Zn148 is a modular synthetic metallo-β-lactamase inhibitor that reverses carbapenem resistance in Gram-negative pathogens in vivo

Article abstract of DOI:10.1128/AAC.02415-19

Carbapenem-resistant Gram-negative pathogens are a critical public health threat and there is an urgent need for new treatments. Carbapenemases (β- lactamases able to inactivate carbapenems) have been identified in both serine β-lactamase (SBL) and metallo-β-lactamase (MBL) families. The recent introduction of SBL carbapenemase inhibitors has provided alternative therapeutic options. Unfortunately, there are no approved inhibitors of MBL-mediated carbapenem-resistance and treatment options for infections caused by MBL-producing Gram-negatives are limited. Here, we present ZN148, a zinc-chelating MBL-inhibitor capable of restoring the bactericidal effect of meropenem and in vitro clinical susceptibility to carbapenems in >98percent of a large international collection of MBL-producing clinical Enterobacterales strains (n = 234). Moreover, ZN148 was able to potentiate the effect of meropenem against NDM-1-producing Klebsiella pneumoniae in a murine neutropenic peritonitis model. ZN148 showed no inhibition of the human zinc-containing enzyme glyoxylase II at 500 μM, and no acute toxicity was observed in an in vivo mouse model with cumulative dosages up to 128 mg/kg. Biochemical analysis showed a time-dependent inhibition of MBLs by ZN148 and removal of zinc ions from the active site. Addition of exogenous zinc after ZN148 exposure only restored MBL activity by -30percent, suggesting an irreversible mechanism of inhibition. Mass-spectrometry and molecular modeling indicated potential oxidation of the active site Cys221 residue. Overall, these results demonstrate the therapeutic potential of a ZN148-carbapenem combination against MBL-producing Gram-negative pathogens and that ZN148 is a highly promising MBL inhibitor that is capable of operating in a functional space not presently filled by any clinically approved compound.

Full text of DOI:10.1128/AAC.02415-19

AAC Accepted Manuscript Posted Online 16 March 2020
Antimicrob. Agents Chemother. doi:10.1128/AAC.02415-19
Copyright © 2020 Samuelsen et al.
This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International license.
1
2
Title: ZN148 – a modular synthetic metallo-β-lactamase inhibitor reverses carbapenem-
resistance in Gram-negative pathogens in vivo
3
Running title: ZN148 - a synthetic metallo-β-lactamase inhibitor
4
5
Authors: Ørjan Samuelsena,b,1 (ORCID: 0000-0002-5525-2614), Ove Alexander Høgmoen
Åstrandc (ORCID: 0000-0002-1490-9632), Christopher Fröhlicha,d (ORCID: 0000-0003-2682-
2267), Adam Heikale (ORCID: 0000-0002-5604-7145), Susann Skagsethd, Trine Josefine Olsen
Carlsend, Hanna-Kirsti S. Leirosd (ORCID: 0000-0002-2726-6322), Annette Bayerf (ORCID:
0000-0003-3481-200X), Christian Schnaarsc, Geir Kildahl-Andersenc, Silje Lauksunda,b, Sarah
Finkee, Sandra Huberg, Tor Gjøene, Adriana Magalhaes Santos Andresene, Ole Andreas Økstade
(ORCID: 0000-0001-9351-8535), Pål Rongvedc,1 (ORCID: 0000-0001-6678-1952)
Author Affiliations: aNorwegian National Advisory Unit on Detection of Antimicrobial
Resistance, Department of Microbiology and Infection Control, University Hospital of North
Norway, Tromsø, Norway. bDepartment of Pharmacy, UiT The Arctic University of Norway,
Tromsø, Norway. cSection for Pharmaceutical Chemistry, Department of Pharmacy, University
of Oslo, Oslo, Norway. dThe Norwegian Structural Biology Centre (NorStruct), UiT The Arctic
University of Norway, Tromsø, Norway. eCentre for Integrative Microbial Evolution (CIME)
and Section for Pharmacology and Pharmaceutical Biosciences, Department of Pharmacy,
University of Oslo, Blindern, Oslo, Norway. fDepartment of Chemistry, UiT The Arctic
University of Norway, Tromsø, Norway. gDepartment of Laboratory Medicine, Division of
Diagnostic Services, University Hospital of North Norway, Tromsø, Norway.
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
1 To whom correspondence may be addressed: E-mail: orjan.samuelsen@unn.no or
pal.rongved@farmasi.uio.no
1

25
Abstract:
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
Carbapenem-resistant Gram-negative pathogens are a critical public health threat and there is an
urgent need for new treatments. Carbapenemases (β-lactamases able to inactivate carbapenems)
have been identified in both serine β-lactamase (SBL) and metallo β-lactamase (MBL) families.
The recent introduction of SBL carbapenemase-inhibitors has provided alternative therapeutic
options. Unfortunately, there are no approved inhibitors of MBL-mediated carbapenem-
resistance and treatment options for infections caused by MBL-producing Gram-negatives are
limited. Here, we present ZN148, a zinc-chelating MBL-inhibitor capable of restoring the
bactericidal effect of meropenem and in vitro clinical susceptibility to carbapenems in >98% of a
large international collection of MBL-producing clinical Enterobacterales strains (n=234).
Moreover, ZN148 was able to potentiate the effect of meropenem against NDM-1-producing
Klebsiella pneumoniae in a murine neutropenic peritonitis model. ZN148 showed no inhibition
of the human zinc-containing enzyme glyoxylase II at 500 µM and no acute toxicity was
observed in an in vivo mouse model with cumulative dosages up to 128 mg/kg. Biochemical
analysis showed a time-dependent inhibition of MBLs by ZN148 and removal of zinc ions from
the active site. Addition of exogenous zinc after ZN148 exposure only restored MBL activity by
~30%, suggesting an irreversible mechanism of inhibition. Mass-spectrometry and molecular
modelling indicated potential oxidation of the active site Cys221 residue. Overall, these results
demonstrate the therapeutic potential of a ZN148-carbapenem combination against MBL-
producing Gram-negative pathogens and that ZN148 is a highly promising MBL inhibitor,
capable of operating in a functional space not presently filled by any clinically approved
compound.
47
48
2

49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
Introduction:
The global increase in antimicrobial resistance is currently undermining our ability to treat
bacterial infections and has become a critical public health threat worldwide. A cornerstone
treatment of serious and life-threatening infections caused by multidrug-resistant (MDR) Gram-
negative bacterial pathogens such as Klebsiella pneumoniae and Escherichia coli has been the
carbapenem β-lactam antibiotics (e.g. meropenem) (1). The major advantage of carbapenems has
been their relative stability towards β-lactamases, such as the extended-spectrum β-lactamases
(ESBLs) and AmpCs, which constitute common resistance mechanisms against β-lactams (2).
However, we now observe a global increase in dissemination and diversity of β-lactamases
(carbapenemases) with the ability to inactivate carbapenems (3). Estimates indicate that
carbapenem-resistant E. coli and K. pneumoniae caused around 3.6 million bloodstream or other
serious infections globally in 2014 (4). The impact of carbapenem-resistance is further illustrated
in a European study where carbapenem-resistance was shown to be the major contributor to the
burden of infections by antibiotic-resistant bacteria in many countries (5). Moreover, a common
feature of carbapenemase-producing Gram-negative bacteria is MDR, including resistance
towards non-β-lactam antimicrobials, resulting in severely limited treatment options (6).
β-Lactamases are divided into two main families and four classes, the serine β-lactamases
(SBLs, class A, C and D) and the metallo-β-lactamases (MBLs, class B) (2). The main
distinction between SBLs and MBLs is that SBLs possess an active site serine, while MBLs
require the presence of zinc ions for activity. β-lactamases with carbapenemase activity have
been identified in both of these families including SBLs such as KPC and OXA-48-like, and the
MBLs NDM, VIM and IMP (2). The recent introduction of serine carbapenemase inhibitors such
as avibactam, vaborbactam and relebactam used in combination with β-lactams, have provided
3

