A fine-tuned lipophilicity/hydrophilicity ratio governs antibacterial potency and selectivity of bifurcated halogen bond-forming nbtis

Article abstract of DOI:10.3390/antibiotics10070862

Herein, we report the design of a focused library of novel bacterial topoisomerase inhibitors (NBTIs) based on innovative mainly monocyclic right-hand side fragments active against DNA gyrase and Topo IV. They exhibit a very potent and wide range of antibacterial activity, even against some of the most concerning hard-to-treat pathogens for which new antibacterials are urgently needed, as reported by the WHO and CDC. NBTIs enzyme activity and whole cell potency seems to depend on the fine-tuned lipophilicity/hydrophilicity ratio that governs the permeability of those compounds through the bacterial membranes. Lipophilicity of NBTIs is apparently optimal for passing through the membrane of Gram-positive bacteria, but the higher, although not excessive lipophilicity and suitable hydrophilicity seems to determine the passage through Gram-negative bacterial membranes. However, due to the considerable hERG inhibition, which is still at least two orders of magnitude away from MICs, continued optimization is required to realize their full potential.

Full text of DOI:10.3390/antibiotics10070862

antibiotics
Article
A Fine-Tuned Lipophilicity/Hydrophilicity Ratio Governs
Antibacterial Potency and Selectivity of Bifurcated Halogen
Bond-Forming NBTIs
1,2,†
Anja Kolaricˇ
, Maja Kokot 1,2,†, Martina Hrast 2 , Matjaž Weiss 2 , Irena Zdovc 3 , Jurij Trontelj 4,
Simon Žakelj 4 , Marko Anderluh 2,
*
and Nikola Minovski 1,
*
1
Laboratory for Cheminformatics, Theory Department, National Institute of Chemistry, Hajdrihova 19,
SI-1001 Ljubljana, Slovenia; anja.kolaric2@um.si (A.K.); maja.kokot@ki.si (M.K.)
Pharmaceutical Chemistry, Faculty of Pharmacy, University of Ljubljana, Aškercˇeva cesta 7,
SI-1000 Ljubljana, Slovenia; martina.hrast@ffa.uni-lj.si (M.H.); matjaz.weiss@ffa.uni-lj.si (M.W.)
Institute of Microbiology and Parasitology, Veterinary Faculty, University of Ljubljana, Gerbicˇeva 60,
SI-1000 Ljubljana, Slovenia; irena.zdovc@vf.uni-lj.si
Biopharmaceutics and Pharmacokinetics, Faculty of Pharmacy, University of Ljubljana, Aškercˇeva cesta 7,
SI-1000 Ljubljana, Slovenia; jurij.trontelj@ffa.uni-lj.si (J.T.); simon.zakelj@ffa.uni-lj.si (S.Ž.)
Correspondence: marko.anderluh@ffa.uni-lj.si (M.A.); nikola.minovski@ki.si (N.M.)
These authors contributed equally to this work.
2
3
4
*
†
Abstract: Herein, we report the design of a focused library of novel bacterial topoisomerase inhibitors
(NBTIs) based on innovative mainly monocyclic right-hand side fragments active against DNA gyrase
and Topo IV. They exhibit a very potent and wide range of antibacterial activity, even against some of
the most concerning hard-to-treat pathogens for which new antibacterials are urgently needed, as
reported by the WHO and CDC. NBTIs enzyme activity and whole cell potency seems to depend on
the fine-tuned lipophilicity/hydrophilicity ratio that governs the permeability of those compounds
through the bacterial membranes. Lipophilicity of NBTIs is apparently optimal for passing through
the membrane of Gram-positive bacteria, but the higher, although not excessive lipophilicity and
suitable hydrophilicity seems to determine the passage through Gram-negative bacterial membranes.
However, due to the considerable hERG inhibition, which is still at least two orders of magnitude
away from MICs, continued optimization is required to realize their full potential.
ꢀꢁꢂꢀꢃꢄꢅꢆꢇ
ꢀꢁꢂꢃꢄꢅꢆ
Citation: Kolaricˇ, A.; Kokot, M.;
Hrast, M.; Weiss, M.; Zdovc, I.;
Trontelj, J.; Žakelj, S.; Anderluh, M.;
Minovski, N. A Fine-Tuned
Lipophilicity/Hydrophilicity Ratio
Governs Antibacterial Potency and
Selectivity of Bifurcated Halogen
Bond-Forming NBTIs. Antibiotics
2021, 10, 862. https://doi.org/
10.3390/antibiotics10070862
Keywords: antibacterials; drug discovery; DNA gyrase inhibitors; topoisomerase IV inhibitors;
Academic Editor: Carlos M. Franco
intercalators; novel bacterial topoisomerase inhibitors
Received: 2 June 2021
Accepted: 12 July 2021
Published: 15 July 2021
1. Introduction
The loss of effective antibiotics is undermining our ability to fight bacterial infectious
diseases and managing the infectious complications that were considered treatable not
long ago. Life-threatening multi-drug resistant bacterial infections are notably common in
hospital environments, where patients with compromised immune systems are significantly
exposed and portray a particularly vulnerable risk group. In 2008 ESKAPE (Enterococcus
faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
aeruginosa, and Enterobacter species) pathogens [1] were introduced as extremely concerning
bugs in hospitals due to their escalating resistance to many effective antimicrobial agents.
Similarly, in 2013 Centers for Disease Control and Prevention (CDC) presented antibiotic
resistance threats report, dividing them in three categories, urgent, serious, and concerning
Copyright:
© 2021 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
based on the antibiotic resistance threats in the United States [2]. In 2017 the World Health
Organization (WHO) announced, for the first time, the antibiotic resistance priority list,
which became a global R&D guidance explicitly focused on specific pathogens; critical, high,
and medium priority pathogen categories were defined, ranking them according to the
urgency of need for new antibacterials [3]. “Critical” pathogens group include A. baumannii,
Antibiotics 2021, 10, 862. https://doi.org/10.3390/antibiotics10070862
https://www.mdpi.com/journal/antibiotics

Antibiotics 2021, 10, 862
2 of 30
P. aeruginosa, and Enterobacteriaceae species (e.g., Escherichia coli and K. pneumoniae), which
are posing particular threat among hospitalized patients with mostly fatal infections such
as sepsis and pneumonia. Several other bacterial strains like methicillin-resistant S. aureus
(MRSA) have been included in the “High” priority list, but they pose no less threat to
human health than bacterial strains from the “Critical” list. The only difference is probably
few more years that we have to find agents that are effective against the “High” priority
group, while for the “Critical” group alarm clock is already ringing—some strains do
not respond to any known therapeutic antibacterial agent [4]. Similar conclusions were
also drawn for the latest list of prioritized bacterial pathogens released in 2019 by CDC,
updating urgent, serious, and concerning bacteria that require attention now, because
the antibiotic resistance is still spreading [5]. Accordingly, a global concern was raised,
emphasizing the necessity for new classes of antibacterials.
In response to this global campaign, novel bacterial topoisomerase inhibitors (NBTIs)
have become a highly promising class of new antibacterial agents active against well-
validated antibacterial targets, the DNA gyrase and its paralogous form topoisomerase
IV (Topo IV) [
they prevent native enzyme function that consequently ends with bacterial cell death [
Contrary to the well-established quinolones, NBTIs bind to an alternative binding site
thereby avoiding the cross-resistance [ ]. They are comprised of three major parts; (1) a left-
6–
10]. By forming a ternary complex with the bacterial enzyme and DNA
9].
9
hand side (LHS) that intercalates between DNA base pairs and stabilizes only single-strand
cleavage breaks, which is connected through (2) a linker to (3) the right-hand side (RHS)
fragment that binds in a pocket midway between both gyrase A (GyrA) subunits [10].
The secondary amine on linker’s chain was identified as an essential element for activ-
ity making a contact with Asp83 at the entrance of GyrA hydrophobic binding pocket
(Figure 1) [9–13]. The formation of complexes with single-strand DNA breaks has been
recognized as a mechanism of inhibition of the enzyme function. We recently solved a 2.3-Å
crystal structure of compound 13 in complex with S. aureus DNA gyrase and DNA (PDB
ID: 6Z1A, Figure 1) [13]. It demonstrates for the first time a co-crystallized inhibitor in a
single orientation that does not suffer from static disorder as all previously reported gyrase
crystal structures complexed with NBTIs [14]. Due to crystallization of the inhibitor in a
single orientation, we were able to relate the compound orientation with the conformation
of the catalytic pocket, which allowed us to demonstrate the course of stabilization mecha-
nism. The intercalation of the LHS fragment between DNA base pairs induce a specific
spatial rotation that is mandatory for accomplishing optimal stacking interactions. In a
concomitant fashion, this rotation affects ribose and phosphates, as well that brings scissile
phosphate in one catalytic pocket closer to the catalytic metal and enables catalytic activity,
while in the second catalytic pocket scissile phosphate moves away from the catalytic metal
and deactivates any catalytic activity. Consequently, DNA breaks can only occur in one
strand [13,14].
Notwithstanding the excellent antibacterial properties of existing NBTIs published in
the literature, a major obstacle for their progression into clinical trials and more importantly
appearance on the market lies in their hERG toxicity that is near to potential therapeutic
concentrations. The endeavors to minimize this hERG-related cardiotoxic potential have
been particularly invested through modification of NBTIs physicochemical properties, such
as basicity and lipophilicity. Such approaches led to noticeable hERG improvements, but
unfortunately in many cases at the expense of lowering their antibacterial activity [15].
However, some recent reports have shown that it is possible to improve hERG cardiotoxicity
and maintain good antibacterial activity despite reduced basicity [16,17]. This prompted us
to focus on optimizing NBTIs in increasing their antibacterial potency across the bacterial
spectrum to provide acceptable therapeutic window even with observed hERG toxicity.
Namely, increasing antibacterial potency can be achieved either by increasing potency on
isolated enzyme or by balanced activity on two related enzymes (e.g., DNA gyrase and
Topo IV). By utilizing structural and physicochemical property optimizations, we have
managed to obtain among the best-in-class NBTIs remarkably active against a panel of

Antibiotics 2021, 10, 862
3 of 30
Gram-positive and some Gram-negative bacterial pathogens from the WHOs “Critical”
and “High” priority lists with a high potential for further progress.
Figure 1. Binding mode representation of an NBTI ligand (inset, grey, compound 13) in a complex
with S. aureus DNA gyrase and DNA molecule (PDB ID: 6Z1A) [13]. The GyrA subunit is shown
in green, the gyrase B (GyrB) subunit is red, while the DNA molecule is orange. Ionic interaction is
shown as cyan dots, while the bifurcated symmetrical halogen bonds are represented as yellow dots.
2. Results and Discussion
2.1. NBTIs Design Strategy
The strategy for designing optimized NBTI compounds was substantiated on our pre-
vious study [18], where we initially synthesized a tiny series of five NBTIs with innovative
RHS moieties active against S. aureus DNA gyrase [18,19]. Due to their inadequate solubil-
ity, excessive rigidity, and instability (compounds were disintegrating), these compounds
failed to outperform reported NBTIs in terms of both enzyme inhibition and antibacterial
potency [18]. In the present study, the first optimization step was aimed at increasing
overall compounds solubility by replacing previously used methoxy-quinoline LHS with a
more optimal, naphthyridine LHS as well as the cyclohexyl-amide linker with aminopiperi-
dine linker (Figure 2); these structural alterations introduce a higher flexibility and ensure
a proper spatial arrangement of the LHS and RHS, respectively [9]. By retaining identical
RHS fragments from our previous study [18], compounds were synthesized (Table 1).
3–7
Replacement of LHS-linker construct contributed in increasing the overall polarity (i.e.,
lowered logD, from 2.59 to 1.75, Figure 2) as well as the basicity (an additional basic center)
indicating an improved compounds solubility and unexpectedly increased stability. To
our surprise, compound
5 showed 2.5-fold improvements in S. aureus gyrase IC50 and
higher antibacterial potency compared to its predecessor. Since, adjusting physicochemical
properties (i.e., lowering logD and increasing pKa) was recognized as an appropriate step
toward enhancing NBTIs permeation across membranes of Gram-negative bacteria [20], we
envisaged it as an appropriate design feature for improving the whole-cell activity, as well.
The solubility of all studied compounds (Supplementary Materials Table S1) was
evaluated by determining the maximum single-dose, which could still dissolve in 500 mL
of the fluid assumed to be available for dissolution in the intestinal tract and thereby
presenting a limit for drug development [21] (Supplementary Materials Table S1). The
most restrictive standard pH (6.8) for basic compounds was chosen for this evaluation.
Especially the compounds
single-dose, while compounds
9
,
10, and 16 appear to be soluble at this pH in any conceivable
13 15 19, and 20 would require either dosing below
8
,
– ,
100 mg or formulation aided solubilization. It is also worth noting that enhanced solubility
could also be expected in the more acidic conditions of the stomach, which combined with
the high permeability is still favorable for absorption in the upper small intestine. The
PAMPA permeability of all tested compounds except 7, 16, and 17 is comparable to that

Antibiotics 2021, 10, 862
4 of 30
of highly permeable reference drugs and clearly higher than that of poorly permeable
reference drugs [22].
Figure 2. The impact of lipophilicity (logD) and structural flexibility on the change in IC50 for a com-
pound containing methoxy-quinoline cyclohexyl-amide LHS-linker combination against S. aureus
DNA gyrase from our previous study (left) [18] and compound
5 (right) comprising naphthyri-
dine aminopiperidine LHS-linker construct confirming the rationally optimized physicochemical
properties and reduced rigidity (amide vs. ethylene) leading to the improved potency and antibacte-
rial activity.
Table 1. DNA supercoiling inhibition of S. aureus and E. coli DNA gyrase and topoisomerase IV and decatenation activity of
human topoisomerase IIα affected by NBTIs.
DNA Gyrase IC50 (µM) 1
Topoisomerase IV IC50 (µM) 1
Human Topo II
(% RA) 2
Cmpd
Structure
S. aureus
E. coli
S. aureus
E. coli
3
1.60 ± 0.35
>100
>100
1.18 ± 0.09
94.71 ± 4.82
4
5
0.51 ± 0.25
0.33 ± 0.08
13.27 ± 2.84
3.00 ± 0.55
>100
0.12 ± 0.01
88.41 ± 0.15
82.10 ± 2.52
29.59 ± 0.54
0.085 ± 0.002
6
7
>100
>100
>100
>100
>100
>100
>100
>100
99.36 ± 9.93
90.35 ± 7.56

Antibiotics 2021, 10, 862
5 of 30
Table 1. Cont.
DNA Gyrase IC50 (µM) 1
Topoisomerase IV IC50 (µM) 1
Human Topo II
Cmpd
8
Structure
(% RA) 2
S. aureus
E. coli
S. aureus
E. coli
0.24 ± 0.04
7.83 ± 0.55
>100
0.21 ± 0.00
94.17 ± 2.90
9
0.16 ± 0.01
1.10 ± 0.17
1.02 ± 0.02
0.55 ± 0.06
2.95 ± 0.56
16.70 ± 0.93
40.60 ± 1.97
16.95 ± 1.69
40.38 ± 0.09
>100
0.23 ± 0.00
0.28 ± 0.01
2.04 ± 0.20
0.28 ± 0.03
101.25 ± 6.82
95.69 ± 6.42
96.97 ± 9.58
100.81 ± 11.22
10
11
12
>100
>100
13
14
0.035 ± 0.01
1.71 ± 0.05
0.57 ± 0.06
16.37 ± 0.56
7.22 ± 0.55
0.031 ± 0.00
0.042 ± 0.003
101.98 ± 0.84
98.25 ± 2.72
0.007 ± 0.00 3
0.011 ± 0.003 3
4.39 ± 0.55
15
16
0.28 ± 0.02
2.13 ± 0.11
0.021 ± 0.001
1.21 ± 0.05
88.26 ± 3.33
97.73 ± 1.08
56.56 ± 3.32
>100
17
1.82 ± 0.11
>100
>100
1.15 ± 0.03
97.80 ± 2.21

