenez-Castella-
[34] E. Zhang, M.M. Wang, S.C. Huang, S.M. Xu, D.Y. Cui, Y.L. Bo, P.Y. Bai, Y.G. Hua,
C.L. Xiao, S. Qin, NOTA analogue: a first dithiocarbamate inhibitor of metallo-
beta-lactamases, Bioorg. Med. Chem. Lett 28 (2018) 214e221.
[35] X.F. Shi, M.M. Wang, S.C. Huang, J.X. Han, W.C. Chu, C. Xiao, E. Zhang, S. Qin,
H2depda: an acyclic adjuvant potentiates meropenem activity in vitro against
metallo-beta-lactamase-producing Enterobacterales, Eur. J. Med. Chem. 167
(2019) 367e376.
b
-lactamase-1, J. Med. Chem. 60 (2017)
ꢀ
nos, M.B. Avison, J. Spencer, C.W.G. Fishwick, C.J. Schofield, Structural basis of
metallo-b-lactamase, serine-b-lactamase and penicillin-binding protein inhi-
bition by cyclic boronates, Nat. Commun. 7 (2016) 12406.
[17] A. Krajnc, J. Brem, P. Hinchliffe, K. Calvopina, T.D. Panduwawala, P.A. Lang,
[36] D.Y. Cui, Y. Yang, M.M. Bai, J.X. Han, C.C. Wang, H.T. Kong, B.Y. Shen, D.C. Yan,
C.L. Xiao, Y.S. Liu, E. Zhang, Systematic research of H2dedpa derivatives as
potent inhibitors of New Delhi metallo-beta-lactamase-1, Bioorg. Chem. 101
(2020) 103965.
~
J.J.A.G. Kamps, J.M. Tyrrell, E. Widlake, B.G. Saward, T.R. Walsh, J. Spencer,
C.J. Schofield, Bicyclic boronate VNRX-5133 inhibits metallo- and serine-
lactamases, J. Med. Chem. 62 (2019) 8544e8556.
b-
ꢀ
[37] D.M. Weekes, C.F. Ramogida, M.d.G. Jaraquemada-Pelaez, B.O. Patrick, C. Apte,
[18] B. Liu, R.E.L. Trout, G.-H. Chu, D. McGarry, R.W. Jackson, J.C. Hamrick,
D.M. Daigle, S.M. Cusick, C. Pozzi, F. De Luca, M. Benvenuti, S. Mangani, J.-
D. Docquier, W.J. Weiss, D.C. Pevear, L. Xerri, C.J. Burns, Discovery of tani-
T.I. Kostelnik, J.F. Cawthray, L. Murphy, C. Orvig, Dipicolinate complexes of
gallium(III) and lanthanum(III), Inorg. Chem. 55 (2016) 12544e12558.
[38] J. Brem, S.S. van Berkel, D. Zollman, S.Y. Lee, O. Gileadi, P.J. McHugh,
borbactam (VNRX-5133): a broad-spectrum serine- and metallo-
b
-lactamase
T.R. Walsh, M.A. McDonough, C.J. Schofield, Structural basis of metallo-b-
inhibitor for carbapenem-resistant bacterial infections, J. Med. Chem. 63
(2020) 2789e2801.
lactamase inhibition by captopril stereoisomers, Antimicrob. Agents Chemo-
ther. 60 (2016) 142e150.
[19] Y. Xiang, Y.-J. Zhang, Y. Ge, Y. Zhou, C. Chen, W.Y. Wahlgren, X. Tan, X. Chen,
K.-W. Yang, Kinetic, thermodynamic, and crystallographic studies of 2-
[39] N. Aoki, Y. Ishii, K. Tateda, T. Saga, S. Kimura, Y. Kikuchi, T. Kobayashi,
Y. Tanabe, H. Tsukada, F. Gejyo, K. Yamaguchi, Efficacy of calcium-EDTA as an
triazolylthioacetamides as Verona integron-encoded metallo-
(VIM-2) inhibitor, Biomolecules 10 (2020) 72.
[20] C. Chen, K. Yang, Ruthenium complexes as prospective inhibitors of metallo-
beta-lactamases to reverse carbapenem resistance, Dalton Trans. 49 (2020)
14099e14105.
[21] W.B. Jin, C. Xu, Q. Cheng, X.L. Qi, W. Gao, Z. Zheng, E.W.C. Chan, Y.-C. Leung,
T.H. Chan, K.-Y. Wong, S. Chen, K.-F. Chan, Investigation of synergistic anti-
microbial effects of the drug combinations of meropenem and 1,2-
benzisoselenazol-3(2H)-one derivatives on carbapenem-resistant enterobac-
teriaceae producing NDM-1, Eur. J. Med. Chem. 155 (2018) 285e302.