72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
treatment options against serine carbapenemase-producing Gram-negative pathogens (7, 8).
Unfortunately, none of these β-lactamase inhibitors possess inhibitory activity against MBLs.
The recent Italian outbreak of NDM-producing Enterobacteriaceae is significant due not only to
its size but also the change in epidemiology of carbapenem-resistant Enterobacteriaceae (CRE)
from endemic KPC-producing CRE to NDM-producing CRE and the subsequent reduction in
treatment options (9). Consequently, new treatment options for infections caused by MBL-
producing Gram-negatives, including NDM-producing Enterobacterales, are urgently required.
Possible treatment options includes cefiderocol (10) and the combination aztreonam-
avibactam (11, 12). Combinations of β-lactams and β-lactam enhancers such as zidebactam (13)
and nacubactam (14) have also shown promising activities. Moreover, several MBL inhibitors,
including aspergillomarasmine A (15), dipicolinic acid derivatives (16), ANT431 (17),
bisthiazolodines (18) and bismuth antimicrobials (19) have been reported. Recently, VNRX-
5133 (taniborbactam), a dual SBL and MBL inhibitor have shown potent activity in combination
with cefepime against MBL-producers (20). However, no direct MBL inhibitors are approved for
clinical use (21). Here, we report the pre-clinical development and characterization of a synthetic
and modular MBL-inhibitor (ZN148) with promising in vitro and in vivo efficacy.
Results and discussion:
Synthesis of ZN148. ZN148 is a construct of the zinc chelator tris-picolylamine (TPA) (22)
covalently linked to meglumine, a hydrophilic glucosyl side chain, through an N-methylated
amide bond, in order to lower the lipophilicity and toxicity of the chelator-conjugate (Fig. 1).
TPA is a known lipophilic zinc chelator with high affinity (1011 M) towards Zn2+ and has
previously been shown to inhibit MBLs (23-26). Synthesis of ZN148 is achieved through a high-
4

95
96
yield, 3-step synthesis from commercially available building blocks (Fig. S1), and the compound
exhibits very high aqueous solubility (1.225 g/mL in phosphate buffered saline).
97
98
In vitro activity of ZN148. Unmodified TPA re-sensitized NDM-1-producing K. pneumoniae
and VIM-2-producing P. aeruginosa clinical isolates towards meropenem by reducing the
minimum inhibitory concentration (MIC) from 32-64 mg/L to 0.125 and 1 mg/L, respectively
(Table S1). However, TPA also demonstrated a high degree of toxicity against human
hepatocarcinoma (HepG2) cells (IC50 9.8 µM). In contrast, ZN148 displayed not only a
comparatively reduced toxicity against HepG2 cells (IC50 >100 µM), which is most likely due to
its much lower cell permeability, but also retained the potentiation of meropenem activity and
fully restored meropenem clinical susceptibility according to EUCAST clinical breakpoints (27)
when tested at 50 µM against the same isolates (Table S1). This indicates that ZN148 is able to
penetrate the Gram-negative cell wall and enter the periplasm where the MBLs are located.
Alternatively, ZN148 could lower the environmental pool of available zinc, decreasing
periplasmic zinc and consequently the activity of MBLs (28). ZN148 exhibited no intrinsic
antibacterial activity at concentrations up to 500 µM, confirming that the potentiation of
meropenem is not due to combined antibacterial activity of the compounds.
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
ZN148 at 50 µM was further tested in combination with meropenem against an extended
international collection of 234 MBL-producing clinical Enterobacterales strains expressing
variants of NDM, VIM and IMP enzymes as well as other non-MBL β-lactamase variants.
Overall, the meropenem-ZN148 combination reduced the meropenem MIC to susceptible levels
(≤ 2 mg/L) in >98% of strains (MIC90 meropenem: ≥64 mg/L and MIC90 meropenem-ZN148:
0.5 mg/L) (Fig. 2A, and Table S2). The geographical distribution and diversity of the strain
5

118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
collection shows that ZN148 is not influenced by strain background or specific MBL-variants.
Similarly, when tested against a subset of 173 MBL-producing E. coli and K. pneumoniae,
ZN148 reduced the doripenem and imipenem MIC to susceptible levels in >99% of strains (Fig.
2B) indicating a potential to be used in combination with other carbapenems. Against MBL-
producing P. aeruginosa (n=52) and A. baumannii (n=6) strains, ZN148 exhibited less
potentiation of carbapenems, though still restored clinical susceptibility in 17%, 15% and 25% of
MBL-producing P. aeruginosa clinical strains in combination with meropenem, doripenem or
imipenem, respectively (Fig. 2C). Against NDM-1-producing class D carbapenemase-negative
A. baumannii strains, ZN148 reduced the meropenem MIC >2-fold in 4/6 clinical strains (Table
S3). The cause of the reduced effectiveness against P. aeruginosa and A. baumannii isolates is
not clear, but could be due to a more restricted outer membrane permeability than in
Enterobacterales (29). This could reduce the uptake of carbapenems or ZN148, and/or a range of
efflux systems which could contribute to carbapenem-resistance particularly in P. aeruginosa
(30). Alternatively, if ZN148’s mode of action is through lowering the environmental availability
of zinc, differential zinc uptake between Enterobacterales and A. baumannii/P. aeruginosa could
be involved.
With a few exceptions, no potentiation of meropenem, doripenem and imipenem was
observed in strains co-producing MBLs and class D carbapenemases (Table S4). Further, no
potentiation of meropenem, doripenem or imipenem was observed for ZN148 against strains
harboring only the class A carbapenemase KPC (Table S5), supporting the specificity towards
MBLs.
A time-dependent cell-killing assay revealed that ZN148 restored the bactericidal activity
of meropenem against an NDM-1-producing K. pneumoniae strain (Fig. 3). Meropenem alone (4
6

141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
mg/L) produced an initial bacteriostatic effect (2-log reduction in cell numbers), before
eventually reaching >1x109 colony forming units per milliliter (CFU/mL) after 24 hours. In
contrast, the combination of meropenem (4 mg/L) with either 50 or 100 μM ZN148 was
bactericidal, restoring a time-dependent killing mechanism. The kinetics of killing were similar
for both concentrations of ZN148 tested. Cell numbers were reduced to below the limit of
detection (1x102 CFU/mL) after 8 hours and no re-growth was observed within 24 hours,
indicating sterilizing bactericidal activity.
Following an investigation into the frequency of resistance (FoR) using single step
selection, we observed a concentration dependent reduction in FoR for both meropenem and
ZN148 (Table S6), with the FoR ranging between 10-7 to 10-8 using an NDM-1-producing K.
pneumoniae strain (K66-45). Whole genome shotgun sequencing (WGS) of stable, isolated
mutants revealed mutations within the outer membrane porin OmpK36 (locus tag:
B5G58_03310) and a LysR type transcriptional regulator (locus tag: B5G58_25480) previously
associated with changes in OmpC expression in E. coli (Table S7) (31). Serial passaging of K.
pneumoniae K66-45 against increasing concentrations of meropenem, in the presence of 50 and
100 M ZN148 resulted in 64x and 16x fold increase in MIC, respectively, compared to the
157
158
159
160
161
162
163
control culture (Fig. S3). WGS analysis of stable mutants isolated from the 100 M ZN148
serial-passaging condition (isolated at 16-fold meropenem MIC, 2 mg/L) also identified
mutations in OmpK36 and the same LysR type transcriptional regulator as from the single step
selection (Table S8). Evolution of NDM have shown the emergence of allelic variants with
enhanced zinc binding capability and increased ability to tolerate zinc starvation caused by metal
chelators (32). However, no mutations were observed in the blaNDM-1 gene in either spontaneous
(single-step) or serially passaged mutants emerging during exposure to ZN148, and no mutants
7

164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
reached clinical resistance levels. Taken together, these results demonstrated that development of
resistance to ZN148 and meropenem co-treatment is unlikely to be a barrier to further
development of ZN148 as an MBL inhibitor.
ZN148 potentiates the activity of meropenem in vivo with no acute toxicity in vivo. In a
murine neutropenic peritonitis model, subcutaneous treatment with a combination of meropenem
(33 mg/kg) and ZN148 (10 mg/kg) resulted in a significantly lower CFU/mL of a meropenem-
resistant NDM-1-producing K. pneumoniae strain in both peritoneal fluid (p<0.0001) and blood
(p<0.01), compared to treatment with meropenem alone (Fig. 4). Treatment with ZN148 alone
did not result in a reduction of CFU compared to vehicle treatment, corroborating the lack of
intrinsic antibacterial activity observed in vitro. The same effect was observed for 33 and 100
mg/kg ZN148 (Fig. S2).
As zinc is an essential metal ion for many biological processes (33, 34), inhibitors of
MBLs based on zinc chelation could potentially have off-target toxicity. To investigate this, we
tested the inhibitory activity of ZN148 against the human glyoxylase II enzyme, which shares the
MBL-protein fold and zinc-binding properties (35). In contrast to EDTA, a strong metal chelator,
ZN148 (500 µM) showed no inhibitory activity against recombinant glyoxylase II (Fig. 5A).
This indicates selectivity towards bacterial MBLs and a more specific mode of action. Selectivity
towards bacterial MBLs has also been shown for other zinc-chelating MBL inhibitors such as
aspargillomarasmine A (15). Moreover, the inhibitory activity of ZN148 was not influenced by
human serum albumin and α1-acid glycoprotein, indicating negligible serum protein binding
(Fig. S4). Additionally, no acute toxicity was observed in vivo in a maximum tolerated dose
study using female Balb/C mice. In four groups, containing twelve animals each, after weekly
8