Antibiotics 2021, 10, 862
6 of 30
Table 1. Cont.
DNA Gyrase IC50 (µM) 1
Topoisomerase IV IC50 (µM) 1
Human Topo II
Cmpd
18
Structure
(% RA) 2
S. aureus
E. coli
S. aureus
E. coli
0.067 ± 0.01 11.89 ± 0.83
>100
>100
94.19 ± 1.33
19
20
0.35 ± 0.03
3.90 ± 0.06
29.04 ± 1.89
0.10 ± 0.01
96.36 ± 2.30
0.33 ± 0.02
0.16 ± 0.02
4.41 ± 0.16
6.91 ± 1.62
>100
0.06 ± 0.00
94.80 ± 0
21
22
18.45 ± 1.32
>100
95.49 ± 2.39
0.34 ± 0.05
0.37 ± 0.02
7.24 ± 0.34
0.24 ± 0.04
>100
>100
97.05 ± 6.94
Gepotidacin
8.30 ± 0.37
0.05 ± 0.00
ND
1
2
IC50 value represents the mean value of two to four independent measurements. The mean
±
SD percentage of residual activity of the
3
enzyme at compound concentration 10
assays, as detailed in [13].
µM obtained by two independent measurements. IC50s were determined by using gel-based
Once proper physicochemical properties of LHS-linker fragment were achieved (Sup-
plementary Materials Table S1), we focused on increasing target affinity (GyrA only at
the beginning) and consequently antibacterial potency. Considering that the structure–
activity relationship (SAR) of such LHS-linker fragment is relatively well established [10],
we recognized a major potential for antibacterial activity improvement in amplifying
their binding to DNA gyrase through applying various RHS alterations. While LHS is
responsible for DNA intercalation, RHS rules GyrA binding, and consequently influences
the overall NBTIs binding affinity and inhibitory potency. After a careful review of the
most prosperous NBTIs from the literature [
moieties, we observed that the majority of NBTIs contain rigid bicyclic RHSs. In contrast to
them, our compound comprises a more flexible 1-phenyl-1H-pyrazole RHS moiety, which
was identified as the most active one (S. aureus DNA gyrase IC50 = 0.33 M in Figure 2)
among compounds (Table 1); consequently, we directed our improvement strategy
9,10,12,23], with a special focus on their RHS
5
µ
3–7

Antibiotics 2021, 10, 862
7 of 30
toward introducing more flexible RHSs. By comparing the length of RHS fragments in
terms of measuring the interatomic distances between linkers’ secondary amine and the
geometrically most distant RHS heavy atom (Supplementary Materials Table S2), this
distance was found to be more than 8 Å for our flexible RHSs (compounds 3–8), comparing
to ~7 Å or less for known bicyclic RHSs. Our further optimization goal was thus directed
to slight shortening the RHS length that was achieved by introducing monocyclic moieties
preferably substituted on para-position with aliphatic groups (compounds 9–22), thereby
maintaining some degree of flexibility. Such RHSs would enable better adaptation into
gyrase binding pocket (Figure 3, Supplementary Materials Figures S1 and S2). The majority
of known RHS fragments form mainly van der Waals interactions with the RHS-binding
pocket delineated with hydrophobic amino acid residues (i.e., Ala68, Gly72, Met75, Met121
for S. aureus gyrase). On the contrary, our idea for strengthening NBTIs binding and conse-
quently improving their overall antibacterial activity was substantiated on targeting the
backbone carbonyl oxygens of Ala68 residues from both GyrA subunits with p-substituted
RHSs. A similar approach was already previously used in achieving excellent enzyme
inhibitory activity in the discovery of GSK299423, an NBTI capable to establish unusual hy-
drogen bonds between the methylene group of the oxathiolopyridine RHS and both Ala68
backbone carbonyl oxygens [9,13]. Apart from important interactions, the added benefit of
targeting backbone carbonyls lies in their presumably lower susceptibility to mutations.
Although this cannot be stated without a reasonable doubt, very few mutations, if any, may
result in backbone cabonyls’ eradication, and few may mask their availability due to steric
hindrance, making them less likely to contribute to developing NBTIs acquired resistance.
Therefore, monocyclic p-phenyl RHSs were substituted with various groups (compounds
12–17) that would strengthen the binding by hydrogen (-CH2OH, -CONH2) or halogen
(F, Cl, Br, I) bonding with the backbone carbonyl oxygen of at least one of Ala68 or other
non-bonded interactions. We have predicted the possibility of such hydrogen/halogen
bonds formation by flexible molecular docking calculations within NBTIs gyrase binding
site of S. aureus (PDB ID: 2XCS) [9]. The design of compounds and prediction of their
ability to form such strong interactions was described in detail in our recent study [13].
Since, NBTIs act on both Gram-positive and Gram-negative bacteria, the ability of such
hydrogen/halogen bond interactions was predicted by flexible molecular docking calcu-
lations within NBTIs binding pocket of a constructed E. coli gyrase homology model, as
well [18]. Furthermore, Topo IV is a secondary [11] and equivalently important NBTI target
to DNA gyrase, therefore a balanced inhibition of both targets would be advantageous in
achieving more potent antibacterial activity and broad spectrum. Since, the enzymes do
not act synergistically [24], a balanced activity would also be an advantage in preventing
development of bacterial resistance through maintaining activity in one enzyme target in
the event of mutations in another. To predict how the compounds would bind to Topo IV
and which interactions they may form (especially interactions with the backbone carbonyl
oxygen of corresponding Ala64), we also performed flexible molecular docking calculations
within NBTI binding site of our S. aureus and E. coli Topo IV assembled homology models,
respectively. The predicted binding mode of p-halogenated compounds (Supplementary
Materials Figures S1 and S2) is exemplified by compound 14 (Figure 3). Taking into account
the defined X···O bond lengths (Cl···O < 3.27 Å, Br···O < 3.37 Å, I···O < 3.50 Å) and angles
(140◦ ≤ Θ1 [C-X···O]
≤
180◦ and 90◦ ≤ Θ2 [X···O = C] ~ 120◦) [25], a high probability of
halogen bond formation with the backbone carbonyl oxygen of Ala68 residue from one
GyrA subunit of S. aureus was noticed (Figure 3a, Supplementary Materials Figure S1a,c,e).
Likewise, halogen-bonding was predicted for S. aureus and E. coli Topo IV with the back-
bone carbonyl oxygen of the corresponding Ala64 from one ParC subunit (Figure 3b,c and
Supplementary Materials Figure S2). Surprisingly, no halogen or any other interactions
with the corresponding alanine (Ala67) was noticed in the case of E. coli DNA gyrase
(Supplementary Materials Figure S1b,d,f), and no hydrogen bonds were anticipated for
any of the investigated NBTIs.

Antibiotics 2021, 10, 862
8 of 30
Figure 3. Halogen-bonding interactions for compound 14 as predicted by flexible molecular docking calculations within the
RHS binding site of: ( ) S. aureus DNA gyrase (PDB ID: 2XCS) [ ], ( ) S. aureus Topo IV homology model, and ( ) E. coli
a
9
b
c
Topo IV homology model. Enzymes are shown in green (DNA gyrase) and blue (Topo IV) ribbons, respectively, compound
14 is colored by element (ball and sticks representation), while DNA in pale orange. Halogen-bonding interactions are
represented as yellow dots.
2.2. Synthesis
The free amine
pound (DSK-1030, DSK InnoSciences Pvt. Ltd., Hosur, Tamil Nadu, India) with 1M HCl
in acetic acid. All final compounds 22 were obtained by one-step reductive amination
using various aldehydes and NaCNBH3 as a reducing agent, as shown on the Scheme 1.
2 was provided by Boc deprotection of commercially available com-
1
3–
Scheme 1. Synthesis of final compounds
30 min, rt; (ii) RCHO, MeOH, CH3COOH, 2 h, then NaCNBH3, 0 C, 10 min, then rt.
3
–
22. Reagents and conditions: (i) 1M HCl in CH3COOH,
◦
2.3. In Vitro Measurement of Inhibitor Potency and Antibacterial Activity
The enzyme inhibition profile of optimized NBTI compounds was assessed by in vitro
enzyme inhibition assay on isolated bacterial targets, DNA gyrase and Topo IV, originating
from wild-type S. aureus and E. coli, respectively. In vitro antibacterial susceptibility ex-
pressed as minimal inhibitory concentration (MIC) values was determined against a panel
of Gram-positive (S. aureus, methicillin-resistant S. aureus, Streptococcus agalactiae, Entero-
coccus faecalis, and Streptococcus pneumoniae), Gram-negative (E. coli, Salmonella alachua, P.
aeruginosa, A. baumannii, and Campylobacter jejuni), and Mycobacterium strains (Mycobac-
terium smegmatis and Mycobacterium avium), enriched by various MRSA and E. coli strains,
resistant to several antibiotics as well as E. coli strains with increased membrane permeabil-
ity and lack of efflux pump, as well. Moreover, selectivity of the compounds for bacterial
enzymes over orthologous human topoisomerase IIα was also assessed.
As displayed in Table 1, the compounds exhibit stronger enzymatic inhibition (lower
IC50s) against S. aureus DNA gyrase and E. coli Topo IV compared to E. coli DNA gyrase and
S. aureus Topo IV. Compounds 14 and 15 with the most potent IC50 values on all enzymes
come closest to the balanced inhibition; however a difference of 1–2 orders of magnitude
in enzyme inhibitory activity (of the same bacteria) is still too high to provide desired
balanced DNA gyrase/Topo IV inhibition profile. All compounds expressed selectivity
for bacterial enzymes over the orthologous human topoisomerase IIα enzyme, with 15
achieving approximately 1000-fold higher selectivity for bacterial enzyme.
Antibacterial activity is presented in Table 2. The potencies against all MRSA strains
were rather comparable for each specific compound. Compounds with bicyclic RHS

Antibiotics 2021, 10, 862
9 of 30
moiety, such as
6
and
7
were inactive against all strains, while compound
3
showed mod-
erate inhibition and antibacterial potency of S. aureus and other Gram-positive strains
but was generally poorly active against Gram-negatives. Compounds and compris-
ing 1-phenylpyrazole RHS inhibited S. aureus DNA gyrase and E. coli Topo IV enzyme,
with higher potency on E. coli Topo IV (0.33 M and 0.09 M, respectively). The overall
with unsubstituted phenyl RHS was found to be more
; therefore, it represented a starting point for further
4
5
µ
µ
antibacterial profile of compound
favorable than its fluoro analog
5
4
optimization. The same trends were also observed in our previous study for compounds
containing the same RHS fragments in combination with methoxy-quinoline cyclohexyl-
amide LHS-linker fragment [18]. The addition of another nitrogen into pyrazole ring (
resulted in slightly decreased potency against Gram-negative strains. According to our
design hypothesis, shorter and more flexible RHS fragments, such as isopropyl ( ) and allyl
10) chains were introduced into pyrazole ring. The shorter isopropyl group ( ) improved
S. aureus gyrase IC50 (0.16 M) compared to (0.33 M), while allyl (10) group reduced
8)
9
(
9
µ
5
µ
gyrase inhibition and antibacterial activity against all strains, even 8-fold for S. aureus,
E. faecalis, and A. baumannii. Both RHS fragments decreased Topo IV enzyme activity.
As a further optimization step, the substituted heteroaromatic pyrazole was replaced
with phenyl to optimize the ligand efficiency (LE, Table 3) for their targets, S. aureus and
E. coli DNA gyrase/Topo IV. Unsubstituted phenyl (11) showed considerable reduction
of enzymatic inhibition as well as lack of antibacterial potency against all strains (up to
32-fold) compared to compounds with pyrazole RHS. However, a notable optimization
was achieved by the introduction of p-halogenated phenyl RHSs (12–15), resulting in a
significant improvement of DNA gyrase inhibition. This offered an initial confirmation
that the predicted halogen bonds were most probably established and was in accordance
with our intentions to obtain acceptable value of LE > ~0.3 [26]. Substantially lower LE
values were observed for compounds 12–15 for E. coli DNA gyrase and S. aureus Topo
IV in comparison to S. aureus DNA gyrase and E. coli Topo IV, which correlate with
observed enzymes activities. Compound 12 with p-fluorophenyl RHS enhanced inhibitory
activity and potency comparable to
4, which contains the same fragment. p-Chlorophenyl
substitution (13) led to remarkable inhibition improvement of all enzymes, which was
evident in overall antibacterial spectrum. Furthermore, introduction of a bromine (14
)
and an iodine (15) showed an exceptional enzymatic inhibitory performances with IC50
of 0.007 µM and 0.011 µM for S. aureus DNA gyrase, which resulted in highest LE for all
enzyme targets among the compounds. Considering that on-target activity increases with
increasing the size of halogen, thereby also increasing lipophilicity (calculated values of
clogP and logD are available in the Supplementary Materials, Table S1), which challenges
our explanation that bifurcated halogen-bonding guide the enzyme inhibitory activity of
presented NBTIs. Although, increasing halogen size certainly increases van der Waals
forces, the highest difference in IC50 values in the series 11–15 is between p-fluoro (12)
and p-chloro (13) derivatives on all enzyme targets, while modest increase in potency is
seen for p-bromo (14) and p-iodo (15) derivatives. p-Fluoro substituent does not form
halogen bonds, while p-chloro, p-bromo, and p-iodo does, which offers a reasonable
confirmation that halogen-bonding is indeed taking place and contribute to the inhibitory
potency. Nevertheless, enhancing the lipophilicity of compounds 11–15 increases their cell
permeability (see results of PAMPA permeability in Supplementary Materials, Table S1),
which most likely results in stronger whole cell activity.
A comparably low nanomolar inhibitory activity has been detected only for GSK299423
(IC50 = 0.014
±
5 µM) [9], however such an outstanding MIC of 0.0078 µg/mL against vari-
ous S. aureus strains for 15 is comparable or even better comparing to the reported NBTIs.
Although, both 14 and 15 showed comparable enzymatic inhibition, which is probably due
to steric hindrance effect of the iodine atom, the difference in antibacterial potency against
all strains is not negligible. 15 shows generally 4-fold higher overall potency than 14 and
remarkable MRSA and S. pneumoniae impact (MIC = 0.0078
µg/mL and 0.015 µg/mL,
respectively), so it was identified as the most active and promising compound among

Antibiotics 2021, 10, 862
10 of 30
the series. Reasonably potent antibacterial activity against other Gram-negative strains,
A. baumannii (MIC = 0.25
against P. aeruginosa (MIC = 32
an excellent M. smegmatis and moderate M. avium antibacterial activity with MIC values
of 0.015 g/mL and 1 g/mL, respectively. This trend in increase of overall antibacterial
µg/mL), C. jejuni (MIC = 0.015
µg/mL), and a moderate activity
µg/mL) was detected, as well. Moreover, 15 also exhibited
µ
µ
performance with halogen size (H < F < Cl < Br < I) is as predicted by molecular docking
calculations due to halogen bond formation, which was confirmed in our recently solved
crystal structure of p-chloro compound 13 in a complex with S. aureus DNA gyrase and
DNA (PDB ID: 6Z1A, Figure 1; Figure 4) [13] in which p-chlorophenyl RHS moiety is bound
in the hydrophobic pocket formed by both GyrA subunits, delineated with amino acid
residues Ala68, Gly72, Met75, and Met121. The most interesting feature related to our
crystal structure is the formation of not a single one, but symmetrical bifurcated halogen
bond between p-chloro and the backbone carbonyls of Ala68 residues from both GyrA
subunits. Such halogen-bonding has not been previously identified in a biological system
between a ligand and its macromolecular target and, moreover, it also contradicts the
classical interpretation for angles and distance of halogen bonds. Detailed description of
the crystal structure was described in our previous work [13].
Important conclusions can be drawn when comparing potency of 15 with Gepotidacin,
an NBTI in phase III clinical trials for a potential first-in-class antibiotic. 15 outperforms
Gepotidacin in isolated enzyme inhibition assay on virtually all four enzymes tested
(Table 1). The more potent on-target activity results in more than two orders of magnitude
more potent antibacterial activity on S. aureus (WT) while retaining the potency on E. coli
(Table 2). This proves that we were successful in increasing NBTI potency, at least for the
Gram-positive bacteria.
Figure 4. Symmetrical bifurcated halogen-bonding of compound 13 in complex with S. aureus DNA
gyrase and DNA molecule (PDB ID: 6Z1A) [13]. GyrA subunit is shown in green, compound 13
is colored in grey (ball and sticks representation) and DNA in pale orange. Halogen bonds are
represented as yellow dots.

