Homodimeric Tobramycin Adjuvant Repurposes Novobiocin as an Effective Antibacterial Agent against Gram-Negative Bacteria

Article abstract of DOI:10.1021/acs.jmedchem.9b00876

Low permeability across the outer membrane is a major reason why most antibiotics are ineffective against Gram-negative bacteria. Agents that permeabilize the outer membrane are typically toxic at their effective concentrations. Here, we report the develo

Full text of DOI:10.1021/acs.jmedchem.9b00876

Subscriber access provided by McMaster University Library
Article
Homodimeric Tobramycin Adjuvant Repurposes Novobiocin as
an Effective Antibacterial Agent against Gram-negative Bacteria
Temilolu Idowu, Derek Ammeter, Heather Rossong, George G Zhanel, and Frank M. Schweizer
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b00876 • Publication Date (Web): 26 Sep 2019
Downloaded from pubs.acs.org on September 30, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 51
Journal of Medicinal Chemistry
1
2
3
Homodimeric Tobramycin Adjuvant Repurposes Novobiocin as an Effective Antibacterial Agent
against Gram-negative Bacteria
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
†
†
†
‡
,†,‡
Temilolu Idowu, Derek Ammeter, Heather Rossong, George G. Zhanel, and Frank Schweizer*
†Department of Chemistry, University of Manitoba, Winnipeg, MB R3T 2N2, Canada
‡Department of Medical Microbiology/Infectious Diseases, University of Manitoba, Winnipeg, MB R3T
R9, Canada
1
KEYWORDS: adjuvant, antibacterial, antimicrobial chemotherapy, Galleria mellonella, homodimer,
novobiocin, outer membrane, synergy, tobramycin
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ABSTRACT
Low permeability across the outer membrane is a major reason why most antibiotics are ineffective
against Gram-negative bacteria. Agents that permeabilize the outer membrane are typically toxic at their
effective concentrations. Here, we report the development of a broad-spectrum homodimeric tobramycin
adjuvant that is non-toxic and more potent than the gold standard permeabilizing agent, polymyxin B
nonapeptide. In pilot studies, the adjuvant confers potent bactericidal activity on novobiocin against
Gram-negative bacteria, including carbapenem-resistant and colistin-resistant strains bearing plasmid-
borne mcr-1 genes. Resistance development to the combination was significantly reduced, relative to
novobiocin alone, and there was no induction of cross-resistance to other antibiotics, including the
gyrase-acting fluoroquinolones. Tobramycin homodimer may allow the use of lower doses of
novobiocin, overcoming its twin-problem of efficacy and toxicity.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
ACS Paragon Plus Environment

