4-Alkyl-1,2,4-triazole-3-thione analogues as metallo-β-lactamase inhibitors

Article abstract of DOI:10.1016/j.bioorg.2021.105024

In Gram-negative bacteria, the major mechanism of resistance to β-lactam antibiotics is the production of one or several β-lactamases (BLs), including the highly worrying carbapenemases. Whereas inhibitors of these enzymes were recently marketed, they only target serine-carbapenemases (e.g. KPC-type), and no clinically useful inhibitor is available yet to neutralize the class of metallo-β-lactamases (MBLs). We are developing compounds based on the 1,2,4-triazole-3-thione scaffold, which binds to the di-zinc catalytic site of MBLs in an original fashion, and we previously reported its promising potential to yield broad-spectrum inhibitors. However, up to now only moderate antibiotic potentiation could be observed in microbiological assays and further exploration was needed to improve outer membrane penetration. Here, we synthesized and characterized a series of compounds possessing a diversely functionalized alkyl chain at the 4-position of the heterocycle. We found that the presence of a carboxylic group at the extremity of an alkyl chain yielded potent inhibitors of VIM-type enzymes with K_i values in the μM to sub-μM range, and that this alkyl chain had to be longer or equal to a propyl chain. This result confirmed the importance of a carboxylic function on the 4-substituent of 1,2,4-triazole-3-thione heterocycle. As observed in previous series, active compounds also preferentially contained phenyl, 2-hydroxy-5-methoxyphenyl, naphth-2-yl or m-biphenyl at position 5. However, none efficiently inhibited NDM-1 or IMP-1. Microbiological study on VIM-2-producing E. coli strains and on VIM-1/VIM-4-producing multidrug-resistant K. pneumoniae clinical isolates gave promising results, suggesting that the 1,2,4-triazole-3-thione scaffold worth continuing exploration to further improve penetration. Finally, docking experiments were performed to study the binding mode of alkanoic analogues in the active site of VIM-2.

Full text of DOI:10.1016/j.bioorg.2021.105024

Bioorganic Chemistry 113 (2021) 105024
Contents lists available at ScienceDirect
Bioorganic Chemistry
journal homepage: www.elsevier.com/locate/bioorg
4-Alkyl-1,2,4-triazole-3-thione analogues as metallo-β-lactamase inhibitors
,
Laurent Gavaraa *, Alice Legrua, Federica Verdirosab, Laurent Sevaillea, Lionel Nautonc,
Giuseppina Corsicab, Paola Sandra Mercurid, Filomena Sanniob, Georges Fellere,
a
b
b
b
b
´
Remi Coulon , Filomena De Luca , Giulia Cerboni , Silvia Tanfoni , Giulia Chelini ,
,
f
,
,*
Moreno Gallenid, Jean-Denis Docquierb *, Jean-François Hernandeza
a
´
´
´
Institut des Biomolecules Max Mousseron, UMR5247 CNRS, Universite de Montpellier, ENSCM, Faculte de Pharmacie, 34093 Montpellier Cedex 5, France
b
c
d
e
f
`
Dipartimento di Biotecnologie Mediche, Universita di Siena, I-53100 Siena, Italy
´
Universite Clermont-Auvergne, CNRS, SIGMA Clermont, Institut de Chimie de Clermont-Ferrand, 63000 Clermont-Ferrand, France
´
´
´
´
`
`
Laboratoire des Macromolecules Biologiques, Centre d’Ingenierie des Proteines-InBioS, Universite de Liege, Institute of Chemistry B6a, Sart-Tilman, 4000 Liege, Belgium
´
´
´
`
´
`
Laboratoire de Biochimie, Centre d’Ingenierie des Proteines-InBioS, Universite de Liege, Allee du 6 août B6, Sart-Tilman, 4000 Liege, Belgium
´
´
´
`
´
`
Centre d’Ingenierie des Proteines-InBioS, Universite de Liege, Allee du 6 août B6, Sart-Tilman, 4000 Liege, Belgium
A R T I C L E I N F O
A B S T R A C T
Keywords:
In Gram-negative bacteria, the major mechanism of resistance to β-lactam antibiotics is the production of one or
several β-lactamases (BLs), including the highly worrying carbapenemases. Whereas inhibitors of these enzymes
were recently marketed, they only target serine-carbapenemases (e.g. KPC-type), and no clinically useful in-
hibitor is available yet to neutralize the class of metallo-β-lactamases (MBLs). We are developing compounds
based on the 1,2,4-triazole-3-thione scaffold, which binds to the di-zinc catalytic site of MBLs in an original
fashion, and we previously reported its promising potential to yield broad-spectrum inhibitors. However, up to
now only moderate antibiotic potentiation could be observed in microbiological assays and further exploration
was needed to improve outer membrane penetration. Here, we synthesized and characterized a series of com-
pounds possessing a diversely functionalized alkyl chain at the 4-position of the heterocycle. We found that the
presence of a carboxylic group at the extremity of an alkyl chain yielded potent inhibitors of VIM-type enzymes
Metallo-β-Lactamase
1,2,4-triazole-3-thione
Bacterial resistance
β-lactam antibiotic
with Ki values in the μM to sub-μM range, and that this alkyl chain had to be longer or equal to a propyl chain.
This result confirmed the importance of a carboxylic function on the 4-substituent of 1,2,4-triazole-3-thione
heterocycle. As observed in previous series, active compounds also preferentially contained phenyl, 2-hy-
droxy-5-methoxyphenyl, naphth-2-yl or m-biphenyl at position 5. However, none efficiently inhibited NDM-1
or IMP-1. Microbiological study on VIM-2-producing E. coli strains and on VIM-1/VIM-4-producing multidrug-
resistant K. pneumoniae clinical isolates gave promising results, suggesting that the 1,2,4-triazole-3-thione
scaffold worth continuing exploration to further improve penetration. Finally, docking experiments were per-
formed to study the binding mode of alkanoic analogues in the active site of VIM-2.
1. Introduction
resistance reached alarming levels, academics, pharmaceutical in-
dustries and governments are urgently implementing a global response
The development and spread of bacteria resistant to antibiotics have
become a very serious threat to public health [1]. Because bacterial
to resolve this crisis. In this context, WHO has recently established a
“global priority list of antibiotic-resistant bacteria” to fight [2]. The first
Abbreviations: BL, β-lactamase; Boc, tert-butyloxycarbonyl; CFU, colony forming unit; DCM, dichloromethane; DIEA, diethylisopropylamine; DMF, dime-
thylformamide; DMSO, dimethylsulfoxide; DPT, dipyridylthionocarbonate; EDTA, Ethylene diamine tetraacetic acid; EtOAc, ethyl acetate; EUCAST, European
Committee on Antimicrobial Susceptibility Testing; HEPES, 4-(2-Hydroxyethyl)-1-piperazine-ethanesulfonic acid; HPLC, high performance liquid chromatography;
IMP, imipenemase; ITC, isothermal calorimetry; KPC, Klebsiella pneumoniae carbapenemase; LC-MS, liquid chromatography coupled to mass spectrometry; MBL,
metallo-β-lactamase; MEM, Meropenem; MHB, Mueller-Hinton broth; MIC, Minimum Inhibitory Concentration; NDM, New Delhi MBL; OXA, oxacillinase; T3P,
propylphosphonic anhydride; TFA, trifluoroacetic acid; VIM, Verona Integron encoded MBL.
* Corresponding authors.
E-mail addresses: laurent.gavara@umontpellier.fr (L. Gavara), jddocquier@unisi.it (J.-D. Docquier), jean-francois.hernandez@umontpellier.fr (J.-F. Hernandez).
https://doi.org/10.1016/j.bioorg.2021.105024
Received 4 March 2021; Received in revised form 19 May 2021; Accepted 22 May 2021
Available online 26 May 2021
0045-2068/© 2021 Elsevier Inc. All rights reserved.