Antibiotics 2021, 10, 862
11 of 30
Table 2. Antimicrobial susceptibility of NBTIs against a panel of Gram-positive, Gram-negative, and Mycobacterium bacterial strains.
Compd
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Gepotidacin
MIC [µg/mL]
S. aureus (ATCC 29213)
MRSA
1
2
8
2
4
4
4
2
2
2
2
4
8
8
8
4
0.25
0.5
0.5
1
1
1
>128 >128
>128 >128
>128 >128
>128 >128
>128 >128
>128 >128
>128 >128
0.5
0.5
0.5
0.5
1
0.5
1
0.25
0.125
0.25
1
1
1
1
1
0.5
2
2
4
4
4
4
8
1
4
8
8
8
8
8
16
2
1
2
4
4
4
4
8
2
0.125
0.25
0.5
0.25
0.5
0.25
1
0.5
0.031
0.062
0.0625
0.0625
0.125
0.125
1
0.0078
0.015
4
4
4
4
8
8
8
2
2
4
4
4
8
8
128
16
0.125
0.25
0.5
0.25
0.5
0.5
1
0.25
1
2
2
2
2
2
16
2
0.5
1
0.5
1
2
1
0.25
0.5
1
0.5
1
0.5
2
1
0.5
1
1
2
2
2
2
1
0.125
ND
ND
ND
ND
ND
ND
ND
MRSA 1 (QA-11.7)
MRSA 2 (QA-11.6)
MRSA 3 (QA-11.4)
MRSA 4 (QA-11.2)
E. faecalis DRK 057 (ATCC 29212)
S. pneumoniae (QA-10/02)
S. agalactiae RDK 047
(QA-990/02)
0.0078
0.03125
0.0625
0.03125
1
1
0.5
4
1
>128
128
0.125
0.015
<1
2
1
128
>128
0.5
0.5
4
16
8
1
0.5
0.125
8
>128
0.25
<1
2
4
1
ND
E. coli (ATCC 25922)
>128
64
64
128
>128
16
32
16
16
8
16
2
16
2
2
2
1
>128 >128
>128 >128
>128 >128
64
32
4
32
32
1
8
8
2
4
1
16
16
4
8
8
64
64
32
32
16
8
32
16
8
8
16
4
4
2
1
2
2
2
1
1
2
0.25
1
128
128
32
32
16
>128
>128
128
128
128
16
32
8
4
8
8
32
4
2
4
16
1
8
4
2
1
2
32
8
2
8
4
16
8
2
4
4
1
E. coli 5 (QA-11.8)
ND
ND
ND
0.125
0.016
ND
ND
E. coli 6 (QA-11.7)
E. coli 7 (QA-11.3)
128
>128 >128
128 >128
>128 >128
>128
1
1
E. coli D228
0.125
0.125
1
0.125
0.078
0.25
32
E. coli N43 9 (CGSC# 5583)
A. baumannii 8C6 (GES-14)
P. aeruginosa RDK 184 (DSM 939)
S. alachua RDK 030c
(QA-1482/04)
1
4
0.25
2
1
32
0.5
4
16
128
1
8
0.5
4
1
16
1
4
>128
32
32
64
32
128
8
>128 >128 >128 >128 >128 >128 >128 >128 >128 >128 >128
128
>128 >128 >128 >128 >128 >128 >128
>128
128
64
>128 >128 >128
64
128
>128
128
32
16
8
>128 >128
128
128
64
128
64
ND
C. jejuni RDK 135 (ATCC 33560)
M. smegmatis
1
4
128
≤1
128
16
0.25
64
>128 >128
64 16
>128 >128
4
0.5
32
0.5
0.25
128
8
32
8
1
2
32
2
0.25
4
0.5
0.125
2
0.015
0.015
1
32
32
>128
128
16
128
4
4
128
2
4
128
0.015
0.5
128
128
0.25
4
2
16
>128
ND
ND
ND
≤1
M. avium
>128
>128 >128
1
2
QA-labeled isolates were obtained from quality assurance schemes sent by reference laboratories. Resistant to: cefoxitin, ciprofloxacin, clindamycin, erythromycin, tetracycline, thiamulin, trimethoprim.
3
4
5
Resistant to: cefoxitin, gentamicin, sulfamethoxazole, tetracycline. Resistant to: cefoxitin. Resistant to: cefoxitin, chloramphenicol, clindamycin, linezolid, tetracycline, thiamulin, trimethoprim. Resistant to:
ampicillin, cefotaxime, ceftazidime, ciprofloxacin, meropenem, nalidixic acid, sulfamethoxazole, tetracycline, trimeptoprim, cefepim, cefoxitn, ertapenem, imipenem. Resistant to: ampicillin, azithromycin,
cefotaxime, ceftazidime, sulfamethoxazole, trimeptoprim, cefepim, cefoxitin, ertapenem, imipenem. Resistant to: ampicillin, cefotaxime, ceftazidime, chloramphenicol, ciprofloxacin, colistin, gentamicin,
meropenem, nalidixic acid, sulfamethoxazole, tetracycline, trimeptoprim, cefepim, cefoxitin, imipenem. Bears a mutation in the lpxC gene that increases membrane permeability. AcrA knockout strain
(knockout of cell membrane efflux pump). ND: not determined.
6
7
8
9

Antibiotics 2021, 10, 862
12 of 30
Table 3. Calculated ligand efficiency (LE) values of compounds.
LE 1
DNA Gyrase
Cmpd
Topo IV
S. aureus
E. coli
S. aureus
E. coli
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
0.25
0.25
0.27
NA
NA
0.27
0.31
0.27
0.29
0.30
0.35
0.39
0.38
0.24
0.25
0.32
0.26
0.27
0.29
0.30
NA
0.20
0.23
NA
NA
0.21
0.25
0.22
0.21
0.23
0.27
0.29
0.31
0.19
NA
0.22
0.22
0.22
0.22
0.23
NA
NA
0.19
NA
NA
NA
0.20
NA
NA
NA
0.23
0.24
0.27
NA
NA
NA
0.18
NA
0.20
NA
0.25
0.28
0.29
NA
NA
0.28
0.30
0.30
0.28
0.31
0.35
0.35
0.36
0.27
0.26
NA
0.28
0.30
NA
NA
1
The LE was calculated according to the following equation: LE = (1.37/HA)
of non-hydrogen (heavy) atoms [26]. NA: not applicable.
×
pIC50, where HA is the number
Substitution of phenyl with p-hydroxymethyl (16) and amide group (17) led to a
significant decrease in activity, with the complete loss of E. coli gyrase and S. aureus Topo
IV on-target activity as well as loss of Gram-negative potency. According to the results for
target inhibition and whole cell activity, no hydrogen-bonding with gyrase or Topo IV was
established. Regarding the compounds which are not able to form hydrogen or halogen
bonds, it was interesting to note that compound 18 comprised of p-dimethylaminophenyl
RHS was identified as highly active against S. aureus gyrase (IC50 = 0.067 µM), with
moderate activity on E. coli gyrase, while no Topo IV activity was determined. According to
our docking calculations, each of Ala68 is forming a pseudohydrogen bond with one methyl
group, which might be the reason for high S. aureus gyrase activity [13]. Introduction of
di-substituted phenyl RHS as in 19 and 20 demonstrated modest and balanced antibacterial
activity on S. aureus gyrase and E. coli Topo IV, which was slightly better for 20 containing
free hydroxyl group. Particularly interesting was its antibacterial activity against C. jejuni,
which was comparable to the most efficient compound 15, although its antibacterial potency
on all other Gram-negative strains was rather moderate. Compounds 21 and 22 comprising
substituted pyridine showed gyrase inhibitory activity on both bacterial strains comparable
to 19 and 20, but unfortunately lost Topo IV inhibitory activity.
A relatively high correlation was observed between the gyrase pIC50 values and
corresponding pMIC of R2 = 0.864 and 0.856 for S. aureus and E. coli, respectively (Figure 5)
,
indicating that DNA gyrase is the pivotal target responsible for the observed whole cell
antibacterial activity in S. aureus strains. The correlation for Topo IV was slightly lower,
with R2 = 0.788 and 0.790 for E. coli and S. aureus, respectively. Although, the compounds
expressed potent enzyme inhibitory activity against E. coli Topo IV, their whole cell potency
on E. coli (expressed as pMIC) is somehow less proportional to Topo IV pIC50 (Figure 5).
The same ratio is observed for S. aureus pair of enzymes. This is a direct consequence of
DNA gyrase being the pivotal bacterial target for S. aureus and as such its inhibition governs
the antibacterial activity. The comparison on E. coli pair of enzymes is not that clear; even
though Topo IV is the primary target in isolated enzyme assay, lower R2 is observed. This
might be a consequence of the intrinsic difference in DNA gyrase and Topo IV importance

Antibiotics 2021, 10, 862
13 of 30
in bacterial cell survival, or even an outcome of the lower compounds’ accumulation in
E. coli due to efflux pumps, which does not follow the same SAR as the one for Topo IV. The
observed selectivity of NBTIs for two related targets might be explained by the differences
in amino acid residues constituting NBTIs’ binding pocket of DNA gyrase and Topo IV for
both strains (Supplementary Materials Figure S3). The main difference is the amino acid
residue Met75 in S. aureus DNA gyrase that corresponds to Ile74 in E. coli DNA gyrase,
Ile71 in S. aureus Topo IV and Leu71 in E. coli Topo IV, respectively. Ile residues (S. aureus
Topo IV, E. coli DNA gyrase) are expected to pose higher steric hindrance thus preventing
the binding of even smaller NBTI RHSs, while Met75 (S. aureus DNA gyrase) and Leu71
(E. coli Topo IV) residues allow just the right space to accommodate monocyclic RHSs [10].
Figure 5. The correlation between antibacterial potency (pMIC =
−
logMIC in molar) of NBTIs and the corresponding
) DNA gyrase and ( ) Topo IV, respectively. A
S. aureus and E. coli inhibitory activity (pIC50 logIC50 in molar) of: (
=
−
a
b
linear correlation may be observed for DNA gyrase of both bacterial strains (R2 = 0.864 and 0.856 for S. aureus and E. coli,
respectively), while slightly inferior linear correlation can be obtained for Topo IV (R2 = 0.788 and 0.790 for E. coli and
S. aureus).
Since the compounds exhibit potent enzyme inhibitory activity in general, their whole
cell potency depends particularly on the uptake into Gram-positive and Gram-negative
bacteria (apart from being substrates for the previously mentioned efflux pumps), which
depends on their physicochemical properties. In this context, one should be aware about
that cell envelopes of Gram-positive and Gram-negative pathogens differ substantially
in composition. This is particularly important in the case of Gram-negatives whose cell
envelope, in particular its outer layer (that is missing in Gram-positives) is extremely
protective for passage of highly lipophilic molecular entities due to tight packing of its
lipopolysaccharide (LPS) rich content [27]. Additionally, the aquaporins are an integral
part of the outer layer and restrict the diffusion of hydrophilic molecules with molecular
weight (MW) around 700 Da [28]. Put differently, more positively charged (i.e., increased
basicity) and suitably polar (i.e., lower logD) compounds with appropriate size (MW ~
400–500, as for the majority of NBTIs) tend to have better permeation in Gram-negative
bacterial pathogens [29]. Taking this into consideration, we further examined the relation-
ship between the compound’s lipophilicity and the enzyme inhibitory activity and the
whole cell potency by plotting the data on a 3D graph (Figure 6). The 3D plot distinguishes
separated clusters for each of the bacterial targets (S. aureus gyrase (green), Topo IV (blue)
and E. coli gyrase (red), Topo IV (yellow)) divided by the correlation between pIC50 and
logD as independent variables, and pMIC as a dependent variable. On one hand, the
increase in MICs is a reasonable consequence of the increase in the halogen size (12–15)
and aforementioned rise in halogen interactions potency. On other hand, the increase in
lipophilicity (logD) potentiates antibacterial activity (MIC), which is particularly notice-
able for compounds 13–15 and 20 having highest logD values (~1.9–2.2), compared to
compounds 16 and 17 with the lowest logD (0.44 and 0.19, respectively). These findings
are congruent with the previously determined, specific, low to moderate lipophilicity

Antibiotics 2021, 10, 862
14 of 30
range (logD ~ 0–3) that is optimal for passive cytoplasmic permeation, particularly in
Gram-positive bacteria [30,31]. Nevertheless, the data represented in Figure 6 point to
an important conclusion, i.e., a deft balance in NBTIs lipophilicity and on-target potency
has to be achieved for high antibacterial potency. On one hand, the secondary amine and
overall hydrophilicity has to be kept in order to achieve crucial interactions with DNA
gyrase and Topo IV as well as to increase the overall optimal ligands’ solubility (that is
decisive for passing the aquaporins rich outer layer of cell envelope in Gram-negative
bacteria). On the other hand, higher, but not excessive lipophilicity increases cellular per-
meability (Supplementary Materials Table S1) and thus antibacterial activity especially on
Gram-positive bacteria. Indeed, on-target activity is a prerequisite for antibacterial activity;
however, how lipophilicity/hydrophilicity influence the antibacterial activity could be best
observed by comparing DNA gyrase/Topo IV inhibition and antibacterial activity of com-
pounds 18–20. Although, 19 is distinguished by the most equilibrated inhibition of DNA
gyrase/Topo IV, both 18 and 20 surpass the latter in antibacterial activity. For 18 this might
be a consequence of more potent inhibition of S. aureus DNA gyrase. 20 has in general
more potent antibacterial activity on both Gram-positive and Gram-negative bacteria albeit
having very comparable on-target potency as 19. The reason for such a difference in activity
can be attributed to the fine-tuned lipophilic/hydrophilic properties of 20; these allow
sufficient compound solubility and cell permeability; however, due to slightly increased
hydrophilicity, 20 is probably a better substrate for Gram-negative aquaporins (for passing
the highly restrictive outer layer), i.e., optimally lipophilic (for passing not only LPS com-
partment of the outer layer, but also the inner, lipid rich cytoplasmic layer) [31,32] and/or a
worse substrate for the efflux pumps (see below). Further evidence on the influence of the
lipophilicity/hydrophilicity balance on the antibacterial activity may be seen by comparing
the most potent compound from our previous study [18] and compound
5 (both depicted
in Figure 2). Approximately 4-fold increase in potency on E. coli DNA gyrase resulted in
the same increase in antibacterial potency (evaluated by comparing MIC values). However,
although we observed approximately the same increase in the potency on isolated enzyme
from S. aureus, the antibacterial potency increased even more (8-fold). This increase in
antimicrobial potency is apparently not influenced solely by the on-target activity. Since
5
is more hydrophilic than the amidocyclohexane derivative, one would expect it to have
poorer permeability properties. Instead, the increase in antibacterial potency on S. aureus
of
efflux pumps. Increase in permeability is unlikely since the starting compound is slightly
more lipophilic, but it is likely that lower lipophilicity of makes it poorer substrate for
S. aureus efflux pumps [33 34]. A similar conclusion can also be reached when comparing
compounds activity on E. coli wild-type, and E. coli D22 and N43 strains. The increased
membrane permeability of E. coli D22 increases the antibacterial potency up to 16 (2 in
majority of cases), where the latter is improved the same or even more in E. coli N43 strain
with removed efflux pumps (32 in majority of cases). This led us to a conclusion that the
5 can only be explained by better permeability properties or the non-susceptibility to the
5
,
×
×
×
compounds have some difficulties in passing into the E. coli and are allegedly pumped-out
by the efflux pumps. As a consequence, the accumulation of the compounds in E. coli is
significantly reduced, and it is most likely the pivotal reason for their lower whole cell
activity. This is in line with the previous observation that the efflux pump containing AcrA
protein is known to bind preferably lipophilic compounds [28,35] and shows that delicate
balance, i.e., a “sweet spot” in compounds lipophilicity has to be achieved in order to
assure both permeability and the absence of efflux pumps susceptibility. This “sweet spot”
needs to be defined for each organism, and in S. aureus seems to be close to logD = 2.