187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
doubling of single doses from 4 mg/kg up to 128 mg/kg and a 7-week cumulative administration
of 252 mg/kg, no significant differences in clinical signs (hair loss or bristly hair, immobility,
prostration) or loss of weight (<10% as compared to initial weight) (Fig. 5B) were observed
between the test and control groups.
ZN148 irreversibly inhibits MBLs. The rationale behind inclusion of the TPA moiety in
ZN148 was to inhibit MBLs by the removal of their active site zinc ions. Using ICP-MS for
analysis of purified VIM-2 and NDM-1, we found that ZN148 removed ~1.8 molar and ~1.3
molar equivalents of zinc from VIM-2 and NDM-1, respectively (Fig. 6A). We also synthesized
three ZN148 analogues, ZN222, ZN223 and ZN228, where one or several 2-pyridinyl rings were
replaced with benzene rings, resulting in reduced chelator strength (23). Analysis of NDM-1 and
VIM-2 enzyme kinetics revealed that the inhibitory efficiency (kinact/Ki) of these compounds was
lower than for ZN148 and correlated with the theoretical chelator strength (Table 1). The relative
position of the pyridine- and benzene rings in the ZN148 derivatives (ZN223 and ZN228)
affected the inhibitory activity (Table 1). However, none of the ZN148 analogues were able to
potentiate meropenem activity towards an NDM-1-producing K. pneumoniae or a VIM-2-
producing P. aeruginosa strain, i.e. the meropenem MIC was unchanged (32 mg/L) both alone
and in combination with ZN222, ZN223 or ZN228, confirming that zinc chelation is required for
activity and constitutes the likely mode of inhibition for ZN148.
In vitro inactivation kinetics revealed that ZN148 inhibited VIM-2 (kinact/Ki: 6.6 min-1
mM-1) more effectively than NDM-1 (kinact/Ki: 0.19 min-1 mM-1) and showed a time-dependent
inhibition of VIM-2 and NDM-1 by ZN148 (Table 1), suggestive of an irreversible mechanism
of inhibition (36, 37). Following incubation of VIM-2 with ZN148, addition of exogenous zinc
9

210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
restored only ~30% of the MBL activity compared to the untreated control (Fig. 6B). Unfolding
and step-wise refolding of VIM-2 following incubation with ZN148 also resulted in restoration
of ~30% of the control activity (Fig. 6C). In contrast, incubation of VIM-2 with EDTA followed
by addition of zinc resulted in complete, or near complete (~80%) restoration of activity (Fig. 6B
and 6C). Taken together, these results demonstrate a difference in the mode of inhibition for
ZN148 compared to EDTA, and support an irreversible mechanism of inhibition of VIM-2 by
ZN148. The mechanism behind the irreversible inhibition is unclear, but several studies have
postulated that chelating agents increase the susceptibility of active site amino acids to chemical
modifications, such as oxidation of Cys221 to Ocs221 (38-40). ESI-MS analysis of purified
VIM-2 incubated with ZN148 revealed an increase in mass of 46.2±0.1Da (25,572.7±0.1 Da to
25,618.9±0.1 Da) (Fig. 6D), potentially representative of a deprotonated cysteine sulfonic acid
(47 Da). The observed mass increase indicates that removal of zinc by ZN148 renders the
enzyme more susceptible to irreversible chemical modification (e.g. oxidation of Cys221) (38,
39, 41), thus preventing the restoration of enzymatic activity by refolding upon addition of
exogenous zinc.
Molecular modeling of ZN148 into the active sites of VIM-2 (PDB ID: 5LSC) and NDM-
1 (PDB ID: 4RL0) predicted favorable aromatic interactions between ZN148 and F61, Y67, and
Y224 and cation-pi interactions to R228 in VIM-2 (Fig. 6E). The stacking of these aromatic side
chains would likely favor greater hydrophobic and cation-pi interactions of ZN148 with the
VIM-2 active site compared to the equivalent residues (M61, V67, K224 and A228) in NDM-1.
ZN148 in NDM-1 would then have fewer enzyme-inhibitor interactions (Fig. 6F), and a less
favorable ZN148 binding compared to VIM-2 (Table 1). This could indicate formation of a
transient enzyme:Zn:inhibitor complex followed by the removal of zinc ions from the active site
10

233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
as shown for several other MBL inhibitors (42). The inactivated Zn-depleted VIM-2 enzyme
demonstrated an increased mass of +46.2±0.1 Da, which is likely due to the oxidation of Cys221,
since oxidation to Ocs221 is observed more often in VIM (40, 43, 44) than in NDM-1 crystal
structures. Second shell residues, fold or zinc affinity might also account for an easier Cys221
oxidation in VIM-2 compared to NDM-1.
Conclusions. The lack of available therapies for infections caused by MBL-producing Gram-
negative bacteria has prompted the World Health Organization to classify them as ‘priority
pathogens’ for research and development of new and effective treatments (45). Discovery and
development of new antibiotics is fraught with difficulty and suffers from high attrition rates at
early stages of development (46). The successful clinical introduction of SBL-carbapenemase
inhibitors demonstrated an alternative strategy, preserving the efficacy of existing antibiotics -
however, this success has not to date been replicated in the treatment of MBL-producing CRE.
To meet the challenge posed by MBL-producing CRE, we adopted a strategy of
developing a synthetic, module zinc-chelator. Taken together, the activity spectrum of ZN148 in
combination with carbapenems, in vivo efficacy and limited in vivo toxicity demonstrate that
ZN148 has the potential to enter pre-clinical studies for further development. Furthermore, as a
modular and easy-to-synthesize MBL inhibitor that restores the activity of carbapenems towards
a range of clinically important Gram-negative MDR pathogens, ZN148 is well placed to enter
the clinical development pipeline along with other promising recently approved molecules (47)
in meeting the urgent, therapeutic challenge of global importance.
255
Materials and Methods:
11

256
257
The chemical synthesis of ZN148 and analogues ZN222, ZN223 and ZN228 is described in the
Supplemental material and Fig. S1.
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
Solubility measurements of ZN148 in phosphate buffered saline (PBS):
The solubility of ZN148 was evaluated in PBS buffer at pH 7.4. Increasing amounts of ZN148
was dissolved in 1.0 mL at room temperature by gentle swirling of the solution in a glass vial. A
total of 1.225 g ZN148 was successfully dissolved in 1.0 mL PBS buffer, giving a total volume
of 1.8 mL. This solution was stable at 4°C for more than 6 weeks without precipitation.
Antibacterial susceptibility testing and time-kill experiments:
Minimum inhibitory concentrations (MICs) of carbapenems alone and in combination with
inhibitors were determined by broth microdilution according to the Clinical and Laboratory
Standards Institute (CLSI) guidelines (48) either using pre-made plates (TREK Diagnostic
Systems/Thermo Fisher Scientific, East Grinstead, UK) with meropenem or in-house prepared
plates with meropenem, imipenem and doripenem. ZN148 and other inhibitors were tested at a
fixed concentration of 50 µM. Cation-adjusted Mueller Hinton medium (TREK Diagnostic
Systems/Thermo Fisher Scientific, East Grinstead, UK/Becton-Dickinson, Franklin Lakes, NJ,
USA) were used as growth medium. The plates were incubated for 20 h at 37°C. ZN148, ZN222,
ZN223, ZN228 and TPA were initially tested in combination with meropenem using two
indicator strains, one NDM-1-producing K. pneumoniae (49, 50) and one VIM-2-producing P.
aeruginosa (51). The activity of ZN148 were subsequently tested against two extended strain
collections: (i) in combination with meropenem against MBL-producing Enterobacterales
(n=62), MBL-producing P. aeruginosa (n=9) and MBL- and class D carbapenemase co-
producing Enterobacterales (n=5) from the strain collection at the Norwegian National Advisory
12