L. Gavara et al.
Bioorganic Chemistry 113 (2021) 105024
and critical priority concerns multi- to extremely-drug resistant Gram-
negative opportunistic pathogens (i.e. Acinetobacter baumannii, Pseudo-
monas aeruginosa and Enterobacteriaceae like Klebsiella pneumoniae and
Escherichia coli), which are resistant to carbapenems, considered last
resort antibiotics in the hospital, and to late-generation cephalosporins.
These drugs belong to the family of β-lactam antibiotics, the most widely
used group of antibacterial agents.
In this context, we are developing MBL inhibitors using the 1,2,4-tri-
azole-3-thione scaffold as an original and promising ligand for the di-
zinc active sites of these enzymes [34,35]. Otto Dideberg’s group pre-
viously identified the binding mode of the 4-amino-5-subtituted
analogue IIIA (Fig. 1) in the di-zinc subclass B3 MBL L1 [36]. The
compound simultaneously bound to the two Zn ions through two atoms
of its heterocyclic moiety (N2 for Zn1 and S3 for Zn2). A similar binding
was more recently established for the di-zinc subclass B1 MBLs VIM-2
[37,38] and NDM-1 [39]. It is also noticeable that several other
studies, including random in silico and experimental screenings, iden-
tified this heterocycle as a potential scaffold for the inhibition of IMP-1
[40], VIM-2 [37] and NDM-1 [41]. Compared to other zinc-binding
groups, the specific behaviour of the 1,2,4-triazole-3-thione moiety
should favour selectivity toward human metallo-enzymes, which are
mainly mono-zinc.
The most relevant resistance mechanism to β-lactam antibiotics is the
production of enzymes called β-lactamases, which inactivate them via
hydrolysis of their β-lactam ring [3–6]. β-Lactamases are divided into
classes A, B, C, and D [7], according to their primary and tertiary
structures, and their catalytic mechanism. Classes A, C and D are serine
hydrolases, which show a heterogeneous substrate profile, mainly con-
cerning penicillins and cephalosporins. But few of them (e.g. the class A
KPC- and the class D OXA-48-type enzymes) also inactivate carbape-
nems. By contrast, class B enzymes, called metallo-β-lactamases (MBLs)
because they use one or two Zn atom(s) for catalysis, are all efficient
carbapenemases, often characterized by a broad substrate profile,
including penicillins and cephalosporins. MBLs are subdivided into
subclasses B1, B2, and B3, based on structural features [8]. Among these,
the most clinically relevant enzymes are di-zinc MBLs belonging to
subclass B1 (e.g. VIM-, NDM-, and IMP-types). MBLs are produced by
those Gram-negative bacteria belonging to the WHO first priority list,
which are responsible of life-threatening nosocomial infections. In
addition, although initially spreading in the hospital setting, MBL-
producing isolates are now increasingly causing community-acquired
infections [9]. Due to the global dissemination of MBL-producing iso-
lates, these enzymes are major therapeutic targets [10].
After Dideberg’s study [36], we launched an extensive exploration of
the 1,2,4-triazole-3-thione scaffold. We first reported a series of IIIA-
based analogues (Fig. 1), diversely substituted at position 5 and
substituted or not at position 4 by a NH2 group [42]. Activities were
generally modest, but few compounds displayed micromolar and broad-
spectrum inhibition against a panel of representative MBLs, including
VIM-2, NDM-1 and IMP-1. However, no compound showed activity in
microbiological assays. This structure–activity relationship study yiel-
ded valuable information about the most favourable substituents at
position 5. These results were confirmed with a series of Schiff base
analogues diversely substituted at positions 4 and 5 (Fig. 1) [38]. In
addition, the presence of an aryl substituent at position 4 linked to the
heterocycle by a hydrazone-like bond led to the discovery of more
potent compounds with a broad spectrum of inhibition. In particular,
several favourable aryl groups including an o-benzoic group were
identified. Another series of molecules possessing a hydrazine-like bond
(i.e. corresponding to the reduced form of Schiff base analogues, Fig. 1)
yielded similar inhibitory activities although of lower impact [43].
However, again, only limited antibiotic potentiation was obtained in
microbiological assays. These results were mainly explained by a poor
bacteria outer membrane penetration. To try to solve this issue, we
continued to explore other types of substituents at position 4 of the
triazole ring and undertook the synthesis of analogues where the
hydrazone- or hydrazine-like function was replaced by alkyl linkers.
We herein report on a new series of 1,2,4-triazole-3-thione com-
pounds where the position 4 was substituted by diversely functionalized
alkyl chains (Fig. 1). The main results were the restriction of the inhi-
bition spectrum to VIM-type enzymes compared to previous series of
analogues but with compounds showing promising activity in microbi-
ological assays, including against MBL-producing clinical isolates.
Inhibiting β-lactamases to protect β-lactam drugs is a well-known
approach and several β-lactamase inhibitors are currently marketed (e.
g. clavulanate, tazobactam, avibactam, vaborbactam) [11]. However,
they only target some serine-β-lactamases (SBLs) [12]. Two MBL in-
hibitors are under clinical development (phase III trials for tani-
borbactam, phase I for QPX7728) [13–16], and only few other
compounds are promising [17–19]. In fact, one major difficulty is the
existence of numerous MBLs showing subtle differences in their active
sites and the discovery of a broad-spectrum inhibitor is a true challenge
[12,20,21]. In addition, most inhibitors typically possess a zinc-
coordinating group, therefore presenting a potential risk to inhibit
important human metallo-enzymes [22].
Among numerous reported families of MBL inhibitors, the largest one
possesses a thiol group [e.g. [23–25]], which was also largely used to
inhibit human metallo-peptidases (e.g. captopril). The consequence is a
high risk for insufficient selectivity. Another important family contains
compoundswithatleasttwocarboxylategroups,anotherwell-knownZn-
ligand, such as succinate [26] or dipicolinate [27] analogues, or as
aspergillomarasmine [28], whose inhibition mechanism at least partially
involves zinc removal [29]. Insufficient selectivity is also a potential
limitation for these compounds [30]. Another Zn-chelating MBL inhibi-
tor, Zn148, was also recently reported [31]. Finally, bicyclic boronates,
such as taniborbactam and QPX7728, are currently the most promising
candidates [14,15,32,33], being inhibitors of both MBLs and SBLs.
2. Results and discussion
2.1. Synthesis
All compounds were prepared according to [44]. The hydrazide
precursor of substituent at position 5 was first obtained in two steps from
Fig. 1. Structure of compound IIIA [36,42], JMV4690, a representative Schiff base analogue [38], the reduced analogues [43] and synthesized 4-alkyl analogues (R1
= (hereto)aryl-(CH2)n- with n = 0–2; R2 = alkyl functionalized or not).
2

L. Gavara et al.
Bioorganic Chemistry 113 (2021) 105024
the corresponding carboxylic acid. Then, the alkylamines R2-NH2, either
commercially available or prepared as described in Supporting Infor-
mation, were treated with DPT (dipyridylthionocarbonate) to yield the
intermediate isothiocyanates, which were directly reacted with hydra-
zides to form the thiosemicarbazide derivatives. Their basic treatment
yielded the expected 1,2,4-triazole-3-thione compounds. Specific syn-
thetic steps carried out for the preparation of several compounds are
presented in Scheme 2. The hydroxamate 4 was prepared by coupling 3
to O-trityl-hydroxylamine, followed by trityl removal in acidic condi-
tions. Compound 8 was obtained from the Boc-protected precursor and
was reacted with: (i) benzoyl chloride and p-tosyl-chloride to yield
compounds 9 and 12, respectively; (ii) DPT as described in Scheme 1,
followed by condensation with benzhydrazide and hot basic treatment
to give the di-triazole-thione compound 15; (iii) phthalic anhydride in
refluxing pyridine to yield the phthalimide 10, which was opened in
basic conditions to afford 11. Compound 38 was obtained from 39
through Boc removal (the same was performed for 36 from 37) and was
condensed to phenylisocyanate to yield the urea 40.
Overall, five compounds displayed Ki values in the micromolar to
submicromolar range against both VIM-2 and VIM-4. Compounds 1 and 2
had a non-functionalized linear butyl and hexyl chain, respectively, while
compounds 3, 5 and 16 contained a butanoic, a butanol or a pyrrolidinone
moiety, respectively. Few other compounds showed more modest activ-
ities (Ki values in the 10–40 μM range) against VIM-2 and/or VIM-4 (i.e. 4,
6, 10, 11, 13, 14, 15, 17–19) with often a preference for VIM-4. Finally, in
this first series, four compounds, 3, 5, 6 and 19, were tested against VIM-1
andshowedmodesttogoodinhibitorypotencies(Ki’s, 39.8± 6.7μM, 1.69
± 0.12 μM, 6.15 ± 0.33 μM, 4.25 ± 0.32 μM, respectively).
According to these first results and as the presence of a carboxylic
group on the substituent at position 4 has already been highlighted in
previous series [38,43], we kept it in a second group of 20 compounds
bearing various alkanoic moieties and a phenyl ring at the 5-position
(Table 2, compound 3 was added for comparison). We varied the
length of the alkanoic chain (20, 22, 23, 26, 27), introduced an amide
bond (21), heteroatoms (O for 24, S for 25), aromatic rings (28–31),
cyclohexyl moieties (32–34), or an amine group, protected (36, 38, 39)
or not (35, 37). As in the case of the first series, all compounds were
inactive or only moderately active against NDM-1 and IMP-1: the best
compounds against NDM-1 displayed longer alkanoic chains (i.e. hex-
2.2. Evaluation of inhibitory potency toward purified MBLs
Compounds were tested against five representative MBLs, the sub-
class B1 enzymes VIM-2, VIM-4, NDM-1 and IMP-1, and the subclass B3
L1. Few compounds were also tested against VIM-1. Initially, testing was
anoic for 26 andoctanoic for 27) and showed Ki values in the 15–20 μM
range. In this series, VIM-2 and VIM-4 inhibition could be retained or
slightly improved. For these enzymes, the alkyl length was important as
compounds with an alkanoic chain shorter than butanoic (acetic 20,
propionic 22) were poorly or not active, whereas those with longer ones
(pentanoic 23, hexanoic 26 and octanoic 27) showed similar or higher
(i.e. submicromolar range) inhibitory potencies compared to compound
performed at one concentration (100 or 200 μM) and Ki values were
measured for compounds showing significant inhibition (typically >
75%). As all compounds evaluated against L1 were no or low inhibitors
(<50% inhibition at 100 μM) of this enzyme, we chose to not include
these results in the Tables.
3. Compound 27 displayed the best Ki values (0.33 and 0.49 μM,
The results obtained for a first series of 19 compounds are presented
in Table 1. In this series, all compounds possessed a phenyl ring at po-
sition 5 and differed by the alkyl substituent at position 4, itself possibly
carrying additional substituents. Two compounds (1 and 2) contained a
non-functionalized saturated alkyl chain, while others displayed diverse
functional groups such as carboxyl (3), hydroxamate (4), alcohol (5, 6),
sulfonic (7), amine (8, 14), amide (9–11, 16), sulphonamide (12), urea
(13), azide (17), sulfonyl (18) and unsaturated alkyl chain (19). In
particular, compounds 9–14 were derivatives of the amine 8. Finally,
compound 15 can be considered as a dimeric 5-phenyl-1,2,4-triazole-3-
thione where the two monomers were linked by an ethylene group at
their 4-position. Overall, most compounds showed poor or no inhibition
towards any MBL tested. In particular, NDM-1 and IMP-1 were not
significantly inhibited by any compound with the exception of com-
respectively) against VIM-2 and VIM-4 in this series. In addition,
although compound 23 was moderately active against VIM-1 as com-
pound 3, compounds 26 and 27 with a longer alkanoic chain were
micromolar inhibitors of this enzyme. Introducing a sulfur but not an
oxygen in the alkyl chain of compound 23 was favourable for both VIM-
type enzymes. Indeed, whereas the ether analogue 24 was moderately or
not active against VIM-2 and VIM-4, respectively, the thioether 25 dis-
played micromolar to submicromolar Ki values against the same en-
zymes. In addition, this compound was also a micromolar inhibitor of
VIM-1 and moderately inhibited NDM-1, indicating that the sulfur
atom significantly improved MBL inhibition.
Branching a phenyl ring at the β position of the butanoic chain of 3
yielded compounds 28 and 29, which were slightly better inhibitors of
VIM-1, VIM-2 and VIM-4 compared to 3. Constraining the alkanoic
chain with a cyclohexyl moiety as in 32–34 was well tolerated for VIM-2
and VIM-4 inhibition. However, compared to the closest analogues 3
and 26, no improvement was observed, with the exception of 34 which
was a better VIM-1 inhibitor.
pound 1, which showed a Ki of around 30 μM against IMP-1. To note, a
close analogue of the NDM-1-inactive compound 19 differing by the
presence of a 2-hydroxyphenyl ring at position 5 has been previously
reported in the literature as a potent NDM-1 inhibitor [41]. In our hands,
however, this compound was also found inactive against NDM-1 (<20%
As a carboxylate group could be unfavourable for external mem-
brane penetration, the adjunction of a protonated or protonatable ni-
trogen atom in a molecule is a frequent mean to counteract the effect of a
negative charge, leading to a zwitterionic form expected to more
favourably penetrate through porins [45]. Unfortunately, replacing the
phenyl ring of compound 28 by a pyridine (i.e. 30 and 31) was detri-
inhibition at 100 μM), supporting the fact that the allyl group was not
more favourable to NDM-1 inhibition than other substituents placed at
position 4 in this series.
mental. Also, introducing an amine group at the
α position of the
butanoic chain (i.e. 35) abolished VIM-2 inhibition, and the addition of a
Boc group was not favourable either (i.e. 36). The same was true for the
α-amino analogue of compound 23 (i.e. 37). However, in contrast, its
Boc-protected derivative 38 showed similar activities to 23 against VIM-
2 and VIM-4 with Ki values in the low micromolar range, suggesting that
the presence of a hydrophobic and bulky group was suitable at this
position. This was confirmed when attaching a phenyl ring to the amine
via a urea link. Although less potent, the corresponding compound 39
significantly inhibited VIM-type enzymes.
Scheme 1. Synthesis of 4-alkyl-1,2,4-triazole-3-thione derivatives. Reagents
and conditions: (a) EtOH, H2SO4, reflux, 5 h; (b) EtOH, hydrazine hydrate, 100 ◦C,
sealed tube; (c) DPT, DMF, sealed tube, 55 ◦C, 3 h; (d) R1-CONHNH2, DMF, 55 ◦C,
3 h; (e) aqueous KOH or NaHCO3, 100 ◦C, 3 h.
Overall, the presence of a hydrophobic linear or constrained alkyl
chain at position 4 was favourable to inhibition of VIM-type enzymes as
compounds 3, 23, 25, 26, 27 and 32–34 were potent inhibitors of these
3