Antibiotics 2021, 10, 862
15 of 30
Figure 6. 3D plot representing influence of the calculated logD (pH = 7.4) on pIC50 and pMIC against
all antibacterial targets. The graph clearly indicates that increasing lipophilicity increases enzymatic
activity (IC50) and antibacterial potency (pMIC).
2.4. Cytotoxicity Studies
In vitro safety profile was determined by cytotoxicity assessment on human umbilical
vein endothelial (HUVEC) and HepG2 liver cancer cell lines, and by the blockage of hERG
potassium ion channel. Cytotoxicity on human HUVEC and HepG2 cell lines was assessed
by compounds’ effect on cell metabolic activity (Table 4). Slightly higher toxicity was
observed on HepG2 cell line than on HUVEC. In general, cytotoxic IC50 values against
both cell lines were higher in comparison to corresponding S. aureus MIC values. The
most potent compounds 13
between S. aureus MIC and IC50s of both cell lines, with highest cytotoxicity on HepG2 cells
being approximately 10 M, while MIC down to 0.016 M was measured for compound
, 14, 15, and 18 showed up to 3 orders of magnitude difference
µ
µ
15 (Supplementary Materials Table S4). The majority of compounds exhibited higher E.
coli whole cell potency than cytotoxic IC50 values, especially in comparison to HepG2 cell
line. Comparing the data for E. coli N43 strain with removed efflux pumps, two orders
of magnitude higher potency was evident for compounds 14 and 15 on bacterial strains
compared to human HUVEC and HepG2 cell lines (0.16
µM for E. coli N43 vs. ~20 µM for
HUVEC and ~10 M for HepG2 of compound 15; Supplementary Materials Table S4). These
µ
results lead us to a reasonable conclusion that potency on bacterial strains does not overlap
with cytotoxicity and hence more potent compounds on bacteria display better safety
profile. For compounds
cell activity and not cytotoxic IC50 was determined (Supplementary Materials Table S5).
None of the compounds inhibited human topoisomerase II enzyme at concentration
10 M; however, the majority of them appeared active on HUVEC and HepG2 cells, from
which we can conclude that other mechanisms likely participate to the overall cell toxicity.
3, 6, 7, 16, and 17 that were poorly or non-active, only metabolic
α
µ

Antibiotics 2021, 10, 862
16 of 30
Table 4. Results of cytotoxicity on human HUVEC and HepG2 cell lines and hERG-related
cardiotoxicity.
HUVEC
HepG2
hERG
Cmpd
IC50 [µM]
IC50 [µM]
IC50 [µM]
3
4
5
6
>50
9.64 ± 0.03
10.70 ± 0.01
>50
Nd
1.04
0.33
0.62
ND
4.64 ± 0.10
6.33 ± 0.19
ND
7
8
9
ND
ND
ND
39.41 ± 5.47
17.08 ± 0.98
0.37
9.82
6.62
1.05
0.59
0.27
0.29
0.18
3.24
13.45
2.11
0.46
0.18
1.97
3.05
>50
>50
10
11
12
13
14
15
16
17
18
19
20
21
22
>50
>50
54.08 ± 3.45
17.67 ± 2.16
12.67± 0.21
10.16 ± 0.10
9.37 ± 0.07
8.47 ± 0.09
ND
>50
26.96 ± 0.38
25.42 ± 0.01
18.98 ± 0.85
>50
>50
ND
23.85 ± 0.09
19.83 ± 2.58
29.80 ± 1.13
>50
10.03 ± 0.06
6.31 ± 0.09
10.86 ± 0.12
>50
>50
32.03 ± 0.30
ND: not determined.
As has already been recognized for NBTIs reported in the literature, this series of
antibacterials suffers from undesired potassium channel inhibition or the hERG cardiac tox-
icity (Table 4). The hERG blockage induced by compounds 15 and 20 with IC50 of 0.18
is of particular concern, being only one order of magnitude above the target potency
MIC = 0.016 µM; Supplementary Materials Table S4) of the most favorable compound 15
µM
(
.
A general trend between the gyrase enzyme potency and cardiotoxicity accompanying
all NBTIs could be observed here as well; higher potency coincides with stronger hERG
inhibition [15]. We should note that 14 and 15, albeit provoking significant hERG inhibition,
have the ratio between hERG (IC50 = 0.18 µM for 15) and IC50/MIC (IC50 = 0.011 µM and
MIC = 0.016 µM for 15) spanning from 1-2 order of magnitude. This reveals that acceptable
therapeutic window might be achieved; however, further optimizations towards lower
hERG inhibition and better safety profile is still required.
3. Conclusions
In summary, herein we describe a rational design, synthesis, and biological charac-
terization of a series of NBTIs active against DNA gyrase and Topo IV, based on struc-
turally new, mainly monocyclic innovative RHS moieties. In general, these derivatives
display potent broad-spectrum antibacterial activity against a panel of clinically relevant
Gram-positive and Gram-negative bacterial pathogens, with significant effect also on My-
cobacterium species. In the course of our research, we have identified a subseries of highly
active compounds featuring p-halogenated phenyl RHSs with extremely potent whole cell
activity on S. aureus, due to symmetrical bifurcated halogen-bonding with the backbone
carbonyl oxygen of Ala68 from both GyrA subunits, i.e., the first halogen bond of this
type identified between a ligand and its target in a biological system. Bromo (14) and
iodo (15) derivatives displayed remarkable inhibitory activities against DNA gyrase from
S. aureus (IC50 of 0.007
of 0.042 M for 14 and 0.021
NBTIs reported to date. Furthermore, compound 15 showed excellent potency against
Gram-positive strains (MIC = 0.0078 g/mL and 0.015 g/mL for S. aureus and S. pneu-
moniae, respectively) and it was the most active compound against Gram-negative strains,
µM for 14 and 0.011
µM for 15, respectively) and E. coli Topo IV (IC50
µ
µM for 15, respectively), which are among the most potent
µ
µ

Antibiotics 2021, 10, 862
17 of 30
including E. coli (MIC ~ 0.25–2.0
µg/mL), A. baumannii (MIC = 0.25
µg/mL), C. jejuni (MIC
= 0.015 g/mL), and P. aeruginosa (MIC = 32 µg/mL) belonging to the list of most critical
µ
pathogens, as reported by the WHO and CDC. Moreover, it appears that DNA gyrase is
the pivotal antibacterial target for NBTIs activity in Gram-positive bacteria comparing
to Topo IV, as demonstrated by its higher pMIC-pIC50 correlation. In conjunction with
this findings, it seems that NBTIs enzyme activity and whole cell potency are directly
dependent on a specific, fine-tuned lipophilicity/hydrophilicity ratio for these compounds
that governs their permeation through the bacterial membranes, in particular the highly
restrictive outer barrier of Gram-negative bacteria. Indeed, lipophilicity itself (logD ~ 0–2.5)
is apparently optimal for passing of NBTIs through the bi-lipid compartment of Gram-
positive bacterial membranes, but once reached seems not to be a decisive factor. However,
the lower, not excessive lipophilicity as well as the ligand’s size are apparently decisive
physicochemical/molecular determinants for passing of these compounds through Gram-
negative bacterial membranes (in particular the outer layer’s LPS barrier and aquaporins)
and reaching their intracellular targets. We believe and strive to a further optimization
toward achieving much wider antibacterial spectrum and a more balanced Gram-positive
and Gram-negative potency. Due to relatively potent hERG inhibition, our future efforts
would be majorly invested in further improvements of the antibacterial activities as well as
reduction of hERG cardiotoxicity.
4. Materials and Methods
4.1. Biopharmaceutical Classification
Prior to the measurements of solubility and permeability isocratic HPLC methods
were developed. An Agilent 1100 system (degasser, bin. pump, well plate sampler, column
thermostat and diode array detector) was used with a Waters Xterra MS C18, 3.5 µm,
4.6 × 100 mm column and a mobile phase from A (0.5% ammonium phosphate buffer at
pH = 3.0) and B (acetonitrile). The individual method parameters for each compound are
given in the Supplementary Materials Table S1.
4.1.1. Solubility
The equilibrium solubility was determined by first weighing approximately 2 mg
of the tested or reference compound to a 1.5 mL microcentrifuge vial and then adding
0.5 mL of 0.1 mm glass beads and 200 µL of pH 6.8 phosphate buffered saline (PBS—1.0 g
potassium dihydrophosphate, 2.0 g dipotassium hydrophosphate and 8.5 g of sodium
chloride dissolved in 1.0 L of purified water). The vials were then placed in a homogenizer
(Next Advance Bullet Blender, Troy, NY, USA) for 5 min ad speed setting “2” after which
they were kept in a thermostated shaker (Tehtnica HeatMix, Domel, Slovenia) at 37 ◦C
and 1200 RPM for 24 h. The supernatants were obtained after a 10 min centrifugation at
15,000
immediately stabilized by a 12-fold dilution with the same buffer as used for the dissolution
process. Twenty L of these solutions were injected for the HPLC analysis by the described
methods (Supplementary Materials Table S1). External standards for calibration were made
in a mobile phase with 5% acetonitrile in a range of 30–90 mg/L and injected at 20
injection volumes. Less than 15% of the solid was dissolved during the incubation for
most compounds except for compounds 10, and 16 with 63%, 30%, and 20% of the solid
×
g (Eppendorf Centrifuge 5415R, Eppendorf AG, Hamburg, Germany) and were
µ
µL
9,
consumed, respectively. The method accuracy was confirmed by the coefficient of variation
which was 5.0%, 1.2%, 1.1%, and 2.0% for the reference compounds metoprolol, verapamil,
theophylline, and digoxin, respectively.
4.1.2. Permeability
The donor solutions for PAMPA permeability measurements were obtained by dis-
solving the remaining solids from the solubility measurements in 750 µL of pH 7.4 PBS
(2.38 g disodium hydrophosphate, 0.19 g potassium dihydrophosphate, and 8.0 g of sodium
chloride) with 20% of methanol in the same homogenizer and the same settings as de-

Antibiotics 2021, 10, 862
18 of 30
scribed above and then diluting the supernatants obtained under the same centrifugation
conditions two-fold with pH 7.4 PBS to yield 10% methanol as a donor solution medium. A
96-well precoated three-layer PAMPA plates (Corning® Gentest
™
Pre-coated PAMPA Plate
System, Corning, NY, USA) were used according to the manufacturer’s instructions and as
described previously [36] with 300
and 200
and a 4-h incubation time. The initial and end donor solutions were injected at a 5
injection volume. The initial donor solutions were also diluted 100-fold and injected at
a 100 L injection volume to serve as external references for the end acceptor solutions
µ
L of the donor solutions in the bottom compartment
◦
µ
L of the acceptor solution (pH 7.4 PBS) in the upper compartment at 37 C
µL
µ
which were injected at the same injection volume. The permeability calculations were made
with a manufacturer-provided Excel spreadsheet directly from the measured peak areas
accounting for the relevant dilutions and injection volumes. The concomitantly measured
standards for permeability classification metoprolol (high permeability), verapamil (high),
theophyllin (high), digoxin (intermediate), atenolol (low), and furosemide (low) were
prepared at 200 µM concentrations for the donor solutions in a PBS with 10% methanol
and thereafter treated the same as the tested compounds. Measurements were made in
quadruplicates with the coefficient of variation up to 26% for the highly permeable com-
pounds. The “high” permeability of the tested compounds was designated when it was
higher than that of theophylline (lowest permeability of a “high” permeability reference
drug and “intermediate” permeability was designated when it was above the value for
digoxin and below teophyllin.
4.2. Homology Modeling
The structural homology model for the enzyme Topo IV was built using the protein
structure homology modelling server Swiss model. First, the target sequence for both
subunits was downloaded from UniProt repository in FASTA format (protein ID for E. coli
parC: P0AFI2, parE: P20083 and for S. aureus parC: Q2FYS4, parE: Q2FYS5). Then the
Swiss model template library was searched for all possible templates. Suitable templates
were selected (PDB ID: 3KSA for E. coli and PDB ID: 3RAF for S. aureus), considering the
following values: coverage, GMQE score, identity and resolution (Supplementary Materials
Table S3). After building the model the correspondence between the homologous model
and the template was visually checked.
4.3. Molecular Docking Calculations
Molecular docking calculations of novel NBTI compounds within the NBTI binding
pocket of S. aureus and E. coli DNA gyrase and S. aureus and E. coli topo IV were performed
by GOLD docking suite in a flexible fashion [37]. Crystal structure of S. aureus DNA gyrase
in complex with DNA and an intercalated NBTI ligand (GSK299423; PDB ID: 2XCS) and
our constructed E. coli DNA gyrase [18], S. aureus topo IV and E. coli topo IV homology
models were used for calculations. The experimental coordinates of the co-crystallized
NBTI ligand (GSK299423) were used to define the binding site (cavity radius of 16 Å for
S. aureus DNA gyrase, 17 Å for E. coli DNA gyrase and 15.5 Å for S. aureus and E. coli topo IV,
respectively) and all ligands were protonated. The same settings and technical parameters
of the GOLD genetic algorithm (population size = 100, selection pressure = 1.1, number of
operations = 100.000, number of islands = 5, niche size = 2, migrate = 10, mutate = 95, cross-
over = 95) were used for all calculations by docking each molecule 10 times into the binding
site. Amino acids Met75/Leu74/Ile71/Leu74 on
α3 helix, Asp83/Asp82/Asp79/Asp79
on 4 helix and Met121/Met120/Met117/Met118 were treated as flexible during docking.
α
The molecular docking protocol was validated by re-docking the co-crystallized NBTI
ligands (GSK299423) 3-fold to reproduce their spatial conformation and orientation. The
heavy-atoms root-mean-square deviation (RMSD
≤
2.0 Å) between each calculated docking
pose and co-crystallized ligand conformation served as decisive criteria for quality of all
structure-based settings [38]. The quality of the calculated docked poses was first visually
examined in terms of their correct spatial orientation and conformation within the NBTI

Antibiotics 2021, 10, 862
19 of 30
binding pocket relative to the natively present co-crystallized ligand conformation and
additionally by the calculated GOLDScore Fitness function [37].
4.4. Analytical Techniques
All chemicals were obtained from commercial sources and were used without further
purification. The reactions requiring anhydrous conditions were carried out under inert
argon atmosphere with anhydrous solvents. Reactions were monitored by thin-layer
chromatography (TLC) on Merck silica gel (60 F254) plates (0.25 mm thick), visualized by
UV light and/or staining reagents. Column chromatography for compound purification
of the final compounds was carried out on silica gel 60 (particle size 0.040–0.063 mm;
Merck, Kenilworth, NJ, USA). HPLC purity was determined on Thermo Scientific DIONEX
UltiMate 3000 instrument equipped with diode array detector using Acquity UPLC® BEH
C8 column (1.7
acid in water (A) and acetonitrile (B), employing the following gradient: 90% A to 50%
A in 6 min, then 50% A for 3 min, with flow rate 0.3 mL/min and injection volume 5 L.
µm. 2.1 mm
×
50 mm). The solvent system consisted of 0.1% trifluoroacetic
µ
Melting points were determined on a Reichert hot stage microscope and are uncorrected.
NMR spectra (1H and 13C) were obtained on a Bruker AVANCE III spectrometer (Bruker
Corporation. Billerica, MA, USA) at 400 and 100 MHz, respectively in DMSO-d6 or CDCl3
solution with tetramethylsilane (TMS) as an internal standard. High resolution mass
spectra were obtained on Advion CMS (Advion Inc., Ithaca, NY, USA) or VG Analytical
Autospec Q mass spectrometer (Fisons, VG Analytical, Manchester, UK). IR spectra were
recorded on Thermo Nicolet Nexus 470 ESP FT-IR spectrometer (Thermo Fisher Scientific,
Waltham, MA, USA).
4.5. Synthesis
Synthesis and analytical characterization of compounds 11
where [13].
–18 is described else-
4.6. 1-(2-(6-Methoxy-1,5-naphthyridin-4-yl)ethyl)piperidin-4-amine (2)
Tert-butyl 1-(2-(6-methoxy-1,5-naphthyridin-4-yl)ethyl)piperidin-4-ylcarbamate (0.63 g,
1.63 mmol, 1 equiv) was dissolved in 12.5 mL of 1 M HCl in acetic acid and stirred for 30 min
at room temperature. Solvent was evaporated, the residue dissolved in 30 mL of saturated
NaHCO3(aq) and pH adjusted to 12 with 1 M NaOH. Water layer was washed with DCM
(3
×
30 mL), combined organic layers dried over Na2SO4 and concentrated in vacuum to
as dark-brown viscous liquid (0.33 g, 70.7%). Rf = 0.10 (DCM/MeOH
afford compound
4:1 + 1% Et3N). H NMR (400 MHz, CDCl3)
2
1
δ = 1.41–1.51 (m, 2H; N(CH2CH2)2CH),
1.88 (d, J = 11.9 Hz, 2H; N(CH2CH2)2CH), 2.20 (t, J = 11.0 Hz, 2H; N(CH2CH2)2CH),
2.66–2.78 (m, 1H; N(CH2CH2)2CH), 2.73–2.84 (m, 2H; CH2CH2N), 3.05 (d, J = 11.5 Hz, 2H;
N(CH2CH2)2CH), 3.30–3.46 (m, 2H; CH2CH2N), 4.08 (s, 3H; CH3O), 7.11 (d, J = 9.0 Hz, 1H;
Ar–H), 7.41 (d, J = 4.5 Hz, 1H; Ar–H), 8.19 (d, J = 9.0 Hz, 1H; Ar–H), 8.66 (d, J = 4.5 Hz, 1H;
Ar–H) ppm. 13C NMR (100 MHz, CDCl3):
δ
= 28.46, 35.83, 52.37, 53.72, 58.37, 77.23, 116.31,
124.24, 140.33, 140.97, 141.48, 146.64, 147.70, 161.43 ppm. IR (ATR): = 3271, 2941, 2811,
ν
1614, 1590, 1503, 1489, 1438, 1398, 1334, 1125, 1079, 1018, 932, 853, 807, 751, 713, 615, 584,
536 cm–1. HRMS (ESI): m/z calculated for C16H22N4O14 + H+: 287.1872 [M + H]+, found:
287.1875.
4.6.1. General Procedure for Reductive Amination
To a solution of amine 2 (1 equiv) in dry methanol, the appropriate aldehyde
(1–1.3 equiv) was added together with the catalytic amount of acetic acid. The reaction
mixture was stirred for 2 h at the◦room temperature under the inert atmosphere. Next, the
reaction mixture was cooled to 0 C and NaCNBH3 (3–5 equiv) dissolved in small amount
of dry methanol was added drop-wise. After stirring overnight at room temperature under
inert atmosphere, the solvent was evaporated in vacuum and the residue diluted in ethyl
acetate. Organic phase was washed with saturated Na2CO3 three times, dried over Na2SO4,