279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
Unit on Detection of Antimicrobial Resistance, Norway and (ii) in combination with
meropenem, imipenem and doripenem against globally collected MBL-producing E. coli (n=87),
K. pneumoniae (n=85), P. aeruginosa (n=52) and A. baumannii (n=6), MBL- and class A/D
carbapenemase co-producing E. coli (n=11), K. pneumoniae (n=15) and A. baumannii (n=7) at
IHMA Europe Sàrl, Switzerland. E. coli ATCC 25922 and P. aeruginosa ATCC 27853 were
used as quality control strains. The MBL inhibitors were also tested for intrinsic antibacterial
activity against the strains.
To determine the kinetics of killing of the meropenem-ZN148 combination, NDM-1-producing
Klebsiella pneumoniae K66-45 (1x106 CFU/mL) in cation-adjusted Mueller Hinton broth II
(Sigma-Aldrich, St. Louis, USA) was treated with either 4 mg/L meropenem or a combination of
4 mg/L meropenem plus 50 or 100 µM ZN148 and grown at 37°C with shaking (180 r.p.m.).
Cell number was determined over the course of 24 h by a modified Miles-Misra method (52).
Briefly, 10 mL serial dilutions of culture in sterile phosphate buffered saline (PBS) were spotted
and dried on square (10cm x 10cm) Luria-Bertani agar (Oxoid, Basingstoke, UK) plates at room
temperature before incubation at 37°C. Limit of detection was set as two colonies in 10 mL
undiluted culture, representing 100 CFU/mL.
In vivo efficacy study:
The in vivo efficacy of meropenem (Mylan, Mylan Hospitals, Asker, Norway) in combination
with ZN148 were evaluated in a murine neutropenic peritonitis model. Female NRMI mice
(Taconic Biosciences, Lille Skensved, Denmark) were rendered neutropenic by intraperitoneal
(i.p.) injection with cyclophosphamide (Baxter, Søborg Denmark) at four days (200 mg/kg) and
13

302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
one day (100 mg/kg) prior to inoculation. Mice were inoculated i.p. with ~5x106 CFU of NDM-
1-producing K. pneumoniae 50752501. One hour post-inoculation mice were injected
subcutaneously (s.c.) in the neck region with ZN148 corresponding to 10, 33 and 100 mg/kg or
vehicle. 30 min later the mice were injected s.c. with 33 mg/kg meropenem or vehicle. Four mice
were included for each treatment regime. Five hrs post-inoculation the mice were anesthetized
and blood was collected from axillary cutdown and the mice were sacrificed. Subsequently, two
ml sterile saline was injected i.p. and intraperitoneal fluid was sampled. Blood and
intraperitoneal fluid were serially diluted in 0.9% NaCl and plated on blood agar plates with 5%
horse blood (Statens Serum Institut, Copenhagen, Denmark) for CFU determination. Groups
were analyzed with ANOVA Dunnett´s multiple comparisons test in GraphPad Prism 7.04
(GraphPad Software Inc, USA). P-values <0.05 were considered statistically significant. All
animal experiments were approved by the National Committee of Animal Ethics, Denmark
(permission no. 2014-15-0201-00171) and adhered to the standards of EU Directive 2010/63/EU.
Single-step selection of K. pneumoniae K66-45 for growth on meropenem and ZN148 in
combination:
In order to determine the frequency of resistance to the combination of meropenem and ZN148, a
modified single-step selection experiment was carried out as previously described (53). Briefly,
NDM-1-producing K. pneumoniae K66-45 was grown from a single colony to approximately 109
CFU/ml and plated on cation-adjusted Mueller-Hinton broth II (Becton Dickinson, Franklin
Lakes, NJ USA) agar containing 30, 50, 100 or 200 M ZN148 and 0.5, 1, 2, 4 or 8 mg/L
323
324
meropenem, in combination. Colonies were counted after overnight incubation at 37 C. The
concentrations of meropenem chosen ranged from 4x to 64x MIC (MIC values in the absence of
14

325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
ZN148) and correlated with EUCAST clinical breakpoints for Enterobacterales, which define 2
mg/L as sensitive and >8 mg/L as resistant to meropenem (18).
Serial passaging of K. pneumoniae K66-45:
K. pneumoniae K66-45 was passaged in a 96-well microtiter plate with a consistent
concentration of ZN148 and increasing concentrations of meropenem. To start the passaging
cation-adjusted Mueller Hinton Broth II medium (Sigma-Aldrich) containing 0.0625 mg/L (0.5 x
MIC) and 50 or 100 µM of ZN148 was inoculated with 1% of overnight culture which was
grown without selection. Six biological replicates were passaged in parallel, and from each of the
six overnight cultures four wells were inoculated, resulting in 24 subcultures for each ZN148
concentration. After 24 hours of incubation at 37 °C and shaking at 180 rpm, the wells were
visually checked for growth. Passaging was continued with all subcultures showing growth by
transferring 1 µl to 100 µl of fresh medium added 0.125 mg/L meropenem (1 x MIC) and 50 or
100 µM of ZN148, respectively. This process was continued by increasing the meropenem
concentration by 2-fold every 24 hours until none of the subcultures showed growth.
DNA preparation and de novo whole genome sequencing of evolved clones from single-step
selection and serial passaging experiments:
Preparation of genomic DNA from NDM-1 producing K. pneumoniae K66-45 was carried out
using the MO BIO DNeasy UltraClean Microbial kit (Qiagen, USA) as previously described
(49). Whole genome shotgun (WGS) sequencing was carried out at either the Norwegian
Sequencing Centre (Oslo, Norway) or Novogene (Beijing, China) using the Illumina HiSeq
platform. Variations and single nucleotide polymorphisms in mutant WGS sequences with
15

348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
greater than 10x coverage were identified by mapping Illumina reads to the reference genome
(GenBank accession numbers CP020901 – CP020905) (49) using Bowtie2 (54) within Geneious
version 11.1.5 (55) and using the inbuilt Geneious variant/SNP finder command.
In vitro toxicity:
The human hepatoblastoma cell line HepG2 (HB-8065, ATCC, Manassas, VA, USA) was
cultured in DMEM-Glutamax™ (5.5 mM glucose, ThermoFisher, USA) supplemented with 10%
fetal bovine serum (Sigma-Aldrich, USA), streptomycin (100 μg/ml, Gibco/ThermoFisher, USA)
and penicillin (100 units/ml, Gibco/ThermoFisher, USA) at 37 °C in 5% CO2. For viability
assays, cells were seeded in white 96-well Nunclon plates (Sigma-Aldrich, USA) at a density of
20000 cells/well and left overnight to adhere. ZN148 and TPA (Sigma-Aldrich, Darmstadt,
Germany) dissolved in DMSO (Sigma-Aldrich, USA) were added to the white 96-well plates
containing 20000 HepG2 cells/well at concentrations ranging from 1 to 1 x 10-3 mM (DMSO
concentration kept below 1%) and incubated for 24 hours at 37°C in 5% CO2. After 24 hours,
AlamarBlue cell viability reagent (Thermo Fisher, Carlsbad, USA) was added (10% final
concentration) and incubated for 4 hours at 37°C. The proportion of viable cells in each well
was measured in a fluorescence plate reader (Clariostar, BMGlabtech, Germany) at ex550
nm/em603 nm (56). Data from eight replicates were fitted by non-linear regression to determine
IC50 values using GraphPad Prism (GraphPad Software Inc, USA).
In vivo tolerance:
In vivo tolerance of ZN148 was performed by Antineo (www.antineo.fr). Female Balb/c mice (4
weeks old, approximately 20 g, Charles River, L’Arbresle, France) were acclimatized 4 days in
16

371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
the animal facility before initiation of experiments. Groups of six mice were either untreated or
treated with 200 µL ZN148 (0.4-12.8 g/L) intraperitoneally once a week. Doses were doubled
each week (4-128 mg/kg) in the absence of any observed weight loss or modification in
behaviour. Individual weights were followed four days a week. Relative weight was calculated
as the ratio between the weight of the day and the weight at the initiation of the experiment.
Animals were also followed for macroscopic modifications in behavior. The protocol for
experiments in mice was approved by the University of Lyon Animal Ethics Committee (Comité
d'Ethique en Expérimentation Animale de l'Université Claude Bernard Lyon 1, authorization
number DR2015-09).
Protein binding:
Protein binding of ZN148 were investigated using the TransilXL plasma protein binding kit
(Sovicell, Leipzig, Germany) containing human serum albumin and α1-acid glycoprotein in a
ratio of 24:1. 500 M ZN148 in 50 mM HEPES buffer, pH 7.5 were added to different
concentrations (0–140 µM) of the protein mixture and incubated for 12 min at room temperature
with shaking at 1200 rpm. After incubation, the suspensions were centrifuged for 5 min on a
Minifuge (Starlab, Milton Keynes, UK) and the supernatant was diluted 1:10 in 50 mM HEPES
buffer, pH 7.5 supplemented with 1 µM ZnSO4 (Sigma Aldrich, St. Louis, USA). VIM-2 was
added at final concentration of 1 nM and the solution was incubated for 10 and 30 min at 25°C.
The enzyme activity was measured by adding nitrocefin (Merck, Darmstadt, Germany) at a final
concentration of 50 µM and measured at 482 nm at 25°C in a SpectraMax Plus plate reader
(Molecular Devices). Buffer alone was included as a control.
17