Page 3 of 51
Journal of Medicinal Chemistry
1
2
3
INTRODUCTION
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
The problem of antibacterial drug resistance is complex and multifaceted, and antibiotic drug
1
development is being massively outpaced by emerging resistance against available antibiotics. Gram-
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
negative bacteria are more difficult to eradicate because of their dual membrane topology and over
expressed efflux pumps of broad substrate specificities.2,3 The outer membrane (OM) of Gram-negative
bacteria, which serves as a barrier to the permeation of potentially noxious molecules, including
antibiotics, is composed of lipopolysaccharide (LPS) on the outer leaflet and phospholipid on the inner
leaflet. The mechanism by which bacteria assemble this well-organized protective barrier has been well
4
elucidated. To breach the OM, agents that interact directly with LPS stability (e.g. polybasic molecules
5
such as polymyxins, aminoglycosides (AGs), and chelating agents such as EDTA) or indirectly with
mechanisms that synthesize, assemble, and/or transport the LPS (e.g. LpxC inhibitors, novobiocin,
6
,7
etc.) have been investigated as potential resistance breakers for Gram-negative bacteria.
Polybasic compounds destabilize the OM by displacing divalent metals that cross-bridge
adjacent phosphate groups attached to LPS core sugar, thereby altering the well-ordered polyelectrolyte
barrier. Truncation and/or modification of the polymyxin class of drugs has resulted into adjuvants such
8
9
as polymyxin B nonapeptide (PMBN) and SPR741 , where intrinsic antimicrobial activity was
decoupled from their OM-destabilizing properties. Whereas PMBN seems to potentiate several OM-
8
impermeable antibiotics against various colistin-susceptible Gram-negative bacteria, the activity of
9
SPR741 excludes antipseudomonal effects. PMBN was shown to be generally less toxic than
1
0
polymyxin B, but it causes similar proximal renal tubular injury in male rats. Similarly, conjugation
and site-specific modification of AGs has resulted into scaffolds that lose the primary ribosomal
properties of AGs but adopts an enhanced membrane effect.11–13 Among membrane-acting AGs,
tobramycin-based adjuvants seem to be more effective against Pseudomonas aeruginosa than other
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
14
Gram-negative bacteria. A major problem associated with the use of AGs is their propensity to cause
irreversible hearing loss, an effect linked to the lack of precise selectivity for prokaryotic ribosomes.15–17
Synthetic AG analogs with lower bacterial and human mitochondrial ribosome specificities have been
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
18
shown to exhibit reduced ototoxic potentials in cochlear explants, in culture and in guinea pig. Hence,
non-ribosomal AGs may exhibit lower idiosyncratic toxicities and drug-induced hearing loss.
Consequently, we sought to expand the spectrum of activity of tobramycin-derived adjuvants beyond P.
aeruginosa by dimerizing the core scaffold of an amphiphilic tobramycin, such that the hydrophobic
domain is sandwiched between two identical polar heads (Figure 1). This was hypothesized to overcome
the hemolytic problem of classic cationic amphiphiles.19–22
As a proof of concept, we investigated the possibility of repurposing novobiocin against
clinically-relevant Gram-negative bacteria using the newly synthesized adjuvant. Novobiocin (Figure 1)
is an orally active dihydroxy-glycosylated aminocoumarin antibiotic that inhibits DNA gyrase by
7
binding the ATP-binding site in the ATPase subunit. In 2011, the oral form of novobiocin (novobiocin
sodium capsule, 250 mg) was withdrawn from US market for “reasons of safety or effectiveness”.23
However, recent pharmacokinetic trials in non-infected subjects (Phase I and II studies) have
demonstrated novobiocin plasma concentration of 150 µM (~ 90 – 100 mg/L) for 24 h after a 5.5 g dose,
with no serious toxicities.24,25 Novobiocin displays limited activity against Gram-negative bacteria
(
MICs far higher than clinically achievable serum concentrations), even though their GyrB is sensitive
to the antibiotic, due to the LPS-containing OM that act as a permeability barrier. PMBN had previously
been investigated in combination with novobiocin to increase penetrance,26,27 but effective
concentrations and dose-limiting toxicities are of serious concern. Herein, we report the development of
a non-toxic broad-spectrum antibiotic adjuvant that is more potent than PMBN and restores potent
GyrB-dependent activity of novobiocin against multidrug (MDR) and extensively drug-resistant (XDR)
4
ACS Paragon Plus Environment