Antibiotics 2021, 10, 862
20 of 30
filtered and the solvent evaporated under reduced pressure. The residue was purified by
flash column chromatography to afford appropriate substituted product.
4.6.2. 1-(2-(6-Methoxy-1,5-naphthyridin-4-yl)ethyl)-N-((1-methyl-1H-benzo[d]imidazol-
2-yl)methyl)piperidin-4-amine (3)
According to the general procedure for reductive amination, compound
2 (0.15 g,
0.524 mmol, 1 equiv) was used together with the 1-methyl-1H-benzo[d]imidazole-2-
carbaldehyde (0.101 g, 0.631 mmol, 1.2 equiv). The residue was purified by flash column
chromatography using CHCl3/MeOH (15:1 following 9:1 + 1% Et3N) to afford compound
3
(0.083 g, 36.6%) as a brown-red solid. Rf = 0.17 (CHCl3/MeOH 15:1 + 1% Et3N). m.p.
57–62 ◦C. 1H NMR (400 MHz, CDCl3):
δ
= 1.43–1.61 (m, 2H; N(CH2CH2)2CH), 1.99 (d,
J = 11.2 Hz, 2H; N(CH2CH2)2CH), 2.21 (t, J = 10.8 Hz, 2H; N(CH2CH2)2CH), 2.59–2.69 (m,
1H; N(CH2CH2)2CH), 2.78–2.82 (m, 2H; CH2CH2N), 3.06 (d, J = 11.6 Hz, 2H; N(CH2CH2)2
CH), 3.33–3.42 (m, 2H; CH2CH2N), 3.84 (s, 3H; CH3N), 4.08 (s, 3H; CH3O), 4.12 (s, 2H;
NHCH2), 7.11 (d, J = 9.0 Hz, 1H; Ar–H), 7.23–7.31 (m, 2H; 2
Ar–H), 7.41 (d, J = 4.5 Hz, 1H; Ar–H), 7.70–7.75 (m, 1H; Ar–H), 8.18 (d, J = 9.0 Hz, 1H;
Ar–H), 8.66 (d, J = 4.5 Hz, 1H; Ar–H) ppm. 13C NMR (100 MHz, CDCl3):
= 28.43, 29.96,
32.52, 43.89, 52.18, 53.74, 58.39, 77.24*, 109.09, 116.32, 119.46, 121.94, 122.49, 124.25, 136.16,
140.33, 140.96, 141.48, 142.26, 146.57, 147.70, 153.18, 161.43 ppm. IR (ATR): = 3390, 2932,
×
Ar–H) *, 7.32–7.36 (m, 1H;
δ
ν
2798, 1611, 1590, 1501, 1490, 1480, 1447, 1400, 1353, 1333, 1261, 1201, 1187, 1111, 1077, 1012,
874, 843, 770, 642, 584 cm–1. HRMS (ESI): m/z calculated for C25H30N6O+H+: 431.2559 [M
+ H]+, found: 431.2563. HPLC purity (254 nm): 94.9%. * signal is overlapping with residual
solvent peak.
4.6.3. N-((1-(4-Fluorophenyl)-1H-pyrazol-4-yl)methyl)-1-(2-(6-methoxy-1,5-naphthyridin-
4-yl)ethyl)piperidin-4-amine (4)
According to the general procedure for reductive amination, compound
2 (0.220 g,
0.768 mmol, 1 equiv) was used together with the 1-(4-fluorophenyl)-1H-pyrazole-4-carbalde
hyde (0.220 g, 1.157 mmol, 1.5 equiv). The residue was purified by flash column chromatog-
raphy using CHCl3/MeOH (15:1 + 1% Et3N) to afford compound
brown solid. Rf = 0.12 (DCM/MeOH 4:1 + 1% Et3N). m.p. 59–74 ◦C. 1H NMR (400 MHz,
CDCl3): = 1.45–1.55 (m, 2H; N(CH2CH2)2CH), 1.98 (d, J = 12.0 Hz, 2H; N(CH2CH2)2CH),
4 (0.220 g, 41.3%) as a light
δ
2.19 (t, J = 10.9 Hz, 2H; N(CH2CH2)2CH), 2.53–2.70 (m, 1H; N(CH2CH2)2CH), 2.75–2.85 (m,
2H; CH2CH2N), 3.07 (d, J = 11.7 Hz, 2H; N(CH2CH2)2CH), 3.37–3.41 (m, 2H; CH2CH2N),
3.81 (s, 2H; NHCH2), 4.08 (s, 3H; CH3O), 7.09–7.17 (m, 3H; 3
1H; Ar–H), 7.60–7.65 (m, 2H; 2 Ar–H), 7.65 (s, 1H; Ar–H), 7.84 (s, 1H; Ar–H), 8.18 (d,
J = 9.0 Hz, 1H; Ar–H), 8.66 (d, J = 4.5 Hz, 1H; Ar–H) ppm. 13C NMR (100 MHz, CDCl3):
= 28.46, 32.66, 40.73, 52.35, 53.72, 58.41, 77.24 *, 116.22 (d, J = 23.0 Hz), 116.29, 120.66 (d,
J = 8.3 Hz), 124.25, 125.67, 136.57, 140.36, 140.76, 140.98, 141.50, 146.69, 147.72, 161.02 (d,
J = 245.8 Hz), 161.42, 163.59 ppm. IR (ATR): = 2927, 2803, 1613, 1591, 1489, 1458, 1443,
×
Ar–H), 7.42 (d, J = 4.5 Hz,
×
δ
ν
1400, 1334, 11263, 1216, 1157, 1126, 1098, 1077, 1042, 1019, 952, 851, 836, 810, 774, 768, 743,
713, 661, 625, 524 cm–1. HRMS (ESI): m/z calculated for C26H29FN6O + H+: 461.2465 [M +
H]+, found: 461.2452. HPLC purity (254 nm): 90.9%. * signal is overlapping with residual
solvent peak.
4.6.4. 1-(2-(6-Methoxy-1,5-naphthyridin-4-yl)ethyl)-N-((1-phenyl-1H-pyrazol-4-
yl)methyl)piperidin-4-amine (5)
According to the general procedure for reductive amination, compound
2 (0.220 g,
0.768 mmol, 1 equiv) was used together with the 1-phenyl-1H-pyrazole-4-carbaldehyde
(0.198 g, 1.150 mmol, 1.5 equiv). The residue was purified by flash column chromatography
using CHCl3/MeOH (15:1) to afford compound
5
(0.220 g, 43.23%) as an orange-red viscous
= 1.48–1.59
liquid. Rf = 0.19 (DCM/MeOH 4:1 + 1% Et3N). 1H NMR (400 MHz, CDCl3):
δ
(m, 2H; N(CH2CH2)2CH), 2.00 (d, J = 14.3 Hz, 2H; N(CH2CH2)2CH), 2.19 (t, J = 10.8 Hz, 2H;
N(CH2CH2)2CH), 2.60–2.70 (m, 1H; N(CH2CH2)2CH), 2.77–2.83 (m, 2H; CH2CH2N), 3.09

Antibiotics 2021, 10, 862
21 of 30
(d, J = 11.9 Hz, 2H; N(CH2CH2)2CH), 3.36–3.42 (m, 2H; CH2CH2N), 3.82 (s, 2H; NHCH2),
4.08 (s, 3H; CH3O), 7.11 (d, J = 9.0 Hz, 1H; Ar–H), 7.24–7.30 (m, 1H; Ar–H) *, 7.40–7.45 (m,
3H; 3
Ar–H), 8.66 (d, J = 4.5 Hz, 1H; Ar–H) ppm. 13C NMR (100 MHz, CDCl3):
40.69, 52.33, 53.70, 59.19, 58.37, 116.29, 118.88, 119.27, 122.46, 124.23, 125.59, 126.31, 129.42,
140.12, 140.27, 140.73, 140.96, 141.43, 146.72, 147.63, 161.41 ppm. IR (ATR): = 2941, 2807,
×
Ar–H), 7.65–7.69 (m, 3H; 3
×
Ar–H), 7.93 (s, 1H, Ar–H), 8.19 (d, J = 9.0 Hz, 1H;
δ
= 28.43, 32.58,
ν
1612, 1594, 1502, 1489, 1463, 1437, 1398, 1373, 1333, 1259, 1105, 1076, 1017, 953, 851, 807,
712, 690, 653, 610, 596, 584, 508 cm–1. HRMS (ESI): m/z calculated for C26H30N6O + H+:
443.2559 [M + H]+, found: 443.2560. HPLC purity (254 nm): 97.1%. * signal is overlapping
with residual solvent peak.
4.6.5. N-((2,8-Dimethylimidazo[1,2-a]pyridin-3-yl)methyl)-1-(2-(6-methoxy-1,5-
naphthyridin-4-yl)ethyl)piperidin-4-amine (6)
According to the general procedure for reductive amination, compound
2 (0.252 g,
0.880 mmol, 1 equiv) was used together with the 2,8-dimethylimidazo[1,2-a]pyridine-3-
carbaldehyde (0.157 g, 0.901 mmol, 1 equiv). The residue was purified by flash column
chromatography using CHCl3/MeOH (9:1) to afford compound 6 (0.252 g, 62.9%) as an
orange-brown solid. The synthesis of 2,8-dimethylimidazo[1,2-a]pyridine-3-carbaldehyde
was previously reported [18]. Rf = 0.23 (DCM/MeOH 4:1 + 1% Et3N). m.p. 50–61 ◦C. 1H
NMR (400 MHz, CDCl3):
δ = 1.46–1.55 (m, 2H; N(CH2CH2)2CH), 1.97 (d, J = 10.1 Hz, 2H;
N(CH2CH2)2CH), 2.22 (t, J = 10.8 Hz, 2H; N(CH2CH2)2CH), 2.48 (s, 3H; CH3Ar) 2.50–2.58
(m, 1H; N(CH2CH2)2CH), 2.60 (s, 3H; CH3Ar), 2.79–2.87 (m, 2H; CH2CH2N), 3.07 (d,
J = 11.8 Hz, 2H; N(CH2CH2)2CH), 3.35–3.43 (m, 2H; CH2CH2N), 4.07 (s, 3H; CH3O), 4.11
(s, 2H; NHCH2), 6.70 (t, J = 6.8 Hz, 1H; Ar–H), 6.92–7.00 (m, 1H; Ar–H), 7.12 (d, J = 9.0 Hz,
1H; Ar–H), 7.41 (d, J = 4.5 Hz, 1H; Ar–H), 8.15 (d, J = 6.7 Hz, 1H; Ar–H), 8.18 (d, J = 9.0
Hz, 1H; Ar–H), 8.66 (d, J = 4.5 Hz, 1H; Ar–H) ppm. 13C NMR (100 MHz, CDCl3):
δ = 13.38,
17.10, 28.37, 32.51, 39.66, 52.25, 53.73, 58.32, 77.23 *, 111.59, 116.35, 118.45, 122.38, 123.06,
124.26, 126.17, 139.85, 140.37, 140.93, 141.51, 144.94, 146.33, 147.72, 161.46 ppm. IR (ATR):
ν
= 3258, 2925, 2813, 1613, 1590, 1489, 1437, 1398, 1373, 1354, 1333, 1259, 1102, 1020, 983,
853, 807, 767, 745, 641, 584, 568 cm–1. HRMS (ESI): m/z calculated for C26H32N6O + H+:
445.2716 [M + H]+, found: 445.2704. HPLC purity (254 nm): 98.4%. * signal is overlapping
with residual solvent peak.
4.6.6. 8-Cyano-N-(1-(2-(6-methoxy-1,5-naphthyridin-4-yl)ethyl)piperidin-4-
yl)imidazo[1,2-a]pyridine-2-carboxamide (7)
Compound
2 (0.209 g, 0.730 mmol, 1 equiv), TBTU (0.234 g, 0.729 mmol, 1 equiv) and
Et3N (0.31 mL, 2.223 mmol, 3 equiv) were dissolved in 5 mL of CH3CN and stirred at the
room temperature under inert atmosphere for 30 min. 8-cyanoimidazo[1,2-a]pyridine-2-
carboxylic acid (0.209 g, 0.730 mmol, 1 equiv) dissolved in 3 mL of CH3CN was added
drop-wise and the reaction mixture stirred overnight. The precipitate was filtered with
vacuum and dried under the reduced pressure. The residue was purified with flash column
chromatography using CHCl3/MeOH (9:1 + 1% Et3N) to afford compound
7 (0.209 g,
62.7%) as a white solid. The synthesis of 8-cyanoimidazo[1,2-a]pyridine-2-carboxyl◦ic acid
was previously reported [18]. Rf = 0.3 (CHCl3/MeOH 9:1 + 1% Et3N). m.p. 175–179 C. 1H
NMR (400 MHz, CDCl3):
δ = 1.63–1.75 (m, 2H; N(CH2CH2)2CH), 2.09 (d, J = 12.5 Hz, 2H;
N(CH2CH2)2CH), 2.35 (t, J = 11.3 Hz, 2H; N(CH2CH2)2CH), 2.80–2.86 (m, 2H; CH2CH2N),
3.09 (d, J = 11.3 Hz, 2H; N(CH2CH2)2CH), 3.42–3.36 (m, 2H; CH2CH2N), 3.99–4.07 (m, 1H;
N(CH2CH2)2CH), 4.09 (s, 3H; CH3O), 7.12 (d, J = 9.0 Hz, 1H; Ar–H), 7.38 (dd, J1 = 9.5 Hz,
J2 = 1.7 Hz, 1H; Ar–H) 7.42 (d, J = 4.5 Hz, 1H; Ar–H), 7.66–7.70 (m, 1H; Ar–H), 8.19 (d,
J = 9.0 Hz, 1H; Ar–H), 8.25 (s, 1H; Ar–H), 8.63–8.64 (m, 1H; Ar–H), 8.68 (d, J = 4.5 Hz, 1H;
Ar–H) ppm. 13C NMR (100 MHz, CDCl3):
δ = 28.45, 32.29, 46.57, 52.27, 53.72, 58.38, 77.23 *,
100.08, 115.16, 115.86, 116.33, 119.28, 124.22, 125.68, 132.66, 140.35, 140.96, 141.50, 142.33,
143.54, 146.55, 147.69, 160.89, 161.44 ppm. IR (ATR): ν = 3403, 3276, 3128, 2948, 2814, 2365,
2229, 2176, 1661, 1650, 1612, 1591, 1564, 1503, 1430, 1400, 1335, 1302, 1258, 1230, 1137, 1121,