394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
Interaction with human glyoxylase II:
Recombinant human glyoxylase II (rHGly II, R&D systems, USA) was incubated with and
without ZN148 (final concentrations of 125, 250 and 500 µM) for 10 and 30 min at 25°C in 50
mM Tris buffer (Merck, Darmstadt, Germany) supplemented with 250 mM NaCl, pH 7.5. 50 µL
of substrate mixture containing 2 mM S-Lactoylglutathione and 400 µM 5,5`-dthiobis(2-
nitrobenzoic acid (Sigma Aldrich, St. Louis, USA) were mixed together with 50 µL of the pre-
incubated rHGly II at a final enzyme concentration of 0.2 ng/µL. The initial enzymatic velocity
was measured at 405 nm in 96 well plates (Thermo Fisher Scientific, Roskilde, Denmark Nunc)
at 25°C in a SpectraMax Plus plate reader (Molecular Devices). EDTA (Merck, Darmstadt,
Germany) was included as a control.
Time-dependent inactivation kinetics:
Protein expression and purification of NDM-1 and VIM-2 was done as previously described (36,
43, 57). Stock solutions of purified NDM-1 and VIM-2 were prepared in 50 mM HEPES buffer
(Merck, Darmstadt, Germany), pH 7.5. Inhibition of NDM-1 and VIM-2 were determined for
ZN148 and the analogues ZN222, ZN223 and ZN228 at different concentrations of inhibitor
after pre-incubation times of 2, 8, 15, 25 and 32 min in 50 mM HEPES buffer pH 7.5
supplemented with 1 µM ZnSO4 (Sigma Aldrich, St. Louis, USA) and bovine serum albumin
(BSA, Sigma Aldrich, St. Louis, USA, final concentration 2 µg/mL) at 25°C. A concentration of
1 nM VIM-2 and 30 nM NDM-1 were used and the reaction was initiated by the addition of 30
µM nitrocefin (Merck, Darmstadt, Germany) for VIM-2 or 100 µM imipenem (Sigma Aldrich,
St. Louis, USA) for NDM-1. The reaction was measured at 482 nm (VIM-2) or 300 nM (NDM-
1) in either standard 96 well plates (Thermo Fisher Scientific, Roskilde, Denmark) for VIM-2 or
18

417
418
419
420
421
422
423
424
425
426
427
UV-transparent 96 well plates (Corning, Kennebunk, ME, USA) for NDM-1 at 25°C in a
SpectraMax Plus plate reader (Molecular Devices). All enzyme and substrate concentrations
indicated are final concentrations in the assay. The enzyme activity in percent was calculated
based on the initial velocity and compared to the control without inhibitor compound. All tests
were performed at least in duplicates.
The observed rate constant (kobs) per inhibitor concentration was calculated from the slope of
semilog plot of enzyme activity in % versus preincubation time. The individual values of kobs
were plotted against the inhibitor concentration and saturation kinetics were fitted into Eq. 1 by
using Graph Pad Prim 4 (GraphPad Software Inc, USA) based on the following model:
K
k
I
428
429
430
431
432
433
434
E●Zn + I ↔ E●Zn●I →inact E + Zn●I and E*●Zn●I
where KI represents the inhibitor concentration that leads to an half-maximum inactivation of the
enzyme, kinact, states the first-order rate constant, E●Zn2+, the holoenzyme, I, the inhibitor,
E●Zn●I, the enzyme:Zn:inhibtior ternary complex, E, inactive Zn-depleted enzyme, Zn●I, the
zinc-inhibitor complex and E*●Zn●I, the inactive enzyme:Zn:inhibitor ternary complex (39).
kinact [I]
kobs
=
(1)
[S]
+ [I]
KI
435
436
437
By fitting these values, the irreversible kinetic parameters maximum inactivation rate (kinact) and
the inhibitor concentration that produces half-maximal rate of inactivation (KI) were obtained.
19

438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
Finally, the inhibitors were characterized by calculating kinact/KI. Where no saturation curve
could be observed, KI and kinact were determined from the linear part of plot 1/kobs versus 1/[I].
Zinc66 determination by inductively coupled plasma mass spectrometry (ICP-MS):
ICP-MS was used to investigate the chelating property of ZN148 by measuring the zinc
concentration (Zn66) after incubation of VIM-2 with and without ZN148 in zinc depleted 50 mM
HEPES buffer (chelex HEPES buffer), pH 7.5. The chelex buffer were prepared by stirring 2 g
chelex resin (Bio-Rad, Hercules, USA) in 100 mL 50 mM HEPES buffer, pH 7.5. The resin was
subsequently removed by sterile filtration (Merck MilliPore, 0.22 m). VIM-2 (10 g/L) was
diluted to 12.5 mg/L and mixed with ZN148 at different final concentrations (0, 12.5, 125 and
1250 µM) in chelex HEPES buffer. All solutions were allowed to incubate for 30 min on ice and
subsequently concentrated by centrifugation (at 4,000 x g, 4°C for 25 min) to a volume of 250
µL using centrifugal molecular cut-off filters (Merck MilliPore, 10,000 Da). The residual
inhibitor within these samples was diluted by factor of 5,000 using the same type molecular cut-
off filters and 50 mM chelex HEPES buffer, pH 7.5. The VIM-2 protein was concentrated to 0.2
g/L followed by a 1/16 dilution with 750 µL of a diluent mixture containing Rh103 (Inorganic
Ventures, Christiansburg, VA, USA) as internal standard. The diluent mixture consisted of Milli-
Q water (Millipore/ Merck KGaA, Darmstadt, Germany) with 2 µg/ Rh103, 2.5% v/v ammonia
solution (Honeywell Fluka, Bucharest, Romania), 0.08% v/v Triton X-100 (Sigma/Merck KGaA,
Darmstadt, Germany), 10% v/v isopropanol (Honeywell Fluka, Bucharest, Romania) and 0.25
µg/L Au (Inorganic Ventures, Christiansburg, VA, USA) as stabiliser. The samples were
introduced to the nebulizer (N2 gas flow 1.03 mL/min) by an ESI-Fast SC2DX autosampler with
a sample flow rate of 3 rpm and further into the NexION 300D ICP-MS system (Perkin Elmer,
20

461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
Waltham, MA, USA). Per inhibitor concentration of at least three biological replicates were
performed and injected and measured in triplicates. For the MS analysis the kinetic energy
discrimination (KED) mode with a helium flow rate of 5.7 mL/min, 20 sweeps per reading and a
dwell time of 100 ms/AMU for Zn66 and 50 ms/AMU for Rh103 were applied. The measurements
were performed with following instrumental settings: rf power (1,600 W), plasma gas flow (18
mL/min Ar), auxillary gas flow (1.2 mL/min N2), RPQ voltage (0.25 V) integration time (2,000
ms). All zinc concentrations were obtained by the internal standard method followed by a blank
subtraction using the NexION software version 1.5 (Perkin Elmer, Waltham, MA, USA). The
zinc concentration within the samples was determined based on an external calibration curve (0,
290, 580, 1,160 and 2,320 µg/L). As controls, VIM-2 protein without inhibitor, buffer control
and diluent blanks for control of instrumental carry over and one sample for quality control of
the measurements (580 µg/L standard) were included.
Liquid chromatography-electrospray ionisation Q-TOF mass spectrometry (LC-ESI-MS):
VIM-2 (8 µM final concentration) was incubated together without or with 5 mM ZN148 on ice
for 30 min. Incubation was carried out in chelex HEPES buffer, pH 7.5 and the inhibitor were
diluted in Milli-Q water (Millipore/Merck, Darmstadt, Germany) by a factor of 1,000,000 using
centrifugal filters (10,000 Da, Merck, Darmstadt, Germany). LC-ESI-MS (Agilent Technologies,
Santa Clara, CA, USA) was performed in positive ion mode using formic acid 0.1% in Milli-Q
water and acetonitrile (VWR, Radnor, USA) and 0.4 mL/min. Spectra deconvolution was
performed by using BioConfirm and MassHunter Qualitative Analysis (Agilent Technologies,
Santa Clara, CA, USA).
21