Page 5 of 51
Journal of Medicinal Chemistry
1
2
3
Gram-negative bacteria. Concentrations as low as 0.25 g/ml (0.1 M) of the adjuvant were enough to
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
cause a measurable effect and the addition of ≤ 7.1 M resulted in the attainment of MICs levels below
clinically achievable plasma concentration of novobiocin in all 28 isolates studied. We also provide
insights into the mechanism of action and resistance development to this combination.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 1. Structures of newly synthesized and reference compounds. Compounds 1–3 are tobramycin
homodimers conjugated at the C-5 position of tobramycin with different tether lengths, compounds 4
and 5 are fragments of lead structure 1, and compound 6 is an aglycone derivative of novobiocin.
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
RESULTS AND DISCUSSION
Design and Synthesis
The design of tobramycin homodimers 1–3 was guided by previous SAR.13 Amphiphilic
tobramycins with lipophilic groups at the 5-OH of deoxystreptamine (ring I; Figure 1) have been shown
to lose ribosomal activities but retain the ability to permeabilize the OM.13,14 Dimerization of ribosome-
targeting antibiotics has also been shown to result in poor inhibitors of in vitro protein translation.28
Hence, to prepare non-ribosomal amphiphilic-like tobramycin homodimers with potentially broad-
spectrum OM permeabilizing properties, we dimerized two fragments of short-chain amphiphilic
tobramycins ligated at the 4,6-disubstituted 2-deoxystreptamine via a copper(1)-catalyzed azide-alkyne
cycloaddition reaction (Scheme 1). This afforded regioselective 1,4-disubstituted 1,2,3-triazole products
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
–3. Analogs with different tether length were synthesized to investigate the optimal spatial separation
between the two domains while compounds 4 and 5 were prepared to study the SAR of the lead
compound 1. The full synthetic strategy for preparing compounds 1–5 is outlined in Scheme 1.
Compound 6 an aglycone derivative of novobiocin, was prepared to investigate the role of the L-noviose
sugar on the gyrase activity of novobiocin (Scheme 2).
Chemical Synthesis of Tobramycin Homodimers (1–3), Fragments 4–5 and Novobiocin Aglycone
(6).
The two amphiphilic tobramycin domains 4 and 5 were prepared following previously reported
4
9
protocol. Tobramycin 7 was purchased from a commercial source and the amino groups were first
protected using di-tert-butyl dicarbonate (Boc anhydride), followed by silylation of the N-Boc-
tobramycin intermediate with excess TBDMSCl to afford a partially protected derivative 8 with free OH
at the C-5 position of the deoxystreptamine ring. In the presence of a phase-transfer catalyst (TBAHS), 8
6
ACS Paragon Plus Environment

Page 7 of 51
Journal of Medicinal Chemistry
1
2
3
was alkylated with 1,n-dibromoalkane (n = 4, 6, 8) in toluene to afford alkylated TBDMS-Boc-protected
tobramycin intermediates 9a-c. Similarly, alkylation of 8 with iodohexyne under the same conditions
followed by TBDMS deprotection afforded 11. The terminal bromine of 9a-c was then displaced by an
azido nucleophile under anhydrous condition and the TBDMS protecting groups were deblocked – using
TBAF – to give compounds 10a-c, followed by Boc deprotection (using TFA) to give compound 5.
Similarly, compound 4 was prepared by deblocking compound 11 with TFA. Dimerization of
tobramycin was achieved by conjoining compounds 10a-c and 11 via “click chemistry” to afford 12a-c.
Global deprotection of Boc-protecting groups using TFA afforded the final compounds 1–3 (Scheme 1).
Compound 6 was prepared by exposing novobiocin sodium salt to a highly basic condition (5M NaOH)
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
at
80
C
(Scheme
2).
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
Scheme 1. Synthesis of tobramycin homodimers 1–3 and fragments 4-5. Conjugates differ in the length of tether.
8
ACS Paragon Plus Environment