Antibiotics 2021, 10, 862
22 of 30
1074, 1018, 855 cm–1. HRMS (ESI): m/z calculated for C25H25N7O2 + H+: 456.2148 [M +
H]+, found: 456.2143. HPLC purity (254 nm): 99.8%. * signal is overlapping with residual
solvent peak.
4.6.7. 2-Phenyl-2H-1,2,3-triazole-4-carbaldehyde (23)
The compound was synthesized according to the procedure from the literature (Sup-
plementary Materials Scheme S1) [39]. To a suspension of glucose (5 g, 27.8 mmol, 1 equiv)
in 50 mL of water, phenylhydrazine hydrochloride (12 g, 83.3 mmol, 3 equiv), sodium
acetate (14 g, 166.8 mmol, 6 equiv) and catalytic amount of acetic acid were added. The
reaction mixture was refluxed for 1h. After reaction was complete, the yellow residue was
collected by vacuum filtration (7.308 g, 73.4%). Rf = 0.48 (DCM/MeOH 4:1 + 1% Et3N).
1H NMR (400 MHz, DMSO-d6):
CH2(OH)), 4.34–4.37 (m, 1H; CH(OH)), 4.51–4.53 (m, 1H; CH2(OH)), 4.56–4.60 (m, 2H;
2 × CH(OH)), 5.10 (d, J = 5.2 Hz; 1H, CH(OH)C(NNHPh)), 6.83–6.88 (m, 2H; 2 × Ar–H)
7.03 (dd, J1 = 1.2 Hz, J2 = 8.8 Hz; 2H, 2 Ar–H), 7.14 (dd, J1 = 1.2 Hz, J2 = 8.8 Hz, 2H;
Ar–H), 7.87 (s, 1H; CHNNHPh), 10.70 (s, 1H; CHNNHPh),
δ
= 3.40–3.50 (m, 2H; 2
×
CH(OH)), 3.53–3.65 (m, 2H;
,
×
2 × Ar–H), 7.28–7.38 (m, 4H; 4
×
12.29 (s, 1H; CNNHPh) ppm.
Next, the crude product (7.303 g, 20.4 mmol, 1 equiv) was mixed with the copper
sulfate pentahydrate (7.128 g, 28.5 mmol, 1.42 equiv) in water (150 mL). The reaction
was refluxed for 2 h and filtered hot after completion, to remove the copper salts. The
product was crystallized after cooling to room temperature and filtered with vacuum.
As the product was unclean, it was recrystallized from ethanol as a light brown solid
(1.76 g, 32.5%). Rf = 0.40 (DCM/MeOH 4:1 + 1% Et3N). 1H NMR (400 MHz, DMSO-d6):
δ
= 3.54–3.64 (m, 4H; CH(OH)CH2(OH)), 4.39–4.40 (m, 1H; CH(OH)), 4.65–4.69 (m, 2H;
CH(OH)CH(OH)), 5.11 (s, 1H; CH(OH)), 5.29 (s, 1H; CH(OH)), 7.23 (s, 1H; Ar–H), 7.38–7.41
(m, 1H; Ar–H), 7.54–7.57 (m, 2H; 2 × Ar–H), 7.95–7.99 (m, 2H; 2 × Ar–H) ppm.
To the product (1.76 g, 6.64 mmol, 1 equiv), dissolved in water (10 mL), sodium
metaperiodate (0.85 g, 4.0 mmol, 0.3 equiv) was added and the reaction mixture stirred
at the room temperature overnight. Upon the reaction completion, the crystalized crude
product was vacuum filtered and washed with diethyl ether, while the mother liquor was
washed with DCM (3
vacuum. Compound 23 is a dark brown solid (0.282 g, 24.5%). Rf = 0.82 (DCM/MeOH 4:1).
1H NMR (400 MHz, DMSO-d6):
= 7.53–7.57 (m, 1H; Ar–H), 7.63–7.67 (m, 2H; 2 Ar–H),
8.11–8.13 (m, 2H; 2 × Ar–H), 8.70 (s, 1H; Ar–H), 10.19 (s, 1H; CHCO) ppm.
×
10 mL), organic phase was dried over Na2SO4 and concentrated in
δ
×
4.6.8. 1-(2-(6-Methoxy-1,5-naphthyridin-4-yl)ethyl)-N-((2-phenyl-2H-1,2,3-triazol-4-
yl)methyl)piperidin-4-amine (8)
According to the general procedure for reductive amination, compound
2 (0.20 g,
0.698 mmol, 1 equiv) was used together with the 23 (0.169 g, 0.908 mmol, 1.3 equiv). The
residue was purified by flash column chromatography using CHCl3/MeOH (9:1) to afford
compound
8 (0.134 g, 43.3%) as a dark brown solid. Rf = 0.45 (DCM/MeOH 4:1), 0.22
(CHCl3/MeOH 9:1). m.p. 74.1–78.9 ◦C. 1H NMR (400 MHz, CDCl3): δ = 1.50–1.58 (m, 2H;
N(CH2CH2)2CH) *, 1.99–2.02 (m, 2H; N(CH2CH2)2CH), 2.19–2.24 (m, 2H; N(CH2CH2)2CH),
2.60–2.65 (m, 1H; N(CH2CH2)2CH), 2.79–2.83 (m, 2H; CH2CH2N), 3.07–3.10 (m, 2H;
N(CH2CH2)2CH), 3.38–3.42 (m, 2H; CH2CH2N), 4.03 (s, 2H; NHCH2), 4.09 (s, 3H; CH3O),
7.12 (d, J = 9.2 Hz, 1H; Ar–H), 7.33–7.37 (m, 1H; Ar–H), 7.43 (d, J = 4.4 Hz, 1H; Ar–H),
7.46–7.51 (m, 2H; 2
J = 9.2 Hz, 1H; Ar–H), 8.67 (d, J = 4.4 Hz, 1H; Ar–H) ppm. 13C (100 MHz, CDCl3):
32.69, 41.73, 52.31, 53.72, 54.36, 58.42, 77.23 *, 116.29, 118.72, 124.25, 127.33, 129.27, 134.44,
139.85, 140.35, 140.98, 141.49, 146.72, 147.71, 148.88, 164.41 ppm. IR (ATR): = 2915, 2814,
×
Ar–H), 7.75 (s, 1H; Ar–H), 8.04–8.07 (m, 2H; 2
×
Ar–H), 8.19 (d,
δ
= 28.47,
ν
1611, 1593, 1489, 1459, 1399, 1334, 1259, 1131, 1108, 1078, 1019, 990, 966, 851, 807, 756, 713,
692, 662, 603, 585, 545 cm−1. HRMS (ESI): m/z calculated for C25H30ON7 + H+: 444.2506 [M
+ H]+, found: 444.2504. HPLC purity (254 nm): 96.5%. * signal is overlapping with residual
solvent peak.

Antibiotics 2021, 10, 862
23 of 30
4.6.9. N-((1-Isopropyl-1H-pyrazol-4-yl)methyl)-1-(2-(6-methoxy-1,5-naphthyridin-4-
yl)ethyl)piperidin-4-amine (9)
According to the general procedure for reductive amination, compound
2 (0.201 g,
0.702 mmol, 1 equiv) was used together with the 1-isopropyl-1H-pyrazole-4-carbaldehyde
(0.100 mL, 0.782 mmol, 1.1 equiv). The residue was purified by flash column chromatog-
raphy using CHCl3/MeOH (9:1 + 1% Et3N) to afford compound
light brown viscous liquid. Rf = 0.24 (CHCl3/MeOH 9:1 + 1% Et3N). 1H NMR (400 MHz,
CDCl3): = 1.49 (d, J = 6.7 Hz, 6H; 2 CH(CH3)2), 1.60–1.74 (m, 2H; N(CH2CH2)2CH), 2.07
9 (0.201 g, 62.9%) as a
δ
×
(d, J = 10.7 Hz, 2H; N(CH2CH2)2CH), 2.30–2.49 (m, 2H; N(CH2CH2)2CH), 2.72–2.84 (m, 1H;
N(CH2CH2)2CH), 2.90–3.00 (m, 2H; CH2CH2N), 3.18 (d, J = 11.9 Hz, 2H; N(CH2CH2)2CH),
3.36–3.50 (m, 2H; CH2CH2N), 3.79 (s, 2H; NHCH2), 4.08 (s, 3H; CH3O), 4.40–4.54 (m, 1H;
CH(CH3)2) 7.12 (d, J = 9.0 Hz, 1H; Ar–H), 7.44 (d, J = 4.5 Hz, 1H; Ar–H), 7.46 (s, 1H; Ar–H),
7.48 (s, 1H; Ar–H), 8.19 (d, J = 9.0 Hz, 1H; Ar–H), 8.67 (d, J = 4.5 Hz, 1H; Ar–H) ppm. 13C
NMR (100 MHz, CDCl3):
δ = 18.43, 22.97, 28.47, 29.71, 32.76, 40.98, 52.43, 53.64, 53.70, 54.44,
58.45, 116.26, 119.96, 124.23, 125.22, 138.08, 140.33, 141.00, 141.49, 146.81, 147.69, 161.41
ppm. IR (ATR):
ν = 2942, 2814, 2328, 2169, 1613, 1590, 1490, 1439, 1399, 1372, 1334, 1261,
1182, 1122, 1079, 1018, 984, 854, 808, 750, 625, 585, 539 cm–1. HRMS (ESI): m/z calculated
for C23H32N6O + H+: 409.2710 [M + H]+, found: 409.2706. HPLC purity (254 nm): 91.7%.
signal is overlapping with residual solvent peak.
4.6.10. N-((1-Allyl-1H-pyrazol-4-yl)methyl)-1-(2-(6-methoxy-1,5-naphthyridin-4-
yl)ethyl)piperidin-4-amine (10)
According to the general procedure for reductive amination, compound
2 (0.20 g,
0.698 mmol, 1 equiv) was used together with the 1-allyl-1H-pyrazole-4-carbaldehyde (0.112
g, 0.823 mmol, 1.2 equiv). The residue was purified by flash column chromatography
using CHCl3/MeOH (9:1 + 1% Et3N) to afford compound 10 (0.200 g, 59.8%) as a dark
red viscous liquid. Rf = 0.16 (DCM/MeOH 4:1 + 1% Et3N). 1H NMR (400 MHz, CDCl3):
δ
= 1.50–1.61 (m, 2H; N(CH2CH2)2CH), 2.00 (d, J = 11.1 Hz, 2H; N(CH2CH2)2CH) *, 2.30
(t, J = 12.9 Hz, 2H; N(CH2CH2)2CH), 2.60–2.71 (m, 1H; N(CH2CH2)2CH), 2.83–2.91 (m,
2H; CH2CH2N), 3.11 (d, J = 11.8 Hz, 2H; N(CH2CH2)2CH), 3.37–3.46 (m, 2H; CH2CH2N),
3.74 (s, 2H; NHCH2), 4.08 (s, 3H; CH3O), 4.71 (m, 2H; CH2CH=CH2), 5.19–5.31 (m, 2H;
CH2CH=CH2), 6.01 (m, 1H; CH2CH=CH2), 7.12 (d, J = 9.0 Hz, 1H; Ar–H), 7.43 (s, 3H;
3 × Ar–H), 8.19 (d, J = 9.0 Hz, 1H; Ar–H), 8.66 (d, J = 4.5 Hz, 1H; Ar–H) ppm. 13C NMR
(100 MHz, CDCl3):
δ = 28.43, 32.46, 40.64, 44.04, 46.12, 52.32, 53.69, 54.15, 54.73, 58.37,
116.27, 118.65, 120.31, 124.22, 128.00, 132.90, 138.93, 140.28, 140.96, 141.44, 146.72, 147.64,
161.40 ppm. IR (ATR):
ν = 2938, 2805, 2324, 2167, 1612, 1589, 1489, 1439, 1398, 1372, 1333,
1260, 1119, 1077, 1018, 990, 928, 853, 807, 731, 642, 585 cm–1. HRMS (ESI): m/z calculated
for C23H30N6O + H+: 407.2554 [M + H]+, found: 407.2547. HPLC purity (254 nm): 96.0%.
* signal is overlapping with residual solvent peak.
4.6.11. N-(4-(Difluoromethoxy)-3-methoxybenzyl)-1-(2-(6-methoxy-1,5-naphthyridin-4-
yl)ethyl)piperidin-4-amine (19)
According to the general procedure for reductive amination, compound
2 (0.194 g,
0.677 mmol, 1 equiv) was used together with the 4-(difluoromethoxy)-3-methoxybenzaldeh
yde (0.120 mL, 0.748 mmol, 1.1 equiv). The residue was purified by flash column chromatog-
raphy using CHCl3/MeOH (9:1 + 1% Et3N) to afford compound 19 (0.194 g, 54.9%) as a
◦
red-brown solid. Rf = 0.19 (DCM/MeOH 4:1 + 1% Et3N). m.p. 85–93 C. 1H NMR (400 MHz,
CDCl3):
δ = 1.44–1.55 (m, 2H; N(CH2CH2)2CH), 1.96 (d, J = 11.6 Hz, 2H; N(CH2CH2)2CH),
2.19 (t, J = 10.5 Hz, 2H; N(CH2CH2)2CH), 2.48–2.59 (m, 1H; N(CH2CH2)2CH), 2.72–2.84 (m,
2H; CH2CH2N), 3.06 (d, J = 11.7 Hz, 2H; N(CH2CH2)2CH), 3.34–3.41 (m, 2H; CH2CH2N),
3.81 (s, 2H; NHCH2), 3.89 (s, 3H; CH3O), 4.07 (s, 3H; CH3O), 6.53 (t, J = 75.4 Hz, 1H;
OCHF2), 6.88 (dd, J1 = 8.1 Hz, J2 = 1.8 Hz, 1H; Ar–H), 7.01 (d, J = 1.6 Hz, 1H; Ar–H), 7.10 (d,
J = 8.4 Hz, 1H; Ar–H), 7.11 (d, J = 8.8 Hz, 1H; Ar–H), 7.41 (d, J = 4.5 Hz, 1H; Ar–H), 8.18 (d,
J = 9.0 Hz, 1H; Ar–H), 8.66 (d, J = 4.5 Hz, 1H; Ar–H) ppm. 13C NMR (100 MHz, CDCl3):
δ