L. Gavara et al.
Bioorganic Chemistry 113 (2021) 105024
Scheme 2. Specific steps for the synthesis of compounds 3, 4, 8–12, 15, 38–40. Reagents and conditions: (a) O-trityl-hydroxylamine, T3P, DIEA, DMF; (b) TFA, DCM,
triisopropylsilane; (c) 15% conc. HCl in dioxane, anisole; (d) Cl-R, DIEA, DMF; (e) DPT, DMF, sealed tube, 55 ◦C, 3 h; (f) benzhydrazide, DMF, 55 ◦C, 3 h; (g) aqueous KOH,
EtOH, 100 ◦C, 3 h; (h) phthalic anhydride, pyridine, reflux; (i) KOH (10 equiv.), EtOH, water, rt; (j) 15% conc. HCl in dioxane, anisole; (k) phenylisocyanate, DIEA,
EtOH, dioxane.
MBLs with Ki values in the micromolar range, and even submicromolar
in the case of compounds 25–27. Branching a phenyl ring (28, 29) or a
blocked amine group (38, 39) was also compatible with inhibition of the
same enzymes.
variable substituent at position 5 (Table 3). Parent compounds 3 (4-
butanoic) and 23 (4-pentanoic) were included for comparison.
Most substituents here introduced at position 5 were explored in
preceding series [38,42,43]. It included o-toluyl (40, 41), halogenated
phenyls (42–44), methoxy-phenyls (45–47), heteroaryls (48–54,
57–59, 62, 66), biaryls (55–61, 63, 65, 66), phenethyl (64) and
cyclohexyl (67). Some of these were previously shown favourable to
broad-spectrum inhibition, as halogenated aryls, 2-hydroxy-5-methoxy-
phenyl, naphthyls, m/p-biphenyls and p-benzyloxyphenyl.
To try to get more insight in the structure–activity relationships in
this series, we selected some compounds and calculated the theoretical
pKa values of their carboxylic group using the pKa module of Marvin-
Sketch software (see Table S1). We also calculated the ligand efficiencies
(LE, Table S1) toward VIM-2 as a representative VIM-type enzyme.
However, no clear correlation could be drawn between pKa values and
the inhibitory potencies, even for compounds having a carboxylic group
at the same distance from the triazole ring (i.e. . 21, 23–25, 37–39). As
stated above, other features like the alkyl chain length or the presence of
hydrophobic and/or bulky moieties are determining.
Unfortunately, as compounds 3 and 23, all compounds were inactive
or only moderately active (Ki of around 100
μM for compound 60
against NDM-1) against NDM-1 and IMP-1. Several compounds (40–47,
50–53, 55–58, 60, 61) significantly inhibited VIM-2 and/or VIM-4 with
Ki values in the micromolar to submicromolar range. All 5-substituents
of these compounds were aryl or heteroaryl groups directly attached to
the triazole ring. Finally, when the comparison was possible (3/23, 40/
Based on these results, we prepared a third series of 28 compounds
possessing either a butanoic or a pentanoic chain at position 4 and a
4

L. Gavara et al.
Bioorganic Chemistry 113 (2021) 105024
Table 1
Table 1 (continued)
Inhibitory activity of 4-alkyl-1,2,4-triazole-3-thiones 1–19 with R1 = phenyl
Cpd
Structure
R2
Ki (μM) or (% inhibition at 100 μM)
VIM-2
VIM-4
NDM-
IMP-1
NI
1
against various MBLs.
18
19
13.4 ±
5.70 ±
NI
2.2
0.58
17.8 ±
8.8 ± 0.8
NI
(45%)
Cpd
Structure
R2
Ki (μM) or (% inhibition at 100 μM)
4.4
VIM-2
VIM-4
NDM-
IMP-1
NI: No Inhibition (<30% or < 50% inhibition at 100 or 200 µM, respectively).
1
aNot determined. All kinetic assays were performed in triplicate.
1
2
2.43 ±
1.20 ±
NI
31.8 ±
0.21
0.13
3.9
41, 46/47, 55/56), no significant difference could be observed between
butanoic and pentanoic compounds. Overall, compound 60 displayed
the best activities against VIM-1, VIM-2 and VIM-4 in this series with Ki
4.94 ±
0.31 ±
NI
NI
0.19
0.03
3
4
7.23 ±
1.47 ±
NI
NI
NI
NI
values of 5.50, 0.35 and 0.27 μM, respectively, confirming the interest of
0.68
0.20
a m-biphenyl at position 5 [38,42,43].
16.1 ±
10.6 ±
2.4
1.6
2.3. Isothermal calorimetry experiments (ITC)
5
6
6.17 ±
3.86 ±
NI
NI
NI
NI
To validate the binding mode of compounds to VIM-2, isothermal
titration calorimetry was performed with compound 3. The thermody-
namic parameters of binding are given in Table 4 and are also depicted
in Fig. 2A for a more visual inspection. An ITC thermogram is shown in
Fig. 2B.
0.55
0.19
40.2 ±
11.4 ±
13.8
1.7
a
7
8
NI
-
NI
NI
NI
NI
The experiment showed a stoichiometric relationship between VIM
and 2 and the inhibitor, which clearly indicated a 1:1 association.
Compound 3 binding was enthalpy-driven (the enthalpic term is the
NI
NI
NI
major contributor to ΔG◦ and therefore to affinity). The favorable
b
9
NI
NI
NI
enthalpy contribution is an indication of specific interactions between
binding partners and reflects ligand specificity and selectivity. Com-
pound 3 also displayed a favorable entropic contribution, which is
regarded as resulting from the release of organized water molecules at
the surface of individual partners to the bulk solvents. Overall, this
suggests that favorable enthalpic interactions, mainly electrostatic,
occurring in the complex were replacing pre-existing bonds established
with the solvent water [46,47]. Accordingly, the net enthalpic effect is
reduced due to desolvation and, as a result, the affinity remained
moderate. The Kd value was in the same range as the Ki value.
10
11.4 ±
NI
NI
NI
1.2
11
18.4 ±
12.3 ±
NI
NI
3.6
1.1
2.4. In vitro antibacterial synergistic activity and cytotoxicity
12
13
NI
NI
(35%)
NI
NI
NI
The ability of selected compounds to potentiate the activity of
β-lactam antibiotics was first evaluated using a disk diffusion assay
performed on isogenic MBL-producing laboratory Escherichia coli
strains, obtained by transforming strain AS19 (a chemically mutagen-
ized strain with enhanced permeability) or LZ2310 (a triple knock-out
mutant of K12 in which genes encoding components of the major
efflux pumps NorE, MdfA and AcrA were inactivated) with a derivative
of the pLB-II plasmid in which the blaVIM-2 gene was cloned [48]. The
latter strain was used to limit the potential efflux of the compounds,
which would prevent sufficient accumulation in the periplasm. Com-
pounds were tested according to their inhibition profile (Table 5).
In this assay, the two potent VIM-2 inhibitors 3 and 23 significantly
potentiated the activity of cefoxitin on the VIM-2-producing LZ2310
strain (Table 5, Figure S1). Indeed, they restored susceptibility to the
antibiotic (being above the EUCAST resistance breakpoint, 18 mm), as a
diameter of the growth inhibition zone of 19–20 mm was measured.
However, a more limited effect was observed when these compounds
were tested on the AS19 strain, potentially indicating the role of efflux
pumps in preventing a significant accumulation of the compounds in the
periplasm. Nonetheless, compound 23 also provided some detectable
potentiation of cefoxitin on VIM-2-producing AS19 strain. Both com-
pounds 3 and 23 possess a phenyl ring at position 5 and only differ by
one CH2 in the alkyl chain at position 4, which could explain their
17.9 ±
10.4 ±
6.1
1.2
14
15
13.7 ±
7.8 ± 1.3
4.1 ± 0.6
NI
NI
NI
NI
2.9
11.7 ±
1.4
16
17
5.32 ±
2.60 ±
NI
NI
NI
0.76
0.18
27.7 ±
NI
(31%)
5.4
5

L. Gavara et al.
Bioorganic Chemistry 113 (2021) 105024
Table 2
Inhibitory activity of 4-alkyl-1,2,4-triazole-3-thiones 3, 20–39 with R1 = phenyl and R2 containing a carboxylic function against various MBLs.
Cpd
Structure
R2
Ki (
μ
M) or (% inhibition at a100 or b200
μM)
VIM-1
VIM-2
VIM-4
NDM-1
NI
IMP-1
NI
3
39.8 ± 6.7
7.23 ± 0.68
1.47 ± 0.20
c
20
21
-
44.3 ± 0.2
(40%)a
NI
NI
NI
NI
NI
9.54 ± 1.21
6.79 ± 0.60
–
22
23
NI
(30%)a
NI
NI
NI
13.9 ± 1.1
2.81 ± 0.58
1.13 ± 0.12
(60%)b
–
24
25
14.3 ± 2.0
NI
NI
NI
NI
2.36 ± 0.14
0.81 ± 0.04
0.71 ± 0.06
(40%)a
26
27
28
5.05 ± 0.26
1.27 ± 0.08
18.5 ± 1.3
0.72 ± 0.10
0.33 ± 0.02
0.80 ± 0.04
1.36 ± 0.10
0.49 ± 0.02
1.04 ± 0.26
19.7 ± 0.2
15.7 ± 1.8
NI
NI
(40%)a
NI
29
24.3 ± 2.1
1.32 ± 0.17
0.42 ± 0.02
NI
NI
–
30
31
NI
NI
NI
NI
NI
NI
NI
NI
NI
–
32
8.06 ± 0.94
6.54 ± 0.25
NI
NI
–
33
34
6.54 ± 0.64
2.87 ± 0.22
3.08 ± 0.23
0.85 ± 0.08
NI
NI
NI
NI
2.82 ± 0.34
–
–
–
–
35
36
NI
NI
NI
NI
NI
NI
(continued on next page)
6