484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
Reversibility and zinc restoration of enzyme activity:
For reversibility of enzyme activity after incubation with ZN148, 3 mL VIM-2 or NDM-1 were
diluted with and without ZN148 in chelex HEPES buffer, pH 7.5 and subsequently incubated for
1 h on ice. 2.5 mL of each solution was loaded on an equilibrated PD-10 column (GE Healthcare,
Pittsburg, PA, USA), washed with 1 mL and eluted with 1.5 mL chelex HEPES buffer, pH 7.5,
respectively. Samples spiked and non-spiked with inhibitor were diluted to a final enzyme
concentration of 1 nM VIM-2 and 3 nM NDM-1. The initial reaction velocity was determined by
adding nitrocefin (Merck, Darmstadt, Germany) for VIM-2 or imipenem (Sigma Aldrich, St.
Louis, USA) for NDM-1 at a final concentration of 50 µM and 100 µM, respectively. The initial
enzyme reactions were recorded at 25°C. Inhibitor and enzyme alone were included as controls.
All measurements were performed at least in duplicates. For enzyme restorability, pre- and post-
column samples were supplemented with ZnSO4 (Sigma Aldrich, St. Louis, USA) to a final
concentration of 100 µM and allowed incubate for 5 min at 25°C and the initial enzyme velocity
was determined as described above.
Un- and refolding of VIM-2:
VIM-2 (0.4 µM) was pre-incubated on ice for 30 min with and without ZN148. Complete protein
unfolding was archived in 50 mM HEPES buffer, pH 7.5 supplemented with 1 µM ZnSO4
(Sigma Aldrich, St. Louis, USA) and 8 M urea (Merck, Darmstadt, Germany). Refolding was
achieved by a step-wise buffer exchange using centrifugal cut-off filters (Merck, Darmstadt,
Germany, 10 kDa) and decreasing concentrations of urea (0, 2, 4 and 6 M in 50 mM HEPES
buffer, pH 7.5). VIM-2 was diluted to a final concentration of 1 nM and the enzyme activity was
measured by adding nitrocefin (Merck, Darmstadt, Germany) to a final concentration of 50 µM.
22

507
508
509
510
511
512
513
514
515
Modelling of ZN148 into VIM-2 and NDM-1:
ZN148 was modelled into VIM-2 and NDM-1 based on x-ray crystallography data of VIM-7 co-
crystallized with ZN222 (C. Froelich and H-K. S. Leiros et al. unpublished), the protein-
inhibitor/substrate complex structures of VIM-2 with triazolylthioacetamide (PDB ID: 5LSC)
and NDM-1 bound to cefuroxime (PDB ID: 4RL0) by using Phenix version 1.12 (58). The 2-
Pyridinyl rings in ZN148 were first placed at the same sites as for ZN222, and then the linker of
the inhibitor was exiting the active site towards W87, N233 and residue 119 similar to the
inhibitor triazolylthioacetamide in VIM-2 (PDB ID: 5LSC).
516
517
Acknowledgments:
518
519
520
521
522
523
524
We acknowledge the contribution of Ida Kristine Østnes Hansen at the UiT The Arctic
University of Norway for performing the ESI-MS measurements. In vivo efficacy studies were
performed at Statens Serum Institut, Copenhagen, Denmark and we are grateful for the support
from Carina Vingsbo Lundberg. The in vivo tolerance study was performed by Antineo, France
and part of the antimicrobial susceptibility testing was performed by IHMA Europe Sàrl,
Switzerland. The whole genome sequencing was performed at the Norwegian Sequencing
Centre.
525
526
Funding:
527
528
We thankfully acknowledge funding from The Research Council of Norway (BIOTEK2021
program, project no. 244219 and FORNY program, project no. 273430), NOVO Pre Seed
23

529
530
531
532
(NOVO Nordisk Fonden ByGS/SSch), Statoil AS (now Equinor AS) the University of Oslo and
the Northern Norway Regional Health Authority (project no. SFP886-09). The funders had no
role in study design, data collection and interpretation, or the decision to submit the work for
publication.
533
534
Author contributions:
535
536
537
538
539
540
541
542
543
544
O.A.H.Å., A.B., and P.R. conceived and initiated the project; Ø.S., O.A.H.Å., C.F., A.H., H-
K.S.L., S.S., A.B., T.G., O.A.Ø., and P.R. conceived and designed experimental methodology;
O.A.H.Å., C.S., G.K-A., performed chemical synthesis; C.F., and S.L., performed antimicrobial
susceptibility testing; A.H., S.F., O.A.Ø., performed in vitro time-kill experiments, frequency of
resistance, serial passaging experiments and whole-genome sequencing; C.F., T.G., and
A.M.S.A., performed in vitro toxicity studies; S.S., and T.J.O.C., performed recombinant
expression and purification of VIM-2 and NDM-1; C.F. performed in vitro kinetic and mode of
action experiments; H-K.S.L., and C.F. performed molecular modelling; S.H., and C.F.,
performed ICP-MS analysis; all authors analyzed, interpreted and contextualized the data; Ø.S.,
O.A.H.Å., and C.F. wrote the initial draft; All authors contributed to the final version.
545
546
547
548
549
550
Competing interests:
P.R, O.A.H.Å, Ø.S, C.S, G.K-A have
a
patent application on the technology
(WO 2018033719 A1). All other authors declare no competing interests.
References:
24

551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
1.
Papp-Wallace KM, Endimiani A, Taracila MA, Bonomo RA. 2011. Carbapenems: past,
present, and future. Antimicrob Agents Chemother 55(11):4943-60.
2.
59.
3.
Enterobacteriaceae: the impact and evolution of a global menace. 2017. J Infect Dis
215(suppl_1):S28-S36.
4.
Consortium. 2018. Estimating the number of infections caused by antibiotic-resistant
Escherichia coli and Klebsiella pneumoniae in 2014: a modelling study. Lancet Glob Health.
6(9):e969-e79.
Bush K. The ABCD's of β-lactamase nomenclature. 2013. J Infect Chemother 19(4):549-
Logan LK, Weinstein RA. The epidemiology of carbapenem-resistant
Temkin E, Fallach N, Almagor J, Gladstone BP, Tacconelli E, Carmeli Y, DRIVE-AB
5.
Cassini A, Hogberg LD, Plachouras D, Quattrocchi A, Hoxha A, Simonsen GS, Colomb-
Cotinat M, Kretzschmar ME, Devleesschauwer B, Cecchini M, Ouakrim DA, Oliveira TC,
Struelens MJ, Suetens C, Monnet DL, Burden of AMR Collaborative Group. 2019. 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(1):56-66.
6.
Perez F, El Chakhtoura NG, Papp-Wallace KM, Wilson BM, Bonomo RA. 2016.
Treatment options for infections caused by carbapenem-resistant Enterobacteriaceae: can we
apply "precision medicine" to antimicrobial chemotherapy? Expert Opin Pharmacother
17(6):761-81.
7.
Zhanel GG, Lawrence CK, Adam H, Schweizer F, Zelenitsky S, Zhanel M, Lagacé-
Wiens PRS, Walkty A, Denisuik A, Golden A, Gin AS, Hoban DJ, Lynch 3rd JP, Karlowsky JA.
2018. Imipenem-relebactam and meropenem-vaborbactam: two novel carbapenem-β-lactamase
inhibitor combinations. Drugs 78(1):65-98.
8.
Wright H, Bonomo RA, Paterson DL. New agents for the treatment of infections with
Gram-negative bacteria: restoring the miracle or false dawn? 2017. Clin Microbiol Infect
23(10):704-12.
9.
metallo-β-lactamase producing carbapenem-resistant Enterobacteriaceae, Italy, 2018–2019 – 4
June 2019. ECDC: Stockholm; 2019.
European Centre for Disease Prevention and Control. Regional outbreak of New Delhi
10.
Zhanel GG, Golden AR, Zelenitsky S, Wiebe K, Lawrence CK, Adam HJ, Idowu T,
Domalaon R, Schweizer F, Zhanel MA, Lagacé-Wiens PRS, Walkty AJ, Noreddin A, Lynch 3rd
JP, Karlowsky JA. Cefiderocol: a siderophore cephalosporin with activity against carbapenem-
resistant and multidrug-resistant Gram-negative bacilli. 2019. Drugs. 79(3):271-89.
11.
Chew KL, Tay MKL, Cheng B, Lin RTP, Octavia S, Teo JWP. 2018. Aztreonam-
avibactam combination restores susceptibility of aztreonam in dual-carbapenemase-producing
Enterobacteriaceae. Antimicrob Agents Chemother 62:e00414-18.
12.
Aztreonam plus clavulanate, tazobactam, or avibactam for treatment of infections caused by
metallo-β-lactamase-producing Gram-negative bacteria. Antimicrob Agents Chemother
63:e00010-19.
Emeraud C, Escaut L, Boucly A, Fortineau N, Bonnin RA, Naas T, Dortet L. 2019.
13.
Moya B, Barcelo IM, Cabot G, Torrens G, Palwe S, Joshi P, Umarkar K, Takalkar S,
Periasamy H, Bhagwat S, Patel M, Bou G, Oliver A. 2019. In vitro and in vivo activities of β-
lactams in combination with the novel β-lactam enhancers zidebactam and WCK 5153 against
25