Antibiotics 2021, 10, 862
24 of 30
= 28.45, 32.79, 50.51, 52.36, 53.71, 55.91, 58.41, 77.23 *, 112.34, 116.25 (t, J = 259.5 Hz), 116.29,
120.21, 122.22, 124.25, 138.73, 139.72, 140.37, 140.97, 141.50, 146.62, 147.72, 151.08, 161.42
ppm. IR (ATR):
ν = 2937, 2832, 1610, 1592, 1507, 1490, 1465, 1419, 1398, 1374, 1334, 1264,
1212, 1191, 1079, 1034, 1015, 853, 811, 787, 753, 732, 586 cm–1. HRMS (ESI): m/z calculated
for C25H30F2N4O3 + H+: 473.2359 [M + H]+, found: 473.2354. HPLC purity (254 nm): 97.2%.
* signal is overlapping with residual solvent peak.
4.6.12. 2-(Difluoromethoxy)-5-((1-(2-(6-methoxy-1,5-naphthyridin-4-yl)ethyl)piperidin-4-
ylamino)methyl)phenol (20)
According to the general procedure for reductive amination, compound
2 (0.171 g,
0.597 mmol, 1 equiv) was used together with the 4-(difluoromethoxy)-3-hydroxybenzaldehy
de (0.138 g, 0.734 mmol, 1.2 equiv). The residue was purified by flash column chromatogra-
phy using CHCl3/MeOH (9:1 + 1% Et3N) to afford compound 20 (0.171 g, 50.9%) as a light
brown solid. Rf = 0.09 (DCM/MeOH 4:1 + 1% Et3N). m.p. 99–104 ◦C. 1H NMR (400 MHz,
CDCl3):
δ = 1.48–1.57 (m, 2H; N(CH2CH2)2CH), 1.98 (d, J = 11.2 Hz, 2H; N(CH2CH2)2CH),
2.18–2.24 (m, 2H; N(CH2CH2)2CH), 2.55–2.61 (m, 1H; N(CH2CH2)2CH), 2.78–2.86 (m, 2H;
CH2CH2N), 3.08 (d, J = 11.8 Hz, 2H; N(CH2CH2)2CH), 3.36–3.45 (m, 2H; CH2CH2N), 3.78
(s, 2H; NHCH2), 4.10 (s, 3H; CH3O), 6.53 (t, J = 74.0 Hz, 1H; OCHF2), 6.85 (dd, J1 = 8.2 Hz,
J2 =2.0 Hz, 1H; Ar–H), 7.02 (d, J = 2.0 Hz, 1H; Ar–H), 7.06 (d, J = 8.2 Hz, 1H; Ar–H), 7.14 (d,
J = 9.0 Hz, 1H; Ar–H), 7.44 (d, J = 4.4 Hz, 1H; Ar–H), 8.21 (d, J = 9.0 Hz, 1H; Ar–H), 8.68
(d, J = 4.5 Hz, 1H; Ar–H) ppm. 13C NMR (100 MHz, CDCl3):
δ
= 28.12, 31.72, 49.94, 51.96,
53.77, 58.06, 77.23 *, 116.38 (t, J = 260.8 Hz), 116.46, 117.32, 119.91, 120.53, 124.34, 137.80,
138.54, 140.22, 140.89, 141.38, 146.10, 147.61, 148.08, 161.53 ppm. IR (ATR): = 2944, 2338,
ν
1612, 1595, 1490, 1438, 1399, 1334, 1297, 1261, 1113, 1076, 1049, 1014, 851, 809, 782, 752, 714,
674, 641, 586 cm–1. HRMS (ESI): m/z calculated for C24H28F2N4O3 + H+: 459.2202 [M +
H]+, found: 459.2196. HPLC purity (254 nm): 97.9%. * signal is overlapping with residual
solvent peak.
4.6.13. 1-(2-(6-Methoxy-1,5-naphthyridin-4-yl)ethyl)-N-((6-(trifluoromethyl)pyridin-3-
yl)methyl)piperidin-4-amine (21)
According to the general procedure for reductive amination, compound
2 (0.184 g,
0.643 mmol, 1 equiv) was used together with the 6-(trifluoromethyl)nicotinaldehyde (0.118
g, 0.674 mmol, 1.05 equiv). The residue was purified by flash column chromatography
using CHCl3/MeOH (9:1 + 1% Et3N) to afford compound 21 (0.184 g, 61.28%) as a red-
brown solid. Rf = 0.22 (CHCl3/MeOH 9:1 + 1% Et3N). m.p. 75–78 ◦C. 1H NMR (400 MHz,
CDCl3):
δ = 1.43–1.54 (m, 2H; N(CH2CH2)2CH), 1.95 (d, J = 12.1 Hz, 2H; N(CH2CH2)2CH),
2.17 (t, J = 10.9 Hz, 2H; N(CH2CH2)2CH), 2.50–2.57 (m, 1H; N(CH2CH2)2CH), 2.76–2.83 (m,
2H; CH2CH2N), 3.05 (d, J = 11.8 Hz, 2H; N(CH2CH2)2CH), 3.35–3.41 (m, 2H; CH2CH2N),
3.94 (s, 2H; NHCH2), 4.07 (s, 3H; CH3O), 7.11 (d, J = 9.0 Hz, 1H; Ar–H), 7.41 (d, J = 4.5 Hz,
1H; Ar–H), 7.65 (d, J = 8.0 Hz, 1H; Ar–H), 7.91 (dd, J1 = 8.0 Hz, J2 = 1.4 Hz, 1H; Ar–H),
8.18 (d, J = 9.0 Hz, 1H; Ar–H), 8.66 (d, J = 4.5 Hz, 1H; Ar–H), 8.69 (d, J = 1.1 Hz, 1H; Ar–H)
ppm. 13C NMR (100 MHz, CDCl3):
δ
= 28.46, 32.78, 47.65, 52.26, 53.70, 58.38, 77.23 *, 116.31,
120.22 (q, J = 264.1 Hz), 121.64, 124.22, 136.86, 139.76, 140.34, 140.95, 141.48, 146.60, 146.88,
147.69, 149.72, 161.42 ppm. IR (ATR):
ν = 2946, 2796, 1612, 1590, 1490, 1449, 1400, 1373, 1334,
1262, 1172, 1119, 1083, 1014, 980, 837, 799, 751, 631, 586 cm–1. HRMS (ESI): m/z calculated
for C23H26F3N5O + H+: 446.2162 [M + H]+, found: 446.2157. HPLC purity (254 nm): 99.4%.
* signal is overlapping with residual solvent peak.
4.6.14. 1-(2-(6-Methoxy-1,5-naphthyridin-4-yl)ethyl)-N-((6-methoxypyridin-3-
yl)methyl)piperidin-4-amine (22)
According to the general procedure for reductive amination, compound
2 (0.20 g,
0.698 mmol, 1 equiv) was used together with the 6-methoxynicotinaldehyde (0.125 g,
0.908 mmol, 1.3 equiv). The residue was purified by flash column chromatography using
CHCl3/MeOH (9:1) to afford compound 22 (0.187 g, 65.7%) as a yellow-orange solid.

Antibiotics 2021, 10, 862
25 of 30
Rf = 0.20 (DCM/MeOH 4:1), 0.11 (CHCl3/MeOH 9:1). m.p. 67–69 ◦C. 1H NMR (400 MHz,
CDCl3):
δ = 1.44–1.54 (m, 2H; N(CH2CH2)2CH), 1.95–1.98 (m, 2H; N(CH2CH2)2CH),
2.17–2.22 (m, 2H; N(CH2CH2)2CH), 2.52–2.57 (m, 1H; N(CH2CH2)2CH), 2.78–2.82 (m, 2H;
CH2CH2N), 3.05–3.08 (m, 2H; N(CH2CH2)2CH), 3.38–3.42 (m, 2H; CH2CH2N), 3.78 (s, 2H;
NHCH2), 3.95 (s, 3H; CH3O), 4.09 (s, 3H; CH3O), 6.75 (dd, J1 = 8.4 Hz, J2 = 0.4 Hz, 1H;
Ar–H), 7.13 (d, J = 9.2 Hz, 1H; Ar–H), 7.43 (d, J = 4.4 Hz, 1H; Ar–H), 7.61 (dd, J1 = 8.4 Hz,
J2 = 2.4 Hz, 1H; Ar–H), 8.10 (d, J = 1.6 Hz, 1H; Ar–H), 8.20 (d, J = 9.2 Hz, 1H; Ar–H), 8.68
(d, J = 4.4 Hz, 1H; Ar–H) ppm. 13C (100 MHz, CDCl3):
δ
= 28.42, 32.65, 47.49, 52.26, 53.42,
53.72, 58.39, 77.24 *, 110.77, 116.30, 124.25, 128.67, 140.35, 139.00, 140.97, 141.49, 146.08,
146.62, 147.71, 161.42, 163.47 ppm. IR (ATR):
ν = 2929, 2808, 1610, 1591, 1449, 1461, 1397,
1376, 1334, 1313, 1292, 1256, 1107, 1073, 847, 770 cm−1. HRMS (ESI): m/z calculated for
C23H30O2N5 + H+: 408.2394 [M + H]+, found: 408.2391. HPLC purity (254 nm): 98.00%.
* signal is overlapping with residual solvent peak.
4.7. In Vitro DNA Gyrase Inhibitory Activity
A Gyrase Supercoiling High Throughput Plate assay kit, purchased from Inspiralis
(Norwich. UK) was used to determine IC50 values of compounds on S. aureus and E. coli.
The assay was performed on black streptavidin-coated 96-well microtiter plates (Thermo
Scientific Pierce), starting by wells rehydration with supplied wash buffer (20 mM Tris-HCl
(pH 7.6), 137 mM NaCl, 0.005% (w/v) BSA, 0.05% (v/v) Tween 20). Biotinylated oligonu-
cleotide in wash buffer was immobilized onto the wells, and the excess oligonucleotide was
washed off with wash buffer and ultrapure water. 1.5 U of S. aureus or E. coli DNA gyrase
enzyme was incubated together with 0.75
the presence of 3
L inhibitor solution in 10% DMSO and 0.008% Tween 20 at 37 ◦C for
30 min in a final reaction volume of 30 L in buffer (40 mM HEPES, KOH (pH 7.6), 10 mM
µg of relaxed pNO1 plasmid as a substrate in
µ
µ
magnesium acetate, 10 mM DTT, 2 mM ATP, 500 mM potassium glutamate, 0.05 mg/mL
albumin for S. aureus and 35 mM Tris-HCl (pH 7.5), 24 mM KCl, 4 mM MgCl2, 2 mM
DTT, 1.8 mM spermidine, 1 mM ATP, 6.5% (w/v) glycerol, and 0.1 mg/mL albumin for
E. coli). The reactions were stopped by adding the TF buffer (50 mM NaOAc (pH 4.7),
50 mM NaCl, and 50 mM MgCl2) to allow the triplex formation (biotin-oligonucleotide-
plasmid) for additional 30 min. Afterward, the unbound plasmid was washed off using
TF buffer, and the solution of Promega Diamond dye in T10 buffer (10 mM Tris-HCl (pH
8.0), 1 mM EDTA) was added. 10 min after, the solution was mixed and the fluorescence
was read using Tecan Fluorimeter (excitation, 495 nm; emission, 537 nm). Preliminary
screening was performed at four inhibitor concentrations of 100, 10, 1, and 0.1
IC50 values were determined at seven inhibitor concentrations for the compounds that
showed residual enzyme activity at concentration 100 M less than 50%, whereas the
other compounds were noted as inactive (IC50 > 100 M). The concentration of inhibitor
µM. The
µ
µ
where the residual activity of the enzyme is 50% (IC50) was calculated using nonlinear
regression based fitting of inhibition curves using log [inhibitor] versus response-variable
slope (four parameters)–symmetrical equation in GraphPad Prism 6.0 software (GraphPad
Software, La Jolla, CA, USA, www.graphpad.com, accessed on 13 July 2021). IC50 value
represents the mean of two to four independent measurements. Ciprofloxacin was used as
a positive control, showing IC50 93.65 µM and 0.44 µM for S. aureus and E. coli, respectively.
IC50 determination of compounds 14 and 15 was performed by Inspiralis (Norwich, UK)
using their gel-based S. aureus Gyrase Supercoiling assay, due to insufficient sensitivity of
fluorescence-based High Throughput Supercoiling assay and is described elsewhere [13].
4.8. In Vitro Topo IV Inhibitory Activity
A Topo IV Relaxation High Throughput Plate assay kit, purchased from Inspiralis
(Norwich. UK) was used to determine IC50 values of compounds on S. aureus and E. coli.
The assay was performed on black streptavidin-coated 96-well microtiter plates (Thermo
Scientific Pierce, Thermo Fisher Scientific, Waltham, MA, USA), starting by wells rehydra-
tion with supplied wash buffer (20 mM Tris-HCl (pH 7.6), 137 mM NaCl, 0.005% (w/v) BSA,

Antibiotics 2021, 10, 862
26 of 30
0.05% (v/v) Tween 20). Biotinylated oligonucleotide in wash buffer was immobilized onto
the wells, and the excess oligonucleotide was washed off with wash buffer and ultrapure
water. 1.5 U of S. aureus or E. coli DNA gyrase enzyme was incubated together with 0.75
of supercoiled pNO1 plasmid as a substrate in the presence of 3
10% DMSO and 0.008% Tween 20 at 37 ◦C for 30 min in a final reaction volume of 30
µ
g
µ
L inhibitor solution in
µL
in dilution buffer (50 mM Tris-HCl (pH 7.5), 1 mM DTT, 1 mM EDTA, 40% (w/v) glycerol
for S. aureus and 40 mM HEPES, KOH (pH 7.6), 100 mM potassium glutamate, 1 mM DTT,
1 mM EDTA, and 40% (v/v) glycerol for E. coli). The reactions were stopped by adding the
TF buffer (50 mM NaOAc (pH 4.7), 50 mM NaCl, and 50 mM MgCl2) to allow the triplex
formation (biotin-oligonucleotide-plasmid) for additional 30 min. Afterward, the unbound
plasmid was washed off using TF buffer, and the solution of Promega Diamond dye in T10
buffer (10 mM Tris-HCl (pH 8.0), 1 mM EDTA) was added. Ten minutes after, the solution
was mixed and the fluorescence was read using Tecan Fluorimeter (excitation, 495 nm;
emission, 537 nm). Preliminary screening was performed at two inhibitor concentrations
of 100 and 1
compounds that showed residual enzyme activity at concentration 100
whereas the other compounds were noted as inactive (IC50 > 100 M). The concentration
µM. The IC50 values were determined at seven inhibitor concentrations for the
µM less than 50%,
µ
of inhibitor where the residual activity of the enzyme is 50% (IC50) was calculated using
nonlinear regression based fitting of inhibition curves using log [inhibitor] versus response-
variable slope (four parameters)–symmetrical equation in GraphPad Prism 6.0 software
(GraphPad Software, La Jolla, CA, USA, www.graphpad.com, accessed on 13 July 2021).
IC50 value represents the mean of two to four independent measurements. Ciprofloxacin
was used as a positive control, showing IC50 26.25
respectively.
µM and 8.07 µM for S. aureus and E. coli,
4.9. Human Topo II Selectivity Evaluation
A Human topoisomerase II Alpha Relaxation High Throughput Plate assay kit, pur-
chased from Inspiralis (Norwich, UK) was used to evaluate the selectivity of compounds
for the bacterial enzymes over the homologous human topoisomerase II. The assay was
performed on black streptavidin-coated 96-well microtiter plates (Thermo Scientific Pierce),
starting by wells rehydration with supplied wash buffer (20 mM Tris-HCl (pH 7.6), 137 mM
NaCl, 0.005% (w/v) BSA, 0.05% (v/v) Tween 20). Biotinylated oligonucleotide in wash
buffer was immobilized onto the wells and the excess oligonucleotide was washed off
with wash buffer. 1.5 U of Human topoisomerase II
with 0.75 g of supercoiled pNO1 plasmid as a substrate in the presence of 3
solution in 10% DMSO and 0.008% Tween 20, and 1
α
enzyme was incubated together
L inhibitor
L of 30 mM ATP, at 37 C for 30 min in
L in buffer (50 mM Tris-HCl (pH 7.5), 125 mM NaCl, 10 mM
g/mL albumin). The reactions were stopped by adding the
µ
µ
◦
µ
a final reaction volume of 30
MgCl2, 5 mM DTT, and 100
µ
µ
TF buffer (50 mM NaOAc (pH 4.7), 50 mM NaCl, and 50 mM MgCl2) to allow the triplex
formation (biotin-oligonucleotide-plasmid) for additional 30 min. Afterward, the unbound
plasmid was washed off using TF buffer, and the solution of Promega Diamond dye in T10
buffer (10 mM Tris-HCl (pH 8.0), 1 mM EDTA) was added. Ten minutes after, the solution
was mixed and the fluorescence was read using Tecan Fluorimeter (excitation, 495 nm;
emission, 537 nm). Screening was performed at inhibitor concentrations of 10
µM. Raw
data were converted to mean SD percentage of residual activity of the enzyme obtained
±
by two independent measurements.
4.10. Antimicrobial Susceptibility Testing
Antimicrobial testing was carried out by the broth microdilution method in 96-well
plate format following the Clinical and Laboratory Standards Institute (CLSI) guide-
lines [40] and European Committee on Antimicrobial Susceptibility Testing (EUCAST)
recommendations [41]. Bacterial suspension of specific bacterial strain equivalent to 0.5
McFarland turbidity standard was diluted with cation-adjusted Mueller Hinton broth with
TES (Thermo Fisher Diagnostic, Landsmeer, The Netherlands), to obtain a final inoculum