L. Gavara et al.
Bioorganic Chemistry 113 (2021) 105024
Table 2 (continued)
Cpd
Structure
R2
Ki (
μ
M) or (% inhibition at a100 or b200
μM)
VIM-1
VIM-2
VIM-4
NDM-1
IMP-1
–
–
37
38
NI
NI
NI
NI
NI
NI
1.95 ± 0.11
1.51 ± 0.08
39
(57%)a
6.93 ± 1.08
4.14 ± 0.39
(30%)a
NI
NI: No Inhibition (<30% or < 50% inhibition at 100 or 200 µM, respectively). a% inhibition at 100
μ
M; b% inhibition at 200
μ
M; cNot determined. All kinetic assays
were performed in triplicate.
similar in vitro activity.
Unfortunately, this combination did not further decrease the MIC of
meropenem toward the VIM-1-producing clinical isolate (not shown).
Other compounds, despite being similarly potent VIM-2 inhibitors (i.
e. compounds 16, 32, 40, 46 and 55), did not show any antibiotic
potentiation, indicating that they might not be able to efficiently cross
the outer membrane and accumulate in the periplasm at biologically
relevant concentrations. In particular, compounds 40, 46 and 55 are
analogues of compound 3 differing by their substituent at position 5
(respectively, o-toluyl, 2-hydroxy-5-methoxy-phenyl and 2-naphthyl). A
similar influence of this substituent was already observed in the Schiff
base series (Fig. 1), as only the analogue with an unsubstituted phenyl
ring at position 5 was found active in the same assay [38].
Nonetheless, the addition of a subinhibitory concentration (0.12 μg/ml)
of colistin to the medium did dramatically improve the potentiation
activity of 47 (by itself very limited), among the best VIM-4 inhibitors
(meropenem MIC, 16 and 2 μg/ml in the absence and presence of the
inhibitor, respectively; VIM-4-producing K. pneumoniae VA-416/02).
Furthermore, a similar result was obtained with compound NAB741, a
polymyxin derivative without direct antibacterial activity but retaining
the outer membrane permeabilizing effect [50,51]. When tested in the
presence of 4 μg/ml NAB741, meropenem MIC was decreased to 4 μg/ml
We also tested the ability of some compounds to increase the mer-
openem susceptibility of MBL-producing multidrug-resistant clinical
isolates using a broth microdilution method. As the most potent in-
hibitors of the present series are mainly targeting VIM-type enzymes, we
chose two K. pneumoniae isolates producing VIM-1 and VIM-4. Com-
in the presence of the inhibitor. These data overall indicate that the in
vitro synergistic activity of potent inhibitors are strongly dependent on
their ability to rapidly cross the outer membrane barrier and rely on
subtle chemical changes in the inhibitor structure.
Furthermore, a preliminary assessment of compound cytotoxicity
was carried out using a membrane integrity assay on HeLa cells. Strik-
ingly, no cytotoxic effect could be observed at the highest tested con-
pounds were tested at a fixed concentration of 32 μg/ml and a low
micromolar or submicromolar VIM inhibitory activity was necessary to
observe a significant reduction of meropenem MIC on K. pneumoniae
bacteria (Table 6). In the case of VIM-1-producing bacteria, although not
all inhibitory potencies have been determined, several among the most
potent VIM-1 inhibitors (i.e. 5, 6, 19, 25, 27 and 60) decreased the
antibiotic MIC by eight-fold. Less potent inhibitors (i.e. 3, 23 and 29)
also showed lower potentiating activity. When tested with a VIM-4-
producing strain, several compounds showed a 4- to 16-fold potentia-
tion of meropenem. In particular, compound 60, which was the best
VIM-4 inhibitor in this series, also displayed the highest potentiation
effect (16-fold) with this isolate. To note, several compounds with
similar VIM-4 inhibitory potencies in the low micromolar to sub-
micromolar range (i.e. 1, 3, 23, 25–27, 29, 34, 38, 47, 51, 56, 60)
showed different levels of MIC reduction (from zero-fold for 47 and two-
fold for 23 to eight-fold for 3, 26 and 34). These results indicate that
subtle structural changes could have significant effects on outer mem-
brane penetration and rate of accumulation in the periplasm. None of
the tested compounds showed intrinsic antibacterial activity when
centration (IC50 values > 500 μM) with compounds 26, 27 and 60,
selected on the basis of their significant synergistic activity with mer-
openem on MBL-producing clinical isolates.
Overall, we demonstrated that 1,2,4-triazole-3-thione-based com-
pounds could restore the susceptibility of MBL-producing clinical iso-
lates to a carbapenem, while they do not induce cell lysis of mammalian
cells in a cytotoxicity assay, supporting the interest for further medicinal
chemistry efforts to improve their membrane permeation properties.
2.5. Modelling study
We investigated the putative binding mode of several compounds in
the VIM-2 active site with docking experiments. Three compounds
containing a butanoic chain at position 4 and a variable substituent at
position 5 (3, 5-phenyl; 46, 5-(2-hydroxy-5-methoxy-phenyl); 60, 5-(m-
biphenyl)) and a 3 analogue with a hexanoic chain (26) were studied.
The docking experiments were performed with a VIM-2 model
generated from 6YRP available in the Protein data Bank [38] using
AutoDock 4.2. 6YRP is the crystallographic structure of the complex
tested alone (MIC > 128 μg/mL).
These compounds were also evaluated with an NDM-1-producing
E. coli clinical isolate but none showed any detectable meropenem
potentiation (not shown), as expected in consideration of their very low
inhibitory activity on this enzyme.
between VIM and 2 and the Schiff base analogue JMV4690 (Ki = 0.7 μM,
Figure 1, R1 = 2-hydroxy-5-methoxy-phenyl), which mainly differs from
the studied compounds by possessing a rigid hydrazone-like bond and a
2-benzoic acid at position 4 of the triazole ring. (Figure S2). This
structure showed the expected double coordination of the active site
zinc ions by two atoms of the triazole-thione moiety, and a H-bond
network involving the two carboxylate oxygens, the main chain NH of
Asn233 and two structural waters (named 5 and 177). In addition, the
Encouraged by these data, we evaluated whether conditions liable to
increase the compound concentration in the bacterial periplasm could
further increase their synergistic activity. In particular, compound 60
was tested in the presence of the dipeptide Phe-Arg-β-naphthylamide
(PAβN, 25 and 100
μg/mL), a known efflux pump inhibitor [49].
7

L. Gavara et al.
Bioorganic Chemistry 113 (2021) 105024
Table 3
Inhibitory activity of 4-alkyl-1,2,4-triazole-3-thiones 3, 23, 40–67 with R2 = butanoic (n = 1) or pentanoic (n = 2).
Cpd
Structure
R1
Ki (
μ
M) or (% inhibition at a100 or b200
μM)
n
VIM-1
VIM-2
VIM-4
NDM-1
IMP-1
3
1
2
39.8 ± 6.7
13.9 ± 1.1
7.23 ± 0.68
2.81 ± 0.58
1.47 ± 0.20
1.13 ± 0.12
NI
NI
NI
23
(60%)b
c
40
41
1
2
-
10.3 ± 1.1
3.13 ± 0.20
3.17 ± 0.22
NI
NI
NI
NI
–
3.07 ± 0.32
42
43
1
1
NI
6.58 ± 0.48
8.80 ± 2.04
2.10 ± 0.18
4.62 ± 0.32
NI
NI
NI
NI
–
44
45
1
1
–
–
6.39 ± 0.72
3.01 ± 0.19
4.11 ± 0.45
1.70 ± 0.11
NI
NI
NI
NI
46
47
1
2
NI
4.20 ± 0.41
1.47 ± 0.14
1.13 ± 0.15
1.31 ± 0.07
NI
NI
NI
NI
–
48
49
50
1
1
1
–
–
–
NI
NI
NI
NI
NI
(65%)b
NI
NI
NI
8.89 ± 0.58
4.84 ± 0.29
NI
51
52
1
1
–
–
6.58 ± 0.90
9.27 ± 1.53
2.08 ± 0.09
4.41 ± 0.39
NI
NI
NI
NI
53
54
1
1
–
–
14.0 ± 2.2
9.09 ± 1.90
NI
NI
NI
NI
NI
NI
55
56
1
2
NI
5.56 ± 0.42
2.99 ± 0.32
0.91 ± 0.08
1.56 ± 0.12
NI
NI
NI
–
(60%)b
57
58
1
1
–
2.98 ± 0.29
4.62 ± 0.32
1.91 ± 0.18
0.31 ± 0.03
(50%)b
NI
NI
NI
30.8 ± 3.1
(continued on next page)
8

L. Gavara et al.
Bioorganic Chemistry 113 (2021) 105024
Table 3 (continued)
Cpd
Structure
Ki (
μ
M) or (% inhibition at a100 or b200
μM)
R1
n
1
VIM-1
NI
VIM-2
VIM-4
NDM-1
NI
IMP-1
NI
59
18.2 ± 2.8
13.3 ± 4.2
60
61
1
1
5.50 ± 0.67
0.35 ± 0.14
3.29 ± 0.18
0.27 ± 0.13
1.26 ± 0.09
99.8 ± 1.6
NI
NI
NI
NI
62
63
1
1
–
NI
NI
NI
NI
NI
NI
NI
NI
NI
64
65
1
1
–
–
NI
NI
NI
NI
NI
NI
NI
NI
66
67
1
1
NI
NI
NI
NI
NI
NI
NI
NI
–
–
NI: No Inhibition (<30% or < 50% inhibition at 100 or 200 µM, respectively).a% inhibition at 100
μ
M; b% inhibition at 200
μ
M; cNot determined. All kinetic assays
were performed in triplicate.
provided an interaction similar to that of JMV4690, although the longer
alkanoic chain compared to 3 (i.e. hexanoic) did not longer allow the
interaction of the carboxylic group within the Asn233 pocket, but
ideally placed it to interact with the Arg228 side-chain (Fig. 3D).
Concerning the variable substituent at position 5, the 2-hydroxy-5-
methoxy-phenyl moiety of 46 was positioned in a reverse way
compared to the same substituent in JMV4690. In this case, the hydroxyl
group would establish a H-bond with the side chain of Asp119 (Fig. 3B).
The flexibility of the Asp119 side-chain during the docking experiment
and the absence of solvent might explain this result as such experiment
promotes electrostatic interaction. In the case of compound 60, the m-
biphenyl group did not seem to establish particular interactions that
could explain its 10-fold higher inhibitory potency compared to the 5-
phenyl analogue. An indirect effect could be an increase rigidity
because of the bulkier substituent. The hydrophobicity of the biphenyl
substituent could displace solvent molecules from the active site, thus
contributing a favourable entropic stabilization energy.
Table 4
Thermodynamic parameters of compound 3 binding to VIM-2 at 25 ◦C.
Cpd
n
Ka (Mꢀ 1
)
Kd
M)
ΔG◦
ΔН◦
TΔS◦
b
(
μ
(kcabl/
mol)
(kcabl/
mol)
(kcal/mol)
3
1.00 ±
(6.4 ±
1.568
ꢀ 7.9
ꢀ 7.0 ±
0.9
0.01
0.1) 105
0.2
methoxy group of its 2-hydroxy-5-methoxy-phenyl moiety at position 5
interacted with the indole NH of Trp87. π stacking interactions were also
observed between the 5- and 4-aromatic rings and Trp87 and His263,
respectively.
First, all studied compounds could be modelled to create similar
interactions with the two zinc ions. The triazole nitrogen and sulphur are
well positioned to complete the coordination tetrahedron of each zinc.
Both compounds possess a carboxylic group at the end of a flexible alkyl
chain and no aryl group at the 4-position. Therefore, such compounds
lost the advantage of interacting with the His263 side-chain. However,
the carboxylic group of compounds with a butanoic substituent (3, 46,
60) was found to interact as the same group of JMV4690 within the
Asn233 pocket (Fig. 3A, B, C). In addition, the Arg228 side-chain
seemed to be close enough to complete this interaction network. This
is in contrast with the carboxylate group of JMV4690, which was shown
to not interact with this basic residue, likely due to the higher rigidity of
its carboxylate-bearing cyclic moiety. The docking of compound 26
The Table S2 presents the mean free energies of best clusters and
estimated Ki values for all four compounds, which were globally
consistent with experimental ones.
3. Conclusion
By introducing a diversely functionalized alkyl chain at position 4 of
5-substituted 1,2,4-triazole-3-thione derivatives, we found that simple
9