596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
multidrug-resistant metallo-β-lactamase-producing Klebsiella pneumoniae. Antimicrob Agents
Chemother 63:e00128-19.
14.
Mushtaq S, Vickers A, Woodford N, Haldimann A, Livermore DM. 2019. Activity of
nacubactam (RG6080/OP0595) combinations against MBL-producing Enterobacteriaceae. J
Antimicrob Chemother 74(4):953-60.
15.
King AM, Reid-Yu SA, Wang W, King DT, De Pascale G, Strynadka NC, Walsh TR,
Coombes BK, Wright GD. 2014. Aspergillomarasmine A overcomes metallo-β-lactamase
antibiotic resistance. Nature. 510(7506):503-6.
16.
Chen AY, Thomas PW, Stewart AC, Bergstrom A, Cheng Z, Miller C, Bethel CR,
Marshall SH, Credille CV, Riley CL, Page RC, Bonomo RA, Crowder MW, Tierney DL, Fast
W, Cohen SM. 2017. Dipicolinic acid derivatives as inhibitors of New Delhi metallo-β-
lactamase-1. J Med Chem 60(17):7267-83.
17.
Everett M, Sprynski N, Coelho A, Castandet J, Bayet M, Bougnon J, Lozano C, Davies
DT, Leiris S, Zalacain M, Morrissey I, Magnet S, Holden K, Warn P, De Luca F, Docquier J-D,
Lemonnier M. 2018. Discovery of a novel metallo-β-lactamase inhibitor that potentiates
meropenem activity against carbapenem-resistant Enterobacteriaceae. Antimicrob Agents
Chemother 62:e00074-18.
18.
Hinchliffe P, Gonzalez MM, Mojica MF, Gonzalez JM, Castillo V, Saiz C, Kosmopoulou
M, Tooke CL, Llarrull LI, Mahler G, Bonomo RA, Vila AJ, Spencer J. 2016. Cross-class
metallo-β-lactamase inhibition by bisthiazolidines reveals multiple binding modes. Proc Natl
Acad Sci U S A 113(26):E3745-54.
19.
Wang R, Lai TP, Gao P, Zhang H, Ho PL, Woo PC, Ma G, Kao RY-T, Li H, Sun H.
2018. Bismuth antimicrobial drugs serve as broad-spectrum metallo-β-lactamase inhibitors. Nat
Commun 9(1):439.
20.
Hamrick JC, Docquier J-D, Uehara T, Myers CL, Six DA, Chatwin CL, John KJ,
Vernacchio SF, Cusick SM, Trout REL, Pozzi C, De Luca F, Benvenuti M, Mangani S, Liu B,
Jackson RW, Moeck G, Xerri L, Burns CJ, Pevear DC, Daigle DM. 2020. VNRX-5133
(taniborbactam), a broad-spectrum inhibitor of serine- and metallo-β-lactamases, restores activity
of cefepime in Enterobacterales and Pseudomonas aeruginosa. Antimicrob Agents Chemother
64:e01963-19.
21.
World Health Organization. Antibacterial agents in clinical development: an analysis of
the antibacterial clinical development pipeline, including tuberculosis. Geneva.
2017;WHO/EMP/IAU/2017.12.
22.
Blackman A. 2005. The coordination chemistry of tripodal tetraamine ligands.
Polyhedron 24(1):1-39.
23.
Anderegg G, Hubmann E, Podder NG, Wenk F. 1977. Pyridine-Derivatives as
Complexing Agents .11. Thermodynamics of Metal-Complex Formation with Bis[(2-
Pyridyl)Methyl]-Amine, Tris[(2-Pyridyl)Methyl]-Amine and Tetrakis[(2-Pyridyl)Methyl]-Amine
Helv Chim Acta 60(1):123-40.
24.
Schnaars C, Kildahl-Andersen G, Prandina A, Popal R, Radix S, Le Borgne M, Gjøen T,
Andresen AMS, Heikal A, Økstad OA, Fröhlich C, Samuelsen Ø, Lauksund S, Jordheim LP,
Rongved P, Åstrand OAH. 2018. Synthesis and preclinical evaluation of TPA-based zinc
chelators as metallo-β-lactamase inhibitors. ACS Infect Dis 4(9):1407-22.
25.
Kildahl-Andersen G, Schnaars C, Prandina A, Radix S, Le Borgne M, Jordheim LP,
Gjøen T, Andresen AMS, Lauksund S, Fröhlich C, Samuelsen Ø, Rongved P, Åstrand OAH.
26

641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
2019. Synthesis and biological evaluation of zinc chelating compounds as metallo-β-lactamase
inhibitors. Medchemcomm 10(4):528-37.
26.
Prandina AR, Radix S, Le Borgne M, Jordheim LP, Bousfiha Z, Fröhlich C, Leiros, H-
K.S, Samuelsen Ø, Frøvold, E, Rongved P, Åstrand OAH. 2019. Synthesis and biological
evaluation of new dipicolylamine zinc chelators as metallo-β-lactamase inhibitors. Tetrahedron
75(11):1525-40.
27.
interpretation of MICs and zone diameters. 2018;Version 8.1, http://www.eucast.org.
28. Meini MR, Gonzalez LJ, Vila AJ. 2013. Antibiotic resistance in Zn(II)-deficient
environments: metallo-β-lactamase activation in the periplasm. Future Microbiol. 8(8):947-79.
29. Hancock RE, Brinkman FS. 2002. Function of Pseudomonas porins in uptake and efflux.
Annu Rev Microbiol 56:17-38.
30. Lister PD, Wolter DJ, Hanson ND. 2009. Antibacterial-resistant Pseudomonas
European Committee for Antimicrobial Susceptiblity Testing. Breakpoint tables for
aeruginosa: clinical impact and complex regulation of chromosomally encoded resistance
mechanisms. Clin Microbiol Rev 22(4):582-610.
31.
Connolly JP, Gabrielsen M, Goldstone RJ, Grinter R, Wang D, Cogdell RJ, Walker D,
Smith DGE, Roe AJ. 2016. A highly conserved bacterial D-serine uptake system links host
metabolism and virulence. PLoS Pathog 12(1):e1005359.
32.
Bahr G, Vitor-Horen L, Bethel CR, Bonomo RA, González LJ, Vila AJ. 2018. Clinical
evolution of New Delhi metallo-β-lactamase (NDM) optimizes resistance under Zn(II)
deprivation. Antimicrob Agents Chemother 62:e01849-17.
33.
Andreini C, Banci L, Bertini I, Rosato A. 2006. Zinc through the three domains of life. J
Proteome Res 5(11):3173-8.
34.
Chasapis CT, Loutsidou AC, Spiliopoulou CA, Stefanidou ME. 2012. Zinc and human
health: an update. Arch Toxicol 86(4):521-34.
35.
human metallo-β-lactamase fold proteins. Trends Biochem Sci 41(4):338-55.
36. Christopeit T, Albert A, Leiros HS. 2016. Discovery of a novel covalent non-β-lactam
inhibitor of the metallo-β-lactamase NDM-1. Bioorg Med Chem 24(13):2947-53.
37. Bergstrom A, Katko A, Adkins Z, Hill J, Cheng Z, Burnett M, Yang H, Aitha M,
Pettinati I, Brem J, Lee SY, McHugh PJ, Schofield CJ. 2016. The chemical biology of
Mehaffey MR, Brodbelt JS, Tehrani KHME, Martin NI, Bonomo RA, Page RC, Tierney DL,
Fast W, Wright GD, Crowder MW. 2018. Probing the Interaction of Aspergillomarasmine A
with metallo-β-lactamases NDM-1, VIM-2, and IMP-7. ACS Infect Dis 4(2):135-45.
38.
Gardonio D, Siemann S. 2009. Chelator-facilitated chemical modification of IMP-1
metallo-β-lactamase and its consequences on metal binding. Biochem Biophys Res Commun
381(1):107-11.
39.
Siemann S, Brewer D, Clarke AJ, Dmitrienko GI, Lajoie G, Viswanatha T. 2002. IMP-1
metallo-β-lactamase: effect of chelators and assessment of metal requirement by electrospray
mass spectrometry. Biochim Biophys Acta. 1571(3):190-200.
40.
Garcia-Saez I, Docquier JD, Rossolini GM, Dideberg O. 2008. The three-dimensional
structure of VIM-2, a Zn-β-lactamase from Pseudomonas aeruginosa in its reduced and oxidised
form. J Mol Biol 375(3):604-11.
41.
for characterizing cysteine redox status. Circ Cardiovasc Genet 5(5):591.
42. Ju LC, Cheng Z, Fast W, Bonomo RA, Crowder MW. 2018. The continuing challenge of
metallo-β-lactamase inhibition: mechanism matters. Trends Pharmacol Sci 39(7):635-47.
Murray CI, Van Eyk JE. 2012. Chasing cysteine oxidative modifications: proteomic tools
27