Page 9 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 2. Synthesis of novobiocin aglycone 6
Susceptibility and Toxicity Screening
Susceptibilities of different Gram-positive and Gram-negative bacteria to the newly synthesized
molecules 1–3 were determined and compared to the progenitor molecule tobramycin. The lack of
activity of compounds 1–3 (MIC ≥ 16 g/ml, Table S1) against a panel of organisms, relative to
tobramycin, is consistent with loss of ribosomal binding. To investigate toxicity, we tested and found
that compounds 1–3 were: i) non-hemolytic against porcine erythrocytes at 1024 g/ml, ii) non-
cytotoxic to human embryonic kidney (HEK293) and human liver carcinoma (HepG2) cells at 50 M
(>128 g/ml), and iii) non-toxic in vivo against Galleria mellonella wax moths at 200 mg/kg (Figure S1).
On the contrary, a single dose administration of 100 mg/kg colistin was toxic to G. mellonella and killed
90 % of the larvae after 96 h (Figure S1c).
Checkerboard Assay with different Classes of Antibiotics
The lack of antibacterial activity and non-toxic properties of 1–3 further encouraged us to screen
their adjuvant properties. An ideal adjuvant is a bioactive helper molecule that is inactive by itself but
can potentiate the activity of a primary antibiotic and/or delay resistance development when used in
2
9
combination. These types of molecules are less likely to select for resistance. To investigate this,
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
checkerboard assay was used to assess the interactions between compounds 1–3 and nineteen different
antibiotics (representing all major classes) against wild-type P. aeruginosa PAO1. P. aeruginosa was
selected for this initial screen because OM permeability is a major mechanism of intrinsic resistance to
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
30
31
antibiotics, and it is often regarded as a highly challenging model organism for new antibiotics.
Compounds 1–3, investigated at ≤ 7.1 M based on achievable plasma concentrations (20 – 200 M) of
aminoglycosides,32,33 exhibit concentration-dependent synergistic relationships with all antibiotics tested
against PAO1, except tobramycin, vancomycin, and colistin (Figure 2, Table S2). OM-impermeable
antibiotics (such as rifampicin, linezolid, clindamycin and novobiocin), efflux-prone antibiotics (such as
tetracyclines, fluoroquinolones, chloramphenicol, etc.), β-lactam antibiotics (such as monobactams,
carbapenems, cephalosporins, and penicillins), and fosfomycin were all potentiated by 4- to 128-fold
(Table S2). Tobramycin by itself is not synergistic with these antibiotics (Table S3). The antagonistic
relationship between compounds 1–3 and tobramycin or colistin (FICI > 4) is consistent with observed
antagonism between tobramycin and colistin at high concentrations (Figure S2). This is perhaps due to
competition for LPS binding by both polybasic molecules. The lack of potentiation of vancomycin is
consistent with other OM permeabilizing agents such as PMBN and pentamidine,26,34 where synergy is
generally more pronounced with large hydrophobic molecules (e.g. rifampicin) than with large
hydrophilic molecules (e.g. vancomycin). Compound 1 is the most potent (Table S2) and least toxic of
the three (Figure S1), hence, it was used for further studies.
10
ACS Paragon Plus Environment

Page 11 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Antibiotics
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
Synergy
No Interaction
Antagonistic
Figure 2. Interactions of compounds 1–3 (at ≤ 7.1 M) with different antibiotics against P. aeruginosa
PAO1. FICI ≤ 0.5 = Green (synergistic); FICI > 0.5 but < 1 = Yellow (no interaction); FICI > 4 = Red
(
antagonistic)
To investigate the spectrum of activity of the newly synthesized adjuvants, we examined a
combination of compound 1 (at ≤ 7.1 M) and novobiocin against MDR/XDR P. aeruginosa,
Acinetobacter baumannii, Escherichia coli, Klebsiella pneumoniae, and Enterobacter cloacae exhibiting
multiple resistance patterns (Table S4). Novobiocin, an ATPase inhibitor of bacterial DNA gyrase and
topoisomerase IV, was selected for this study because of its unique mechanism of action but lack of
meaningful activity against Gram-negative bacteria even though their GyrB is sensitive to the
antibiotic.26,27 Synergy was retained in 100% of all isolates tested (8- to 256-fold potentiation; Table 1) –
including carbapenem- and colistin-resistant clinical isolates (Table S4), isolates bearing GyrA
mutations (Table S5), and strains carrying the mobilized colistin resistance (mcr-1) gene that confer
3
5
plasmid-mediated resistance to colistin. We conclude that resistance to novobiocin in Gram-negative
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
bacteria is primarily due to OM permeability, and that resistance to fluoroquinolones (via GyrA
mutations) does not necessarily confer resistance to novobiocin. To contextualize the potency of
compound 1, we compared its ability to potentiate novobiocin to the ‘gold standard’ potentiator
molecule, PMBN, and observed that at equimolar molar concentrations of 7.1 µM, compound 1
reproducibly displayed superior activity to PMBN against colistin-resistant Gram-negative bacteria
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(Table 2, Figure S3a). Next, we investigated whether tobramycin by itself would potentiate novobiocin
against Gram-negative bacteria. We tested a combination of tobramycin and novobiocin against
tobramycin-susceptible and tobramycin-resistant Gram-negative bacteria strains and observed no
synergy (FICI ≥ 0.75) (Table S6). Other antibiotics such as minocycline, ceftazidime, rifampicin, etc.
also did not show any synergy with novobiocin, consistent with a previous chemical screen of 30,000
small molecules that identified just a handful of compounds (four hits) capable of synergizing with
3
6
novobiocin. To establish whether the synergistic relationship of compound 1 with novobiocin was
specific to Gram-negative bacteria, we assessed the combination against Gram-positive bacteria.
Whereas novobiocin displayed potent activity against MRSA ATCC 33592 (MIC, 0.125 µg/ml), MRSE
6
1589 (MIC, 0.0625 µg/ml), and E. faecium ATCC 27270 (MIC, 2 µg/ml), consistent with prior
3
7
literature, addition of compound 1 did not improve the activity of novobiocin further (Table S7). This
suggests that OM-permeabilization in Gram-negative bacteria, but not in Gram-positive, is the
predominant mechanism by which compound 1 (and PMBN) enhances the activity of novobiocin.
Notably, a high level of susceptibility to novobiocin (MIC ≤ 1 µg/ml) was attained in wild-type and
MDR A. baumannii clinical isolates in the presence of ≤ 7.1 µM of compound 1 (Table 1).
12
ACS Paragon Plus Environment