L. Gavara et al.
Bioorganic Chemistry 113 (2021) 105024
Fig. 2. (A) VIM-2 binding energetics of compound 3. Data are from Table 4. (B) Isothermal titration calorimetry of VIM-2 by compound 3 at 25 ◦C. Upper panel:
exothermic microcalorimetric trace of compound injections into VIM-2 solution (18.7 µM). Lower panel: Wiseman plot of heat releases versus molar ratio of
injectant/protein in the cell and nonlinear fit of the binding isotherm for n equivalent binding sites. The binding enthalpy corresponds to the amplitude of the
transition curve, Ka is derived from the slope of the transition and the stoichiometry n is determined at the transition midpoint.
Table 5
Table 6
In vitro synergistic activity of selected VIM-2 inhibitors measured by disk
Antibacterial synergistic activity of compounds on VIM-1-and VIM-4-producing
K. pneumoniae clinical isolates with meropenem (MEM) determined by the broth
microdilution method.
diffusion assay.a
Cpdsb
Ki (
μM) on VIM-2
Inhibition zone diameter (mm)c
Cpd
(32
MEM MICa
(μ
g/mL) and Ki (
μM) values of selected inhibitors
AS19
LZ2310
μ
g/mL)
K. pneumoniae 7023
K. pneumoniae VA-416/02
None
EDTA
DMSO
3
–
–
–
16
26
17
ND
17
19
17
17
17
16
14
30
14
20
14
19
15
14
16
14
(bla+VIM-1
)
(bla+VIM-4
)
MEM MIC
Ki (VIM-1)b
MEM MIC
Ki (VIM-4)b
7.23 ± 0.68
5.32 ± 0.76
2.81 ± 0.58
8.06 ± 0.94
10.3 ± 1.1
None
1
16
16
4
–
16
4
–
16
ND
1.20 ± 0.13
1.47 ± 0.14
10.6 ± 1.6
3.86 ± 0.19
11.4 ± 1.7
10.4 ± 1.2
8.75 ± 0.83
1.13 ± 0.12
0.71 ± 0.06
1.36 ± 0.10
0.49 ± 0.02
0.42 ± 0.02
3.08 ± 0.23
0.85 ± 0.08
1.51 ± 0.08
1.31 ± 0.7
2.08 ± 0.09
1.56 ± 0.12
0.27 ± 0.13
23
3
39.8 ± 6.7
ND
2
32
4
4
4
40
5
2
1.69 ± 0.12
6.15 ± 0.33
ND
4
46
4.20 ± 0.41
5.56 ± 0.42
6
2
8
55
13
19
23
25
26
27
29
33
34
38
47
51
56
60
8
4
a
VIM-2-producing E. coli strains were obtained by transforming the pLB-II
plasmid carrying the cloned MBL gene in strains AS19 and LZ2310 (triple
knock-out mutant of efflux pumps).
2
4.25 ± 0.32
13.9 ± 1.1
2.36 ± 0.14
5.05 ± 0.26
1.27 ± 0.08
24.3 ± 2.01
ND
4
8
8
2
4
4
2
2
4
b
A variable volume of the compound dissolved in DMSO was added to a
8
4
cefoxitin disk (30 μg) in order to obtain a final inhibitor quantity of 40 μg. DMSO
4
4
was used as a control and did not affect the diameter of the growth inhibition
4
ND
2
zone. 220 µg EDTA restored full susceptibility to the antibiotic.
8
ND
4
4
ND
16
4
c
4
ND
EUCAST resistance breakpoint of cefoxitin = 18 mm.
8
ND
4
2
5.50 ± 0.67
1
a
MEM, meropenem.
linear alkyl (1, 2), alkanol (5) or alkanoic chains (3, 23, 25–27) were the
most favourable for MBL inhibition. In fact, higher potencies were
generally measured for compounds with a carboxylic group at the ex-
tremity of a linear alkyl chain, the alkyl chain being longer or equal to
b
From Tables 1–3.
10

L. Gavara et al.
Bioorganic Chemistry 113 (2021) 105024
Fig. 3. Binding mode of compounds 3 (A), 46 (B), 60 (C) and 26 (D) in VIM-2 studied by molecular modeling. Compounds (orange) were superimposed with
JMV4690 (light blue) for comparison. VIM-2 side chains that were let flexible are in green. All four compounds interact with the two zinc ions as determined for
JMV4690, while their 5-aryl substituent could present different orientations. The images were produced using UCSF Chimera [52]. (For interpretation of the ref-
erences to colour in this figure legend, the reader is referred to the web version of this article.)
propyl. A branched substituent like a phenyl ring (28, 29) or a protected
amino group (38), and a constrained alkyl chain (32–34) were also
tolerated but without change in potency. VIM-1, VIM-2 and VIM-4 were
the only MBLs significantly inhibited among all tested, with Ki values in
the submicromolar to micromolar range. By contrast, NDM-1 and IMP-1
were not or only modestly inhibited by this series of compounds, high-
lighting a substantial structural heterogeneity between the active sites of
these subclass B1 MBLs. As already observed in previous series of 1,2,4-
triazole-3-thione compounds, the 4-position is a determining factor for
selectivity [38]. Keeping a butanoic or a pentanoic chain at position 4,
the study of substitution at position 5 yielded several aryl moieties that
were reported favourable in previous series [38,43], i.e. 2-hydroxy-4-
methoxy-phenyl (46, 47), naphthyl (55, 56), m-biphenyl (60) and
benzyloxyphenyl (61). Only the m-biphenyl substituent achieved higher
VIM-2 and VIM-4 inhibition with Ki values in the submicromolar range
compared to compound 3, as was observed for the corresponding
hydrazone-like [38] and hydrazine-like [43] compounds with an o-
benzoic group at the 4-position. The modelling study suggested that all
docked compounds could interact with the two zinc ions in the VIM-2
active-site as JMV4690 does in the crystallographic structure 6YRP. In
addition, the carboxylic group of both compounds with a butanoic chain
at position 4, would interact within the Asn233 pocket as the same group
of JMV4690. Increasing the alkanoic chain length to hexanoic would
prevent that interaction but favour ionic interaction with Arg228,
therefore allowing to keep similar binding affinity.
ability to accumulate in the periplasm at significant concentrations and
our data provided some understanding of the structural requirements
modulating this aspect.
Despite the compounds displayed a narrower spectrum of MBL in-
hibition compared to previous series, promising results were obtained in
microbiological and cytotoxicity assays, supporting the interest to
continue developing this series of 1,2,4-triazole-3-thione-based com-
pounds. Further modification at position 4 of the heterocycle is ongoing
to obtain compounds possessing an extended inhibitory spectrum and
being able to potentiate the antibiotic susceptibility of MBL-producing
clinical isolates. This work will be reported soon.
4. Experimental
4.1. Chemistry
Hydrazides R1-CO-NHNH2 and non commercial amines R2-NH2 were
prepared as described in Supporting Information. Physico-chemical
properties of final compounds are also presented in Supporting Infor-
mation. The purity of final compounds was determined to be ≥ 95% by
1H NMR and reverse phase HPLC.
4.1.1. General procedure for the preparation of 4-alkyl-1,2,4-triazole-3-
thiones diversely substituted at positions 4 and 5.
Triazole-thione compounds were prepared following the synthetic
pathway reported by Deprez-Poulain et al. (2007) [44].
Microbiological assays identified two compounds able to potentiate
the activity of a β-lactam antibiotic when tested on two VIM-2-producing
laboratory strains. More importantly, several compound were found to
significantly potentiate the activity of meropenem when tested on two
multidrug-resistant VIM-producing K. pneumoniae clinical isolates.
Compared to previous series [38,43], these inhibitors showed a higher
4.1.1.1. 4.1.1.1. Thiosemicarbazide intermediates. To a solution of the
amine R2-NH2 (1 mmol, 1 equiv.) in anhydrous DMF (4 mL) (in the case
of hydrochloride salt, 1.5 equiv. of Na2CO3 (159 mg) was added) was
added DPT (244 mg, 1.05 mmol, 1.05 equiv.). After stirring at 55 ◦C in a
11