[22] W.B. Jin, C. Xu, Q. Cheung, W. Gao, P. Zeng, J. Liu, E.W.C. Chan, Y.C. Leung,
T.H. Chan, K.Y. Wong, S. Chen, K.F. Chan, Bioisosteric investigation of ebselen:
synthesis and in vitro characterization of 1,2-benzisothiazol-3(2H)-one de-
b
-lactamase 2
inhibitor for metallo-b-lactamase in a mouse model of Pseudomonas aerugi-
nosa pneumonia, Antimicrob. Agents Chemother. 54 (2010) 4582e4588.
[40] B. Kulengowski, D.S. Burgess, Imipenem/relebactam activity compared to
other antimicrobials against non-MBL-producing carbapenem-resistant
enterobacteriaceae from an academic medical center, Pathog Dis 77 (2019).
[41] J. Hoque, M.M. Konai, S. Gonuguntla, G.B. Manjunath, S. Samaddar,
V. Yarlagadda, J. Haldar, Membrane active small molecules show selective
broad spectrum antibacterial activity with No detectable resistance and
eradicate biofilms, J. Med. Chem. 58 (2015) 5486e5500.
[42] K.B. Ishnava, P.P. Shah, Anticariogenic and hemolytic activity of selected seed
protein extracts In vitro conditions, J. Dent. 11 (2014) 576e586.
[43] S. Qin, C. Zhang, S. Schwarz, L. Li, H. Dong, H. Yao, X.D. Du, Identification of a
novel conjugative mcr-8.2-bearing plasmid in an almost pan-resistant
hypermucoviscous Klebsiella pneumoniae ST11 isolate with enhanced viru-
lence, J. Antimicrob. Chemother. 75 (2020) 2696e2699.
rivatives as potent New Delhi metallo-
100 (2020) 103873.
b-lactamase inhibitors, Bioorg. Chem.
ꢁ
€
ꢀ
[23] G.-B. Li, M.I. Abboud, J. Brem, H. Someya, C.T. Lohans, S.-Y. Yang, J. Spencer,
[44] T.W. Price, J. Gallo, V. Kubícek, Z. Bohmova, T.J. Prior, J. Greenman, P. Hermann,
G.J. Stasiuk, Amino acid based gallium-68 chelators capable of radiolabeling at
neutral pH, Dalton Trans. 46 (2017) 16973e16982.
D.W. Wareham, M.A. McDonough, C.J. Schofield, NMR-filtered virtual
screening leads to non-metal chelating metallo-b-lactamase inhibitors, Chem.
Sci. 8 (2017) 928e937.
[45] M. Khannam, S.K. Sahoo, C. Mukherjee, Effect of ligand chirality and hyper-
conjugation on the thermodynamic stability of a tris(aquated) GdIII complex:
synthesis, characterization, and T1-weighted phantom MR image study, Eur. J.
Inorg. Chem. (2019) 2518e2523, 2019.
[46] CLSI, Performance Standards for Antimicrobial Susceptibility Testing; 29th
Informational Supplement. CLSI Document M100, CLSI, Wayne, PA, 2019.
[47] S.B. Falconer, S.A. Reid-Yu, A.M. King, S.S. Gehrke, W. Wang, J.F. Britten,
B.K. Coombes, G.D. Wright, E.D. Brown, Zinc chelation by a small-molecule
adjuvant potentiates meropenem activity in vivo against NDM-1-producing
Klebsiella pneumoniae, ACS Infect. Dis. 1 (2015) 533e543.
[24] Y. Wang, X. Sun, F. Kong, L. Xia, X. Deng, D. Wang, J. Wang, Specific NDM-1
inhibitor of isoliquiritin enhances the activity of meropenem against NDM-
1-positive Enterobacteriaceae in vitro, Int. J. Environ. Res. Publ. Health 17
(2020) 2162.
[25] Y. Liu, K. Yang, Y. Jia, Z. Wang, Repurposing peptidomimetic as potential in-
hibitor of New Delhi metallo-beta-lactamases in gram-negative bacteria, ACS
Infect. Dis. 5 (2019) 2061e2066.
[26] S. Liu, J. Zhang, Y. Zhou, N. Hu, J. Li, Y. Wang, X. Niu, X. Deng, J. Wang, Pter-
ostilbene restores carbapenem susceptibility in New Delhi metallo-beta-
lactamase-producing isolates by inhibiting the activity of New Delhi
14