Antibiotics 2021, 10, 862
27 of 30
of 105 CFU/mL. Compounds dissolved in DMSO and inoculum were mixed together and
incubated for 18–20 h at 35 ◦C. Suspensions of Mycobacterium sp. were prepared into a
tube of Middelbrook 7H9 broth supplemented with 5% OADC (Thermo Scientific, Trek
Diagnostic Systems, West Sussex, UK) and incubated for at least 10 days. After incubation
the minimal inhibitory concentration (MIC) values were determined by visual inspection
as the lowest dilution of compounds showing no turbidity. Tetracycline was used as a
positive control on every assay plate.
4.11. Metabolic Activity Assessment
Cells of the HepG2 line (ATCC® HB-8065 ) and HUVEC (ATCC® CRL-1730
™ ™) were
cultured in Dulbecco’s Modified Eagle Medium (DMEM; Sigma-Aldrich, St. Louis, MO,
USA) supplemented with 10% heat-deactivated FBS (Gibco, Grand Island, NE, USA), 2 mM
L-glutamine, 100 U/mL penicillin, 100 µg/mL streptomycin (all from Sigma-Aldrich, St.
Louis, MO, USA) in a humidified chamber at 37 ◦C and 5% CO2. The experiments were
carried out on passages 7–9 for both cell lines. The tested compounds were dissolved
in DMSO and further diluted in culture medium to a desired final concentration. Cells
were seeded into 96-well plates at 8
to attach onto the wells. Cells were treated with 1
interest or corresponding vehicle as control. For IC50 determination, cells were seeded into
96-well plates at 5 L/well) and after attachment treated with seven
104 cells/mL (100
×
104 cells/mL (100
M and 50
µ
L/well) and incubated for 24 h
µ
µM of each compound of
×
µ
different concentration of the selected compounds or corresponding vehicle as control. The
metabolic activity was assessed after 72 h treatment using the CellTiter 96 Aqueous One
Solution Cell Proliferation Assay (Promega, Madison, WI, USA), in accordance with the
manufacturer’s instructions [42]. The absorbance was measured at 492 nm on an automated
microplate reader BioTek Synergy HT (BioTek Instruments, Inc., Winooski, VT, USA). Data
are presented as the percentage of metabolic activity of control cells stimulated with vehicle
or as IC50 value (mean
duplicate or triplicate. IC50 values were calculated by a non-linear regression model using
Graph Pad Prism 7 software.
±
SD) from two independent experiments, each conducted in
4.12. Cardiovascular hERG Inhibition
Cardiovascular safety profile was entirely outsourced. hERG screening was performed
at TCG Lifesciences Pvt. Ltd., Kolkata, West Bengal, India with FP-based hERG binding
assay (Invitrogen kit).
5. Patents
Antibacterials based on monocyclic fragments coupled to aminopiperidine naphthyri-
dine scaffold: WO 2020169593 A1. Munich: European Patent Organisation (EPO).
Antibacterials based on monocyclic fragments coupled to aminopiperidine naphthyri-
dine scaffold: LU101131. Luxembourg: Office de la propriété intellectuelle Ministère de
l0Économie.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/
10.3390/antibiotics10070862/s1, Figure S1: Predicted binding modes of compound 13
the S. aureus and E. coli DNA gyrase NBTIs binding site, Figure S2: Predicted binding modes of
compound 13 15 within the S. aureus and E. coli Topo IV NBTIs binding site, Figure S3: Differences
–15 within
–
in amino acid residues constituting NBTIs’ binding pocket of DNA gyrase and Topo IV, Scheme
S1: Synthesis of 2-phenyl-2H-1,2,3-triazole-4-carbaldehyde (23), NMR spectra, Table S1: The HPLC
method parameters, equilibrium solubility at pH 6.8 and PAMPA permeability of our NBTIs and
reference compounds, Table S2: Comparison of S. aureus DNA gyrase IC50 and the lengths of RHS
fragments for our previously published compound 13 and known NBTIs, Table S3: Selected templates
for homology modelling, Table S4. Minimal inhibitory concentration (MIC) in molar, Table S5. The
metabolic activity of cells at 1
µM and 50
µM of tested compound represented as mean percent
±
SD.
Author Contributions: Conceptualization, A.K., M.A. and N.M.; methodology, A.K., M.K., M.H.,
M.W., I.Z., J.T. and S.Ž.; formal analysis, A.K., M.H., M.A. and N.M.; writing—original draft prepara-

Antibiotics 2021, 10, 862
28 of 30
tion, A.K. and M.K.; writing—review and editing, M.H., M.A. and N.M.; supervision, M.A. and N.M.;
project administration, M.A. and N.M. All authors have read and agreed to the published version of
the manuscript.
Funding: This research was funded by Slovenian Research Agency grant numbers P1-0017 and
P1-0208.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.
Acknowledgments: The authors thank Stane Pajk (Faculty of Pharmacy, University of Ljubljana, SI)
for performing HPLC analysis as well as Tilen Huzjak for his assistance with synthesis.
Conflicts of Interest: The authors declare no conflict of interest.
References
1.
2.
3.
Rice, L.B. Federal funding for the study of antimicrobial resistance in nosocomial pathogens: No ESKAPE. J. Infect. Dis. 2008, 197,
1079–1081. [CrossRef]
Centers for Disease Control and Prevention. Available online: https://www.cdc.gov/drugresistance/pdf/ar-threats-2013-508
.pdf (accessed on 29 March 2021).
World Health Organization. Available online: https://www.who.int/news-room/detail/27-02-2017-who-publishes-list-of-
bacteria-for-which-new-antibiotics-are-urgently-needed (accessed on 29 March 2021).
4.
5.
Fair, R.J.; Tor, Y. Antibiotics and bacterial resistance in the 21st century. Perspect. Medicin. Chem. 2014, 6, 25–64. [CrossRef]
Centers for Disease Control and Prevention. Available online: https://www.cdc.gov/drugresistance/pdf/threats-report/2019
-ar-threats-report-508.pdf (accessed on 29 March 2021).
6.
Levasseur, P.; Delachaume, C.; Lowther, J.; Hodgson, J. Minimum Inhibitory Concentrations (MIC) And Mutation Prevention
Concentrations (MPC) Of NXL101, A Novel Topoisomerase IV Inhibitor, Against Staphylococcus aureus Including Multi-resistant
Strains. In Proceedings of the 45th Annual Interscience Conference on Antimicrobial Agents and Chemotherapy, Washington,
DC, USA, 16–19 December 2005; American Society for Microbiology: Washington, DC, USA, 2005; p. 184.
Gomez, L.; Hack, M.D.; Wu, J.; Wiener, J.J.M.; Venkatesan, H.; Santillaán, A., Jr.; Pippel, D.J.; Mani, N.; Morrow, B.J.; Motley, S.T.;
et al. Novel pyrazolederivates as potent inhibitors of type II topoisomerases. Part 1: Synthesis and preliminary SAR analysis.
Bioorg. Med. Chem. Lett. 2007, 17, 2723–2727. [PubMed]
Wiener, J.J.M.; Gomez, L.; Venkatesan, H.; Santillaán, A., Jr.; Allison, B.D.; Schwarz, K.L.; Shinde, S.; Tang, L.; Hack, M.D.; Morrow,
B.J.; et al. Tetrahydroindazole inhibitors of bacterial type II topoisomerases. Part 2: SAR development and potency against
multidrug resistant strains. Bioorg. Med. Chem. Lett. 2007, 17, 2718–2722.
7.
8.
9.
Bax, B.D.; Chan, P.F.; Eggleston, D.S.; Fosberry, A.; Gentry, D.R.; Gorrec, F.; Giordano, I.; Hann, M.M.; Hennessy, A.; Hibbs, M.;
et al. Type IIA topoisomerase inhibition by a new class of antibacterial agents. Nature 2010, 466, 935–940. [CrossRef]
10. Kolaricˇ, A.; Anderluh, M.; Minovski, N. Two Decades of Successful SAR-Grounded Stories of the Novel Bacterial Topoisomerase
Inhibitors (NBTIs). J. Med. Chem. 2020, 63, 5664–5674. [CrossRef]
11. Black, M.T.; Stachyra, T.; Platel, D.; Girard, A.-M.; Claudon, M.; Bruneau, J.-M.; Miossec, C. Mechanism of action of the antibiotic
NXL101, a novel non-fluoroquinolone inhibitor of bacterial type II topoisomerases. Antimicrob. Agents Chemother. 2008, 52,
3339–3349. [CrossRef]
12. Singh, S.B.; Kaelin, D.E.; Wu, J.; Miesel, L.; Tan, C.M.; Meinke, P.T.; Olsen, D.; Lagrutta, A.; Bradley, P.; Lu, J.; et al.
Oxabicyclooctane-linked novel bacterial topoisomerase inhibitors as broad spectrum antibacterial agents. ACS Med. Chem. Lett.
2014, 5, 609–614. [CrossRef] [PubMed]
13. Kolaricˇ, A.; Germe, T.; Hrast, M.; Stevenson, C.E.M.; Lawson, D.M.; Burton, N.P.; Vörös, J.; Maxwell, A.; Minovski, N.; Anderluh,
M. Potent DNA gyrase inhibitors bind asymmetrically to their target using symmetrical bifurcated halogen bonds. Nat. Commun.
2021, 12, 1–13. [CrossRef]
14. Bax, B.D.; Murshudov, G.; Maxwell, A.; Germe, T. DNA topoisomerase inhibitors: Trapping a DNA-cleaving machine in motion.
J. Mol. Biol. 2019, 431, 3427–3449. [CrossRef]
15. Kolariccˇ, A.; Minovski, N. Novel bacterial topoisomerase inhibitors: Challenges and perspectives in reducing hERG toxicity.
Future Med. Chem. 2018, 10, 2241–2244. [CrossRef]
16. Li, L.; Okumu, A.A.; Nolan, S.; English, A.; Vibhute, S.; Lu, Y.; Hervert-Thomas, K.; Seffernick, J.T.; Azap, L.; Cole, S.L.; et al.
1,3-Dioxane-linked bacterial topoisomerase inhibitors with enhanced antibacterial activity and reduced hERG inhibition. ACS
Infect. Dis. 2019, 5, 1115–1128. [CrossRef]
17. Okumu, A.; Lu, Y.; Dellos-Nolan, S.; Papa, J.L.; Koci, B.; Cockroft, N.T.; Gallucci, J.; Wozniak, D.J.; Yalowich, J.C.; Mitton-Fry, M.J.
Novel bacterial topoisomerase inhibitors derived from isomannide. Eur. J. Med. Chem. 2020, 199, 112324. [CrossRef] [PubMed]
18. Kolariccˇ, A.; Novak, D.; Weiss, M.; Hrast, M.; Zdovc, I.; Anderluh, M.; Minovski, N. Cyclohexyl amide-based novel bacterial
topoisomerase inhibitors with prospective GyrA-binding fragments. Future Med. Chem. 2019, 11, 935–945. [CrossRef] [PubMed]

Antibiotics 2021, 10, 862
29 of 30
19. Kolaric, A.; Minovski, N. Structure-based design of novel combinatorially generated NBTIs as potential DNA gyrase inhibitors
against various Staphylococcus aureus mutant strains. Mol. BioSyst. 2017, 13, 1406–1420. [CrossRef]
20. Surivet, J.P.; Zumbrunn, C.; Bruyeère, T.; Bur, D.; Kohl, C.; Locher, H.H.; Seiler, P.; Ertel, E.A.; Hess, P.; Enderlin-Paput, M.; et al.
Synthesis and characterization of tetrahydropyranbased bacterial topoisomerase inhibitors with antibacterial activity against
Gram-negative bacteria. J. Med. Chem. 2017, 60, 3776–3794. [CrossRef]
21. Rosenberger, J.; Butler, J.; Muenster, U.; Dressman, J. Application of a refined developability classification system. J. Pharm. Sci.
2019, 108, 1090–1100. [CrossRef] [PubMed]
22. U.S. Food and Drug Administration. Available online: https://www.fda.gov/media/70963/download (accessed on 29 March
2021).
23. Surivet, J.P.; Zumbrunn, C.; Rueedi, G.; Bur, D.; Bruyeère, T.; Locher, H.; Ritz, D.; Seiler, P.; Kohl, C.; Ertel, E.A.; et al. Novel
tetrahydropyran-based bacterial topoisomerase inhibitors with potent anti-gram positive activity and improved safety profile. J.
Med. Chem. 2015, 58, 927–942. [CrossRef]
24. Nayar, A.S.; Dougherty, T.J.; Reck, F.; Thresher, J.; Gao, N.; Shapiro, A.B.; Ehmann, D.E. Target-based resistance in Pseudomonas
aeruginosa and Escherichia coli to NBTI 5463, a novel bacterial type II topoisomerase inhibitor. Antimicrob. Agents Chemother. 2015
,
59, 331–337. [CrossRef]
25. Sirimulla, S.; Bailey, J.B.; Vegesna, R.; Narayan, M. Halogen interactions in protein–ligand complexes: Implications of halogen
bonding for rational drug design. J. Chem. Inf. Model. 2013, 53, 2781–2791. [CrossRef]
26. Hopkins, A.L.; Keserü, G.M.; Leeson, P.D.; Rees, D.C.; Reynolds, C.H. The role of ligand efficiency metrics in drug discovery. Nat.
Rew. 2014, 13, 105–121. [CrossRef]
27. Silhavy, T.J.; Kahne, D.; Walker, S. The bacterial cell envelope. Cold Spring Harb. Perspect Biol. 2010, 2, a000414. [CrossRef]
28. Nikaido, H. Molecular basis of bacterial outer membrane permeability revisited. Microbiol. Mol. Biol. Rev. 2003, 67, 593–656.
[CrossRef]
29. Richter, M.F.; Drown, B.S.; Riley, A.P.; Garcia, A.; Shirai, T.; Svec, R.L.; Hergenrother, P.J. Predictive compound accumulation rules
yield a broad-spectrum antibiotic. Nature 2017, 545, 299. [CrossRef] [PubMed]
30. Manchester, J.I.; Dussault, D.D.; Rose, J.A.; Boriack-Sjodin, P.A.; Uria-Nickelsen, M.; Ioannidis, G.; Bist, S.; Fleming, P.; Hull, K.G.
Discovery of a novel azaindole class of antibacterial agents targeting the ATPase domains of DNA gyrase and topoisomerase IV.
Bioorg. Med. Chem. Lett. 2012, 22, 5150–5156. [CrossRef]
31. Bissachi, G.S.; Manchester, J.I. A new-class antibacterial—almost. Lessons in drug discovery and development: A critical analysis
of more than 50 years of effort toward ATPase inhibitors of DNA gyrase and topoisomerase IV. ACS Infect. Dis. 2015, 1, 4–41.
32. Delcour, A.H. Outer membrane permeability and antibiotic resistance. Biochim. Biophys. Acta 2009, 1794, 808–816. [CrossRef]
33. Jang, S. Multidrug efflux pumps in Staphylococcus aureus and their clinical implications. J. Microbiol. 2016, 54, 1–8. [CrossRef]
34. Fullam, E.; Young, R.J. Physicochemical properties and Mycobacterium tuberculosis transporters: Keys to efficacious antitubercu-
lar drugs? RSC Med. Chem. 2021, 12, 43–56. [CrossRef]
35. Nikaido, H. Multidrug efflux pumps of gram-negative bacteria. J. Bacteriol. 1996, 178, 5853–5859. [CrossRef] [PubMed]
36. Chen, X.; Murawski, A.; Patel, K.; Crespi, C.; Balimane, P. A novel design of artificial membrane for improving the PAMPA model.
Pharm. Res. 2008, 25, 1511–1520. [CrossRef] [PubMed]
37. Jones, G.; Willett, P.; Glen, R.C.; Leach, A.R.; Taylo, R. Development and validation of a genetic algorithm for flexible docking. J.
Mol. Biol. 1997, 267, 727–748. [CrossRef] [PubMed]
38. Verdonk, M.L.; Cole, J.C.; Hartshorn, M.J.; Murray, C.W.; Taylor, R.D. Improved protein-ligand docking using GOLD. Proteins
Struct. Funct. Bioinf. 2003, 52, 609–623. [CrossRef]
39. Gonzaga, D.; Senger, M.R.; da Silva, C.; Ferreira, V.F.; Silva, F.P., Jr. 1-Phenyl-1H- and 2-phenyl-2H-1,2,3-triazol derivatives:
Design, synthesis and inhibitory effect on alpha-glycosidases. Eur. J. Med. Chem. 2014, 74, 461–476. [CrossRef]
40. Patel, J.B.; Cockerill, F.R.; Badfort, P.A.; Elioppoulos, G.M.; Hindler, J.A.; Jenkins, S.G.; Lewis, J.S.; Limbago, B.; Miller, L.A.;
Nicolau, D.P.; et al. Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria that Grow Aerobically, Approved Standard,
10th ed.; CLSI Document M07-A10; Clinical and Laboratory Standards Institute: Wayne, PA, USA, 2015; Available online:
https://clsi.org/media/1632/m07a10_sample.pdf (accessed on 13 July 2021).

Antibiotics 2021, 10, 862
30 of 30
41. The European Committee on Antimicrobial Susceptibility Testing. Breakpoint Tables for Interpretation of MICs and Zone
Diameters. Version 5.0. 2015. Available online: www.eucast.org (accessed on 13 July 2021).
42. Promega Corporation. Technical Bulletin: CellTiter 96®AQueous One Solution Cell Proliferation Assay; Promega: Madison, WI,
USA, 2001; Available online: https://www.promega.com/-/media/files/resources/protocols/technical-bulletins/0/celltiter-96
-aqueous-one-solution-cell-proliferation-assay-system-protocol.pdf (accessed on 13 July 2021).