L. Gavara et al.
Bioorganic Chemistry 113 (2021) 105024
sealed tube for 1 h30, the hydrazide R1-CONHNH2 (1.1 mmol, 1.1
equiv.) was added and the mixture was again heated at 55 ◦C for 1 h30.
After cooling to room temperatur, the solution was diluted in EtOAc and
the organic phase was extracted five times with water, dried over MgSO4
and concentrated in vacuum. If necessary, the product was purified by
gel column chromatography (EtOAc/Hexane).
equiv., 0.18 mmol., 19 µL) in a mixture of dioxane (4 mL), AcOEt (4 mL)
and triethylamine (1 mL) was stirred at room temperature for 2 h. The
solution was diluted in water (10 mL) and washed with EtOAc (15 mL).
The aqueous layer was lyophilized and the crude product was purified
by preparative reverse phase HPLC to give the corresponding urea
compound 39.
4.1.1.2. 4.1.1.2. Cyclization. The thiosemicarbazide intermediates were
solubilized in a mixture of water and ethanol (2:3) and KOH (3 equiv., 3
mmol) was added. After refluxing for 2 h, the mixture was neutralized
with saturated aqueous KHSO4 and extracted twice with DCM. The
organic phases were mixed, dried over MgSO4, filtered and evaporated
under vacuum. The residues were purified by gel column chromatog-
raphy (EtOAc/Hexane) or by reverse phase HPLC.
4.2. Biology
4.2.1. Metallo-β-Lactamase inhibition assays
4.2.1.1. VIM-type enzymes, NDM-1 and IMP-1. The inhibition potency
of the compounds has been assessed as previously reported [38,43].
Briefly, the rate of hydrolysis of a reporter substrate, either 150 μM
imipenem or meropenem (λ, 300 nm), 120
100
μM cefotaxime (λ, 260 nm) or
4.1.2. Specific synthetic procedures for compounds 4, 8–12, 15, 35, 37,
39
μ
M nitrocefin (λ, 482 nm), by a purified MBL enzyme (VIM-1, VIM-
2, VIM-4, NDM-1, and IMP-1; enzyme concentration in the assay ranged
1–70 nM) was measured at 30 ◦C in 50 mM HEPES buffer (pH 7.5) in the
absence and presence of several concentrations of the inhibitor (0.5 µM
–1 mM).
4.1.2.1 Compound 4. A mixture of the carboxylic acid 3 (1 equiv., 0.5
mmol, 135 mg), O-trityl-hydroxylamine (1.2 equiv., 0.6 mmol, 165 mg),
T3P (1.1 equiv, 0.55 mmol, 175 mg) and DIEA (3 equiv., 1.5 mmol, 0.26
mL) in DMF (4 mL) was stirred at room temperature for 4 h. The solution
was then diluted in EtOAc (20 mL) and washed with water (4x30 mL).
The organic layer was dried over MgSO4, filtered and evaporated. The
crude product was purified on silica gel column (EtOAc/Hex) to give the
corresponding trityl-protected compound.
The inhibition constants (Ki) were determined on the basis of a model
of competitive inhibition by analysing the dependence of the ratio v0/vi
(v0, hydrolysis velocity in the absence of inhibitor; vi, hydrolysis velocity
in the presence of inhibitor) as a function of [I] as already described
[53]. The assays were performed in triplicate.
A mixture of the trityl-protected intermediate (1 equiv., 0.34 mmol,
177 mg), TFA (10 equiv, 0.25 mL) and triisopropylsilane (11.5 equiv,
0.8 mL) in dichloromethane (8 mL) was stirred at room temperature for
10 min. The residue obtained upon evaporation was purified on silica gel
column (DCM/MeOH) to give the corresponding hydroxamic acid 4.
4.2.1.2. L1. Inhibition assay was performed as previously described
[38,43]. Briefly, L1 (0.2–0.8 nM) was tested in 25 mM HEPES pH 7.5
with 50 μM ZnCl2. The enzyme and the inhibitor (100 µM) were pre-
incubated for 30 min at 21 ◦C and then incubated at 30 ◦C in the pres-
ence of 100
μ
M nitrocefin as the reporter substrate (Δε482 = 15,000 Mꢀ 1
.
4.1.2.1. Compounds 8, 35 and 37. A mixture of Boc-protected com-
pound (1 equiv., 1 mmol, compounds 36 for 35 and 38 for 37) in
dioxane (4 mL) and hydrochloric acid 37% (2 mL) was stirred at 40 ◦C
for 1 h and then evaporated. The resulting products were obtained in
sufficient purity.
cmꢀ 1). In the present study, no compound inhibited L1 with a residual
activity < 30%, and therefore no Ki was measured [24]. The assays were
performed in triplicate.
4.2.2. Microbiological assays
The ability of selected compounds to restore the susceptibility of
resistant bacteria to a β-lactam antibiotic was assessed as previously
described [38,43], according to the CLSI (Clinical Laboratory Standard
Institute) recommendations [54,55].
4.1.2.2. Compounds 9 and 12. A mixture of the 1,2,4-triazole-3-thione
compound 8 (1 equiv., 0.2 mmol, 51 mg) and benzoyl chloride (com-
pound 9) or p-tosyl chloride (compound 12) (1.1 equiv., 0.22 mmol) in
DMF was stirred at room temperature for 1 h. The solution was then
diluted in EtOAc (10 mL) and washed with water (5x20 mL). The
organic layer was dried over MgSO4, filtered and evaporated. The crude
products were purified on silica gel column (EtOAc/Hex) to yield the
corresponding acylated compounds 9 and 12.
Assays using the agar disk-diffusion method were performed on MBL-
producing isogenic laboratory strains obtained after transforming
Escherichia coli AS19 [56] and LZ2310 [57] strains with a derivative of
the high copy number plasmid pLB-II carrying the cloned blaVIM-2 gene.
After 18 h incubation at 35 ◦C in the presence of cefoxitin (30
μg)-
containing disks supplemented with an inhibitor or DMSO or EDTA
(220 g) as negative and positive controls, respectively, the diameter of
4.1.2.3. Compounds 10 and 11. A mixture of the 1,2,4-triazole-3-thione
compound 8 (1 equiv., 1 mmol, 257 mg) and phthalic anhydride (1.5
equiv., 1.5 mmol, 222 mg) in pyridine (2 mL) was refluxed for 1 h. The
solution was then diluted in EtOAc (20 mL) and washed with water
(4x30 mL). The organic layer was dried over MgSO4, filtered and
evaporated to give a crude powder. The resulting phthalimide 10 was
purified by recrystallization from a mixture of EtOAc-Hexane.
μ
the growth inhibition zone was measured and compared to that obtained
in the absence of inhibitor. Experiments were performed in triplicate.
The minimum inhibitory concentration (MICs) of meropenem
(MEM) was determined in triplicate using Mueller-Hinton broth (MHB)
and a bacterial inoculum of 5 × 104 CFU/well, in both the absence and
presence of a fixed concentration (32
μg/mL) of an inhibitor. Two
A solution of the phthalimide 10 (1 equiv., 0.2 mmol, 70 mg) in a
mixture of EtOH-H2O (2–1/6 mL) was added to a solution of KOH (112
mg, 10 equiv., 2 mmol) and the reaction was stirred at room temperature
for 1 h. The mixture was acidified by an 1 N aqueous solution of KHSO4
(5 mL) and extracted by EtOAc (3x10 mL). The combined organic layers
were dried over MgSO4, filtered and evaporated to give a crude powder.
Recrystallisation from EtOAc led to the pure compound 11.
multidrug-resistant K. pneumoniae isolates producing VIM-1 (7023) and
VIM-4 (VA-416/02) and present in our collection were used [58,59].
4.2.3. Cytotoxicity assays
The potential cytotoxic activity of selected compounds (26, 27 and
60) was evaluated using the commercially available membrane integrity
assay (CytoTox 96® non-radioactive cytotoxicity assay, Promega,
Madison, WI, U.S.A.). HeLa cells were grown in Dulbecco’s Modified
Eagle’s Medium (Euroclone, Pero, Italy) supplemented with 10% fetal
4.1.2.4. Compound 15. This compound was obtained from compound 8
and benzohydrazide by following the general synthetic procedure for 1,2,4-
triazole-3-thiones.. 4.1.2.6. Compound 39. A solution of the amino acid
compound 37 (1 equiv., 0.2 mmol, 58 mg) and phenyl isocyanate (0.9
◦
bovine serum, 4.5 mg/mL glucose and 2 mM L-glutamine (37 C, 5%
CO2). The compounds were tested for their ability to induce cell lysis by
measuring the release of lactate dehydrogenase (LDH) after incubating
12