687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
43.
Christopeit T, Carlsen TJ, Helland R, Leiros HK. 2015. Discovery of novel inhibitor
scaffolds against the metallo-β-lactamase VIM-2 by surface plasmon resonance (SPR) based
fragment screening. J Med Chem 58(21):8671-82.
44.
Ø. 2014. His224 alters the R2 drug binding site and Phe218 influences the catalytic efficiency of
the metallo-β-lactamase VIM-7. Antimicrob Agents Chemother 58(8):4826-36.
45.
Leiros HK, Skagseth S, Edvardsen KS, Lorentzen MS, Bjerga GE, Leiros I, Samuelsen
Tacconelli E, Carrara E, Savoldi A, Harbarth S, Mendelson M, Monnet DL, Pulcini C,
Kahlmeter G, Kluytmans J, Carmeli Y, Ouellette M, Outterson K, Patel J, Cavaleri M, Cox EM,
Houchens CR, Grayson ML, Hansen P, Singh N, Theuretzbacher U, Magrini N, WHO Pathogens
Priority List Working Group. 2018. Discovery, research, and development of new antibiotics: the
WHO priority list of antibiotic-resistant bacteria and tuberculosis. Lancet Infect Dis 18(3):318-
27.
46.
confronting the challenges of antibacterial discovery. Nat Rev Drug Discov 6(1):29-40.
47. Theuretzbacher U, Outterson K, Engel A, Karlen A. 2019. The global preclinical
antibacterial pipeline. Nat Rev Microbiol 10.1038/s41579-019-0288-0.
48. Clinical & Laboratory Standards Instiute. 2015. Methods for dilution antimicrobial
susceptibility tests for bacteria that grow aerobically - Tenth Edition: Approved Standard M07-
A10. CLSI, Wayne, PA, USA.
49.
a multidrug-resistant, blaNDM-1-expressing Klebsiella pneumoniae K66-45 clinical isolate from
Norway. Genome Announc 5:e00601-17.
Payne DJ, Gwynn MN, Holmes DJ, Pompliano DL. 2007. Drugs for bad bugs:
Heikal A, Samuelsen Ø, Kristensen T, Økstad OA. 2017. Complete genome sequence of
50.
Samuelsen Ø, Thilesen CM, Heggelund L, Vada AN, Kummel A, Sundsfjord A.
Identification of NDM-1-producing Enterobacteriaceae in Norway. 2011. J Antimicrob
Chemother 66(3):670-2.
51.
A. 2009. The first metallo-β-lactamase identified in Norway is associated with a TniC-like
transposon in a Pseudomonas aeruginosa isolate of sequence type 233 imported from Ghana.
Antimicrob Agents Chemother 53(1):331-2.
52.
blood. J Hygiene 38(6):732-49.
53. Livermore DM, Mushtaq S, Barker K, Hope R, Warner M, Woodford N. 2012.
Characterization of β-lactamase and porin mutants of Enterobacteriaceae selected with
ceftaroline + avibactam (NXL104). J Antimicrob Chemother 67(6):1354-8.
54.
9(4):357-9.
Samuelsen Ø, Buaro L, Toleman MA, Giske CG, Hermansen NO, Walsh TR, Sundsfjord
Miles AA, Misra SS, Irwin JO. 1938. The estimation of the bactericidal power of the
Langmead B, Salzberg SL. Fast gapped-read alignment with Bowtie 2. 2012. Nat Meth
55.
Kearse M, Moir R, Wilson A, Stones-Havas S, Cheung M, Sturrock S, Buxton S, Cooper
A, Markowitz S, Duran C, Thierer T, Ashton B, Meintjes P, Drummond A. 2012. Geneious
Basic: an integrated and extendable desktop software platform for the organization and analysis
of sequence data. Bioinformatics 28(12):1647-9.
56.
O'Brien J, Wilson I, Orton T, Pognan F. 2000. Investigation of the Alamar Blue
(resazurin) fluorescent dye for the assessment of mammalian cell cytotoxicity. Eur J Biochem
267(17):5421-6.
57.
Lund BA, Leiros HK, Bjerga GE. 2014. A high-throughput, restriction-free cloning and
screening strategy based on ccdB-gene replacement. Microbial Cell Fact 13(1):38.
28

732
733
734
735
58.
Adams PD, Afonine PV, Bunkoczi G, Chen VB, Davis IW, Echols N, Headd JJ, Hung L-
W, Kapral GJ, Grosse-Kunstleve RW, McCoy AJ, Moriarty NW, Oeffner R, Read RJ,
Richardson DC, Richardson JS, Terwilliger TC, Zwart PH. 2010. PHENIX: a comprehensive
Python-based system for macromolecular structure solution. Acta Crystallogr D Biol Crystallogr
66(Pt 2):213-21.
736
737
738
29

739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
Figure legends:
Fig. 1: Chemical structure of ZN148
Fig. 2: Antimicrobial activity of carbapenem-ZN148 combinations. (A) Cumulative
meropenem (MEM) minimum inhibitory concentration (MIC) alone or in combination with 50
µM ZN148 against MBL-producing E. coli (n=112), K. pneumoniae (n=112) and other
Enterobacterales (n=10) and all strains (n=234) combined. (B) Cumulative doripenem (DOR)
and imipenem (IPM) minimum inhibitory concentration (MIC) alone or in combination with 50
µM ZN148, against MBL-producing E. coli (n=87) and MBL-producing K. pneumoniae (n=85).
(C) Cumulative meropenem (MEM), doripenem (DOR) and imipenem (IPM) minimum
inhibitory concentration (MIC) alone or in combination with 50 µM ZN148, against MBL-
producing P. aeruginosa. For MEM the collection included 61 strains, while for DOR and IPM
the collection included 52 strains.
Fig. 3: Time-kill assay. NDM-1-producing K. pneumoniae K66-45 challenged with either
meropenem (MEM) alone (4 mg/L, solid circles) or a combination of either 50 μM ZN148 plus
MEM (4 mg/L, inverted open triangles) or 100 μM ZN148 plus MEM (4 mg/L, open triangles).
Cell viability was expressed as log10 cfu/ml. Error bars represent the standard deviation from
three independent technical replicates, limit of detection is indicated by the dotted line.
Fig. 4: In vivo activity of meropenem (MEM) and MEM-ZN148 combination. Neutropenic
NMRI mice were inoculated intraperitoneally with ~5x106 CFU of NDM-1-producing K.
pneumoniae 50752501 (MEM MIC = 64 mg/L). Mice were treated by subcutaneously injection
in the neck region with vehicle (phosphate buffered saline), MEM (33 mg/kg) or MEM (33
mg/kg) + ZN148 (10 mg/kg). Vehicle and ZN148 were given one hour post-inoculation while
MEM was given 1.5 hrs post-inoculation. Colony counts in blood (A) and peritoneal fluid (B)
30