Page 13 of 51
Journal of Medicinal Chemistry
1
2
3
Table 1. Tobramycin homodimer potentiates novobiocin against Gram-negative bacteria.
Synergistic effects of novobiocin and compound 1 (at ≤ 7.1 µM, i.e. ≤ 16 µg/ml) against wild-type and
clinical isolates of Gram-negative bacteria. MIC of compound 1 is > 128 µg/ml against all strains.
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
†
‡
§
#
Wildtype; Clinical isolate; Carbapenem-resistant; *Colistin resistant; mcr-1 gene positive.
MIC (µg/ml) of Novobiocin
Bacteria
Strain
Fold Potentiation
Alone
512
512
256
512
512
16
+ Compound 1
PAO1†
4
8
128
64
2
2
59-96918‡,§
60-97103‡,§
1
256
64
262-101856‡,§
64-104354‡,§
1433‡,§,
00036‡,§
101243‡,§,
01885‡,§
14228‡,§,
8
2
8
64
P. aeruginosa
9
*
1
16
1
1024
128
1024
512
16
8
128
128
128
32
*
1
1
8
1
*
16
0.5
0.5
0.5
0.25
1
ATCC 17978†
027‡
32
8
16
0
31‡
2247‡,*
8
16
A. baumannii
9
4
16
110193‡
LAC-4‡
256
8
256
16
0.5
2
ATCC 25922†
64
32
9
4393‡,*#
128
256
128
512
256
1024
2
64
E. coli
‡
,
#
94474 *
07115‡,§
16
8
16
1
16
‡
,
1
13250 *
16
16
8
32
113254‡,*
116381‡
K. pneumoniae
16
128
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
17029‡
1024
1024
64
8
32
8
128
32
8
‡
,
E. cloacae
118568 *
‡
,
1
21187 *
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 2. Compound 1 potentiates Novobiocin better than PMBN against colistin-resistant Gram-
negative bacteria. PA = P. aeruginosa; AB = A. baumannii; KP = K. pneumoniae; EC = E. cloacae.
†
‡
Wild type; Clinical isolate.
MIC (µg/mL) of Novobiocin
Organism
Alone
512
16
+ PMBN (7.1 µM)
+ 1 (7.1 µM)
PAO1†
4
4
4
1
PA 91433‡
PA 114228‡
PA 101243‡
E. coli 94393‡
E. coli 94474‡
AB 92247‡
KP 113254‡
EC 118568‡
512
128
128
256
4
128
8
16
1
4
2
32
1
16
0.25
16
32
256
1024
64
512
To investigate SAR of tobramycin homodimers, we prepared the constituent fragments (alkyne 4
and azide 5) of compound 1, the most potent of the synthesized derivatives (Scheme 1). We also
examined the adjuvant properties of a Boc-protected derivative 12a to investigate the role of free amines
as it relates to OM destabilization. Compounds 4 and 5 did not potentiate novobiocin against P.
aeruginosa PAO1 (Table S8), suggesting that covalent linkage of both domains is critical for the
function of the adjuvant. This observation is consistent with optimal hydrophobic-charge threshold that
14
ACS Paragon Plus Environment