L. Gavara et al.
Bioorganic Chemistry 113 (2021) 105024
HeLa cell cultures (20,000 cells/well) for 24 h (medium as above, 37 ◦C,
5% CO2) in the absence and presence of varying concentrations of the
Appendix A. Supplementary material
compound (up to 500
μ
M). Further controls included samples containing
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.bioorg.2021.105024.
the medium only, 1% DMSO or in which cell lysis was induced by the
addition of 9% Triton X-100 (maximum LDH release control). The assay
was performed in triplicate.
References
[1] C. Lee Ventola, The antibiotic resistance crisis: part1: causes and threats, Pharm.
Ther. 40 (2015) 277–283. PMC4378521.
4.3. Isothermal titration calorimetry (ITC) analysis of compound binding
to VIM-2
[2] World Health Organization, Global priority list of antibiotic-resistant bacteria to
guide research, discovery and development of new antibiotics, 27 february 2017.
[3] P. Nordmann, T. Naas, L. Poirel, Global spread of carbapenemase-producing
Enterobacteriaceae, Emerg. Infect. Dis. 17 (2011) 1791–1798, https://doi.org/
10.3201/eid1710.110655.
ITC titration was performed on a MicroCal ITC200 (GE-Malvern)
equipped with a 200 µL Hastelloy sample cell and an automated 40 µL
glass syringe rotating at 1000 rpm. VIM-2 in 10 mM HEPES-NaOH, 0.15
M NaCl, 50 µM ZnSO4, pH 7.5 was diluted to the desired concentration
with the same buffer and was brought to DMSO concentration identical
to that of the injected compound. The tested compound was solubilized
in DMSO at 20 mM concentration and was diluted to 200 µM with the
enzyme buffer, resulting in a final DMSO concentration of 1%.
In a standard experiment, VIM-2 (19 µM) was titrated by one initi-
ating injection (0.5 µL) followed by 19 injections (2 µL) of compound 3
(200 µM) at an interval of 150 s. Dilution heat of compound injections
into buffer, at the corresponding DMSO concentration, were subtracted
from raw data.
[4] T.R. Walsh, M.A. Toleman, The emergence of pan-resistant Gram-negative
pathogens merits a rapid global political response, J. Antimicrob. Chemother. 67
(2012) 1–3, https://doi.org/10.1093/jac/dkr378.
[5] W.C. Reygaert, An overview of the antimicrobial resistance mechanisms of
bacteria, AIMS Microbiol. 4 (2018) 482–501, https://doi.org/10.3934/
microbiol.2018.3.482.
¨
[6] A. Cassini, L.D. Hogberg, D. Plachouras, A. Quattrocchi, A. Hoxha, G.S. Simonsen,
M. Colomb-Cotinat, M.E. Kretzschmar, B. Devleesschauwer, M. Cecchini, D.
A. Ouakrim, T.C. Oliveira, M.J. Struelens, C. Suetens, D.L. Monnet, Burden of AMR
Collaborative Group, Attributable deaths and disability-adjusted life-years caused
by infections with antibiotic-resistant bacteria in the EU and the European
Economic Area in 2015: a population-level modelling analysis, Lancet Infect. Dis.
19 (2019) 56–66, https://doi.org/10.1016/S1473-3099(18)30605-4.
[7] K. Bush, Past and present perspectives on β-lactamases, Antimicrob. Agents
Chemother. 62 (2018) e01076–18, https://doi.org/10.1128/AAC.01076-18.
[8] T. Palzkill, Metallo-β-lactamase structure and function, Ann. N. Y. Acad. Sci. 1277
(2013) 91–104, https://doi.org/10.1111/j.1749-6632.2012.06796.x.
[9] V.R. Gajamer, A. Bhattacharjee, D. Paul, C. Deshamukhya, A.K. Singh, N. Pradhan,
H.K. Tiwari, Escherichia coli encoding blaNDM-5 associated with community-
acquired urinary tract infections with unusual MIC creep-like phenomenon against
imipenem, J. Glob. Antimicrob. Resist. 14 (2018) 228–232, https://doi.org/
10.1016/j.jgar.2018.05.004.
The data so obtained were fitted via the non-linear least squares
minimization method to determine binding stoichiometry (n), associa-
tion constant (Ka), and change in enthalpy of binding (ΔH◦ ) using
ORIGIN 7 software v.7 (OriginLab). In the fitting procedure, tbhe com-
pound concentration was lowered in order to provide n around 1 and to
take compound solubility into account under ITC conditions. The Gibbs
[10] S.E. Boyd, D.M. Livermore, D.C. Hooper, W.W. Hope, Metallo-β-lactamases:
structure, function, epidemiology, treatment options, and the development
pipeline, Antimicrob. Agents Chemother. 64 (2020) e00397–20, https://doi.org/
10.1128/AAC.00397-20.
free energy of binding, ΔG◦ , was calculated from Ka (ΔG◦b = ꢀ RT ln Ka)
b
and the entropic term, TΔS◦ , was derived from the Gibbs-Helmholtz
b
equation using the experimental ΔH◦b value (ΔG◦b = ΔH◦b ꢀ TΔS◦ ).
b
[11] C. Gonzalez-Bello, D. Rodriguez, M. Pernas, A. Rodriguez, E. Colchon, β-Lactamase
inhibitors to restore the efficacy of antibiotics against superbugs, J. Med. Chem. 63
(2020) 1859–1881, https://doi.org/10.1021/acs.jmedchem.9b01279.
[12] J.-D. Docquier, S. Mangani, An update on β-lactamase inhibitor discovery and
development, Drug Resist. Updat. 36 (2018) 13–29, https://doi.org/10.1016/j.
drup.2017.11.002.
4.4. Molecular modelling
The docking studies were performed with AutoDock 4.2.6 [60] as
previously described [43]. For each ligand, free rotations were defined.
Files for the docking were prepared from: (i) the structure of complex
VIM-2/JMV4690 (PDB code 6YRP) [38] (the protein was protonated
with Dockprep module of Chimera and the program was let to itself
protonate His residues, while checking the absence of conflict in the H-
bond network); (ii) compounds and protein pdbqt files prepared with
AutoDockTools (ADT) [61]. For the protein, seven side-chains were
allowed to move: Phe61, Asp62, Tyr67, Asp119, Glu149, Arg228 and
Asn233 (to visualize them, see Figure S1).
[13] C.J. Burns, D. Daigle, B. Liu, D. McGarry, D.C. Pevear, R.E. Trout, β-Lactamase
inhibitors. WO Patent WO 2014/089365 A1.
˜
[14] A. Krajnc, J. Brem, P. Hinchliffe, K. Calvopina, T.D. Panduwawala, P.A. Lang, 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
β-lactamases, J. Med. Chem. 62 (2019) 8544–8556, https://doi.org/10.1021/acs.
jmedchem.9b00911.
[15] 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. Weis, D.C. Pevear, L. Xerri, C.J. Burns, Discovery of
Taniborbactam (VNRX-5133): a broad spectrum serine- and metallo-β-lactamase
inhibitor for carbapenem-resistant bacterial infections, J. Med. Chem. 63 (2020)
2789–2801, https://doi.org/10.1021/acs.jmedchem.9b01518.
Molecular graphics and analyses were performed with UCSF
Chimera, developed by the Resource of Biocomputing, Visualization and
Informatics at the University of California, San Francisco, with support
from NIH P41-GM103311.
[16] J.C. Vazquez-Ucha, J. Arca-Suarez, G. Bou, A. Beceiro, New carbapenemase
inhibitors : clearing the way for the β-lactams, Int. J. Mol. Sci. 21 (2020) 9308,
https://doi.org/10.3390/ijms21239308.
[17] M. Everett, N. Sprynski, A. Coelho, J. Castandet, M. Bayet, J. Bougnon, C. Lozano,
D.T. Davies, S. Leiris, M. Zalacain, I. Morrissey, S. Magnet, K. Holden, P. Warn,
F. De Luca, J.-D. Docquier, M. Lemonnier, Discovery of a novel metallo-β-lactamase
inhibitor that potentiates meropenem activity against carbapenem-resistant
Enterobacteriaceae, Antimicrob. Agents Chemother. 62 (2018) e00074–18, https://
doi.org/10.1128/AAC.00074-18.
Declaration of Competing Interest
[18] S. Leiris, A. Coelho, J. Castandet, M. Bayet, C. Lozano, J. Bougnon, J. Bousquet,
M. Everett, M. Lemonnier, N. Sprynski, M. Zalacain, T.D. Pallin, M.C. Cramp,
N. Jennings, G. Raphy, M.W. Jones, R. Pattipati, B. Shankar,
The authors declare no competing financial interest.
R. Sivasubrahmanyam, A.K. Soodhagani, R.R. Juventhala, N. Pottabathini,
S. Pothukanuri, M. Benvenuti, C. Pozzi, S. Mangani, F. De Luca, G. Cerboni, J.-
D. Docquier, D.T. Davies, SAR studies leading to the identification of a novel series
of metallo-β-lactamase inhibitors for the treatment of carbapenem-resistant
Enterobacteriaceae infections that display efficacy in an animal infection model,
ACS Infect. Dis. 5 (2019) 131–140, https://doi.org/10.1021/acsinfecdis.8b00246.
[19] N. Reddy, M. Shungube, P.I. Arvidsson, S. Baijnath, H.G. Kruger, T. Govender,
T. Naicker, A 2018–2019 patent review of metallo-β-lactamase inhibitors, Exp.
Opin. Ther. Pat. 30 (2020) 541–555, https://doi.org/10.1080/
Acknowledgments
Part of this work was supported by Agence Nationale de la Recherche
(ANR-14-CE16-0028-01, including fellowship to L.S.). We thank Mr
Pierre Sanchez for mass spectrometry analyses. Thanks are also due to
Prof. Liam Good (Royal Veterinary College, London, U.K.) and Prof.
Lynn Zechiedrich (Bayor College of Medicine, Houston, USA) for
providing the E. coli strains AS19 and LZ2310, respectively. We are
extremely grateful to Prof. Martti Vaara (Northern Antibiotics Ltd.,
Espoo, Finland) for providing compound NAB741.
13543776.2020.1767070.
[20] R.P. McGeary, D.T. Tan, G. Schenk, Progress toward inhibitors of metallo-
β-lactamases, Future Med. Chem. 9 (2017) 673–691, https://doi.org/10.4155/fmc-
2017-0007.
13

L. Gavara et al.
Bioorganic Chemistry 113 (2021) 105024
[21] A.R. Palacios, M.-A. Rossi, G.S. Mahler, A.J. Vila, Metallo-β-lactamase inhibitors
inspired on snapshots from the catalytic mechanism, Biomolecules 10 (2020) 854,
https://doi.org/10.3390/biom10060854.
[39] F. Spyrakis, M. Santucci, L. Maso, S. Cross, E. Gianquinto, F. Sannio, F. Verdirosa,
F. De Luca, J.-D. Docquier, L. Cendron, D. Tondi, A. Venturelli, G. Cruciani, M.
P. Costi, Virtual screening identifies broad-spectrum β-lactamase inhibitors with
activity on clinically relevant serine- and metallo-carbapenemases, Sci. Rep. 10
(2020) 12763, https://doi.org/10.1038/s41598-020-69431-y.
[22] A.Y. Chen, R.N. Adamek, B.L. Dick, C.V. Credille, C.N. Morrison, S.M. Cohen,
Targeting metalloenzymes for therapeutic intervention, Chem. Rev. 119 (2019)
1323–1455, https://doi.org/10.1021/acs.chemrev.8b00201.
´
[40] P. Vella, W.M. Hussein, E.W. Leung, D. Clayton, D.L. Ollis, N. Mitic, G. Schenk, R.
´
[23] B.M. Lienard, G. Garau, L. Horsfall, A.I. Karsisiotis, C. Damblon, P. Lassaux,
P. McGeary, The identification of new metallo-β-lactamase inhibitor leads from
fragment-based screening, Bioorg. Med. Chem. Lett. 21 (2011) 3282–3285,
https://doi.org/10.1016/j.bmcl.2011.04.027.
`
C. Papamicael, G.C. Roberts, M. Galleni, O. Dideberg, J.-M. Frere, C.J. Schofield,
Structural basis for the broad-spectrum inhibition of metallo-β-lactamases by
thiols, Org. Biomol. Chem. 6 (2008) 2282–2294, https://doi.org/10.1039/
b802311e.
[41] F. Spyrakis, G. Celenza, F. Marcoccia, M. Santucci, S. Cross, P. Bellio, L. Cendron,
M. Perilli, D. Tondi, Structure-based virtual screening for the discovery of novel
inhibitors of New Delhi Metallo-β-lactamase-1, ACS Med. Chem. Lett. 9 (2017)
45–50, https://doi.org/10.1021/acsmedchemlett.7b00428.
[24] P. Lassaux, M. Hamel, M. Gulea, H. Delbrück, P.S. Mercuri, L. Horsfall,
`
D. Dehareng, M. Kupper, J.-M. Frere, K. Hoffmann, M. Galleni, C. Bebrone,
Mercaptophosphonate compounds as broad-spectrum inhibitors of the metallo-
β-lactamases, J. Med. Chem. 53 (2010) 4862–4876, https://doi.org/10.1021/
jm100213c.
[42] L. Sevaille, L. Gavara, C. Bebrone, F. De Luca, L. Nauton, M. Achard, P. Mercuri,
S. Tanfoni, L. Borgianni, C. Guyon, P. Lonjon, G. Turan-Zitouni, J. Dzieciolowski,
´
`
K. Becker, L. Benard, C. Condon, L. Maillard, J. Martinez, J.-M. Frere, O. Dideberg,
M. Galleni, J.-D. Docquier, J.-F. Hernandez, 1,2,4-Triazole-3-thione compounds as
inhibitors of dizinc metallo-β-lactamase, ChemMedChem 12 (2017) 972–985,
https://doi.org/10.1002/cmdc.201700186.
[25] M.M. Gonzalez, M. Kosmopoulou, M.F. Mojica, V. Castillo, P. Hinchliffe,
I. Pettinati, J. Brem, C.J. Schofield, G. Mahler, R.A. Bonomo, L.I. Llarrull,
J. Spencer, A.J. Vila, Bisthiazolidines: a substrate-mimicking scaffold as an
inhibitor of the NDM-1 carbapenemase, ACS Inf. Dis. 1 (2015) 544–554, https://
doi.org/10.1021/acsinfecdis.5b00046.
[43] L. Gavara, F. Verdirosa, A. Legru, P.S. Mercuri, L. Nauton, L. Sevaille, G. Feller,
D. Berthomieu, F. Sannio, F. Marcoccia, S. Tanfoni, F. De Luca, N. Gresh,
M. Galleni, J.-D. Docquier, J.-F. Hernandez, 4-(N-Alkyl- and -acyl-amino)-1,2,4-
triazole-3-thione analogs as metallo-β-lactamase inhibitors: impact of 4-linker on
potency and spectrum of inhibition, Biomolecules 10 (2020) 1094, https://doi.org/
10.3390/biom10081094.
[26] J.H. Toney, G.G. Hammond, P.M. Fitzgerald, N. Sharma, J.M. Balkovec, G.
P. Rouen, S.H. Olson, M.L. Hammond, M.L. Greenlee, Y.D. Gao, Succinic acids as
potent inhibitors of plasmid-borne IMP-1 metallo-β-lactamase, J. Biol. Chem. 276
(2001) 31913–31918, https://doi.org/10.1074/jbc.M104742200.
[27] 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-β-lactamase-1, J. Med. Chem. 60 (2017)
7267–7283, https://doi.org/10.1021/acs.jmedchem.7b00407.
[44] R.F. Deprez-Poulain, J. Charton, V. Leroux, B.P. Deprez, Convenient synthesis of
4H–1,2,4-triazole-3-thiols using di-2-pyridylthionocarbamate, Tetrahedron Lett. 48
(2007) 8157–8162, https://doi.org/10.1016/j.tetlet.2007.09.094.
´
`
[45] M. Masi, M. Refregiers, K.M. Pos, J.-M. Pages, Mechanisms of envelope
permeability and antibiotic influx and efflux in Gram-negative bacteria, Nat.
Microbiol. 2 (2017) 17001, https://doi.org/10.1038/nmicrobiol.2017.1.
[46] E. Freire, Do enthalpy and entropy distinguish first class from best in class? Drug
Discov. Today 13 (2008) 869–874, https://doi.org/10.1016/j.drudis.2008.07.005.
[47] J.E. Ladbury, Calorimetry as a tool for understanding biomolecular interactions
and an aid to drug design, Biochem. Soc. Trans. 38 (2010) 888–893, https://doi.
org/10.1042/BST0380888.
[28] 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-
β-lactamase antibiotic resistance, Nature 510 (2014) 503–506, https://doi.org/
10.1038/nature13445.
[29] A. Bergstrom, A. Katko, Z. Adkins, J. Hill, Z. Cheng, M. Burnett, H. Yang, M. Aitha,
M.R. Mehaffey, J.S. Brodbelt, K.H. Tehrani, N.I. Martin, R.A. Bonomo, R.C. Page, D.
L. Tierney, W. Fast, G.D. Wright, M.W. Crowder, Probing the interaction of
aspergillomarasmine A with metallo-β-lactamase NDM-1, VIM-2, and IMP-7, ACS
Infect. Dis. 4 (2018) 135–145, https://doi.org/10.1021/acsinfecdis.7b00106.
[30] A. Matsuura, H. Okumura, R. Asakura, N. Ashizawa, M. Takahashi, F. Kobayashi,
N. Ashikawa, K. Arai, Pharmacological profiles of aspergillomarasmines as
endothelin converting enzyme inhibitors, Jpn J. Pharmacol. 63 (1993) 187–193,
https://doi.org/10.1254/jjp.63.187.
`
[48] L. Borgianni, J. Vandenameele, A. Matagne, L. Bini, R. Bonomo, J.-M. Frere, G.
M. Rossolini, J.-D. Docquier, Mutational analysis of VIM-2 reveals an essential
determinant for metallo-β-lactamase stability and folding, Antimicrob. Agents
Chemother. 54 (2010) 3197–3204, https://doi.org/10.1128/AAC.01336-09.
[49] R.P. Lamers, J.F. Cavallari, L.L. Burows, The efflux inhibitor phenylalanine-
arginine-naphthylamide (PAN) permeabilizes the outer membrane of gram-
negative bacteria, PloS one 8 (2013), e60666, https://doi.org/10.1371/journal.
pone.0060666.
¨
[31] O. Samuelsen, O.A.H. Astrand, C. Frohlich, A. Heikal, S. Skagseth, T.J.O. Carlsen,
H.S. Leiros, A. Bayer, C. Schnaars, G. Kildahl-Andersen, S. Lauksund, S. Finke,
S. Huber, T. Gjoen, A.M.S. Andresen, O.A. Okstad, P. Rongved, ZN148 – a modular
synthetic metallo-β-lactamase inhibitor reverses carbapenem-resistance in Gram-
negative pathogens in vivo, Antimicrob. Agents Chemother. 64 (2020) e02415–19,
https://doi.org/10.1128/AAC.02415-19.
[50] M. Vaara, Novel derivatives of polymyxins, J. Antimicrob. Chemother. 68 (2013)
1213–1219, https://doi.org/10.1093/jac/dkt039.
[51] M. Vaara, Polymyxin derivatives that sensitize Gram-negative bacteria to other
antibiotics, Molecules 24 (2019) 249, https://doi.org/10.3390/
molecules24020249.
´
[32] J. Brem, R. Cain, S. Cahill, M.A. McDonough, I.J. Clifton, J.C. Jimenez-Castellanos,
[52] E.F. Pettersen, T.D. Goddard, C.H. Huang, G.S. Couch, D.M. Greenblatt, E.C. Meng,
T.E. Ferrin, U.C.S.F. Chimera, a visualization system for exploratory research and
analysis, J. Comput. Chem. 25 (2004) 1605–1612, https://doi.org/10.1002/
jcc.20084.
M.B. Avison, J. Spencer, C.W. Fishwick, C.J. Schofield, Structural basis of metallo-
β-lactamase, serine-β-lactamase and penicillin-binding protein inhibition by cyclic
boronates, Nat. Commun. 7 (2016) 12406, https://doi.org/10.1038/
ncomms12406.
`
[53] J.-D. Docquier, J. Lamotte-Brasseur, M. Galleni, G. Amicosante, J.M. Frere, G.
[33] S.J. Hecker, K.R. Reddy, O. Lomovskaya, D.C. Griffith, D. Rubio-Aparicio,
K. Nelson, R. Tsivkovski, D. Sun, M. Sabet, Z. Tarazi, J. Parkinson, M. Totrov, S.
H. Boyer, T.W. Glinka, O.A. Pemberton, Y. Chen, M.N. Dudley, Discovery of cyclic
boronic acid QPX7728, an ultra-broad-spectrum inhibitor of serine and metallo-
β-lactamases, J. Med. Chem. 63 (2020) 7491–7507, https://doi.org/10.1021/acs.
jmedchem.9b01976.
M. Rossolini, On functional and structural heterogeneity of VIM-type metallo-
β-lactamases, J. Antimicrob. Chemother. 51 (2003) 257–266, https://doi.org/
10.1093/jac/dkg067.
[54] Clinical Laboratory Standard Institute, Performance standards for antimicrobial
disk susceptibility tests; approved standard, Document M02-A12, 2015, Twelfth
Edition, Wayne, PA, USA.
[34] L. Olsen, S. Jost, H.W. Adolph, I. Pettersson, L. Hemmingsen, F.S. Jørgensen, New
leads of metallo-β-lactamase inhibitors from structure-based pharmacophore
design, Bioorg. Med. Chem. 14 (2006) 2627–2635, https://doi.org/10.1016/j.
bmc.2005.11.046.
[55] Clinical Laboratory Standard Institute, Methods for dilution antimicrobial
susceptibility tests for bacteria that grow aerobically, Document M07-A10, 2015,
Twelfth Edition, Wayne, PA, USA.
[56] M. Avalos, M. Boetzer, W. Pirovano, N.E. Arenas, S. Douthwaite, G.P. van Wezel,
Complete genome sequence of Escherichia coli AS19, an antibiotic-sensitive
variant of E. coli strain B REL606, Genome Announc. 6 (2018) e00385-18, https://
doi.org/10.1128/genomeA.00385-18.. DOI: .
[35] K. Kwapien, M. Damergi, S. Nader, L. El Khoury, Z. Hobaika, R.G. Maroun, J.-
P. Piquemal, L. Gavara, D. Berthomieu, J.-F. Hernandez, N. Gresh, Calibration of
1,2,4-triazole-3-thione, an original Zn-binding group of metallo-β-lactamase
inhibitors. Validation of a polarizable MM/MD potential by quantum chemistry,
J. Phys. Chem. B 121 (2017) 6295–6312, https://doi.org/10.1021/acs.
jpcb.7b01053.
[57] S. Yang, S.R. Clayton, E.L. Zechiedrich, Relative contributions of the AcrAB, MdfA
and NorE efflux pumps to quinolone resistance in Escherichia coli, J. Antimicrob.
Chemother. 51 (2003) 545–556, https://doi.org/10.1093/jac/dkg126.
[58] S. Cagnacci, L. Gualco, S. Roveta, S. Mannelli, L. Borgianni, J.-D. Docquier, F. Dodi,
M. Centanaro, E. Debbia, A. Marchese, G.M. Rossolini, Bloodstream infections
caused by multidrug-resistant Klebsiella pneumoniae producing the carbapenem-
hydrolysing VIM-1 metallo-β-lactamase: first italian outbreak, J. Antimicrob.
Chemother. 61 (2008) 296–300, https://doi.org/10.1093/jac/dkm471.
[59] F. Luzzaro, J.-D. Docquier, C. Colinon, A. Endimiani, G. Lombardi, G. Amicosante,
G.M. Rossolini, A. Toniolo, Emergence in Klebsiella pneumoniae and Enterobacter
cloacae clinical isolates of the VIM-4 metallo-β-lactamase encoded by a conjugative
plasmid, Antimicrob. Agents Chemother. 48 (2004) 648–650, https://doi.org/
10.1128/aac.48.2.648-650.2004.
[36] L. Nauton, R. Kahn, G. Garau, J.-F. Hernandez, O. Dideberg, Structural insights into
the design of inhibitiors of the L1 metallo-β-lactamase from Stenotrophomonas
maltophilia, J. Mol. Biol. 375 (2008) 257–269, https://doi.org/10.1016/j.
jmb.2007.10.036.
[37] T. Christopeit, T.J. Carlsen, R. Helland, H.K. Leiros, Discovery of novel inhibitor
scaffolds against the metallo-β-lactamase VIM-2 by surface plasmon resonance
(SPR) based fragment screening, J. Med Chem. 58 (2015) 8671–8682, https://doi.
org/10.1021/acs.jmedchem.5b01289.
[38] L. Gavara, L. Sevaille, F. De Luca, P. Mercuri, C. Bebrone, G. Feller, A. Legru,
G. Cerboni, S. Tanfoni, D. Baud, G. Cutolo, B. Bestgen, G. Chelini, F. Verdirosa,
`
F. Sannio, C. Pozzi, M. Benvenuti, K. Kwapien, M. Fischer, K. Becker, J.-M. Frere,
[60] G.M. Morris, R. Huey, W. Lindstrom, M.F. Sanner, R.K. Belew, D.S. Goodsell, A.
J. Olson, AutoDock4 and AutoDockTools4: automated docking with selective
receptor flexibility, J. Comput. Chem. 16 (2009) 2785–2791, https://doi.org/
10.1002/jcc.21256.
S. Mangani, N. Gresh, D. Berthomieu, M. Galleni, J.-D. Docquier, J.-F. Hernandez,
4-Amino-1,2,4-triazole-3-thione-derived Schiff bases as metallo-β-lactamase
inhibitors, Eur. J. Med. Chem. 208 (2020), 112720, https://doi.org/10.1016/j.
ejmech.2020.112720.
[61] M.F. Sanner, Python: a programming language for software integration and
development, J. Mol. Graph. Mod. 17 (1999) 57–61. PMID: 10660911.
14