Page 15 of 51
Journal of Medicinal Chemistry
1
2
3
must be maintained by cationic amphiphiles in order to destabilize bacterial membranes.19–22 Similarly,
compound 12a did not potentiate novobiocin against P. aeruginosa PAO1 (Table S8), indicating that
electrostatic interactions between the positively charged amines in compound 1 and the negatively
charged phosphate residues on the OM of Gram-negative bacteria is central to the potency of the
adjuvant.
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Potentiation of Novobiocin is Independent of RND Efflux Pumps
To investigate whether efflux pumps play a role in the restoration of potent antibacterial activity
of novobiocin by tobramycin homodimers, we assessed and compared the presence of synergy in efflux-
deficient P. aeruginosa strains to wild type PAO1. Efflux mutants PAO200 and PAO750, lacking
different clinically-relevant efflux pumps that extrude different classes of antimicrobial agents, exhibited
higher level of susceptibilities to novobiocin (MIC, 32 µg/ml) than wild type PAO1 (MIC, 512 µg/ml),
suggesting that novobiocin is a substrate of the RND efflux pumps. The MIC of compound 1 was
unaffected by these pumps (Table 3). However, compound 1 (at 7.1 µM) further increased the
susceptibilities of the mutants to novobiocin and lowered its MICs in both strains from 32 µg/ml to
0
.125 µg/ml (256-fold potentiation) (Table 3). This suggests that whereas novobiocin is a substrate of
the RND pumps, its potentiation by compound 1 is independent of these pumps.
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 3. Potentiation of novobiocin in P. aeruginosa by compound 1 is independent of RND efflux
pumps. PAO1 = wild-type, PAO200 (∆mexAB-oprM) and PAO750 (∆mexAB-oprM, ∆mexCD-oprJ,
∆mexEF-oprN, ∆mexJK, ∆mexXY, and ∆opmH outer membrane) are efflux-deficient strains.38,39 MICs
are reported in g/ml.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
MIC of Novobiocin
MIC of
Strain
Fold Potentiation
+
7.1 M
Compound 1
Alone
Compound 1
PAO1
>128
>128
>128
512
32
4
128
256
256
PAO200
PAO750
0.125
0.125
32
Time-Kill Assay
To scrutinize the data generated by checkerboard assay and provide more complementary
evidence for the observed synergy, time-kill assays were performed on four different Gram-negative
bacteria. Novobiocin is predominantly bacteriostatic versus most bacteria but could also be bactericidal
against some pathogens. We investigated the kinetics of killing in LB media at fixed concentrations of
32 g/ml (50.4 M) for novobiocin and 16 g/ml (7.1 M) for compound 1 because these
concentrations (in combination) inhibit visible growth of all isolates studied in MHB media (Table 1).
For PAO1 and K. pneumoniae 116381, 32 µg/ml of novobiocin alone resulted in bacterial growth
identical to their respective controls (without drug) while a combination of novobiocin (32 µg/ml) and
compound 1 (7.1 µM) inhibited the growth of both pathogens in LB media (Figure 3). The combination
resulted in bactericidal (PAO1) and synergistic relationships against both strains after 24 h. For A.
baumannii ATCC 17978, 32 µg/ml novobiocin alone was bacteriostatic after 24 h, while a combination
of novobiocin (32 µg/ml) and compound 1 (7.1 µM) was bactericidal (> 3-Log reduction) and
16
ACS Paragon Plus Environment

Page 17 of 51
Journal of Medicinal Chemistry
1
2
3
synergistic after 9 h of incubation (Figure 3). After 24 h, the bacterial culture containing novobiocin and
compound 1 was completely sterilized, representing > 5-Log reduction from the starting inoculum. For
E. coli ATCC 25299, 32 µg/ml novobiocin alone was bactericidal after 24 h, while a combination of
novobiocin (32 µg/ml) and compound 1 (7.1 µM) exhibited bactericidal effects after 6 h of incubation
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(Figure 3). Novobiocin and compound 1 became synergistic against E. coli ATCC 25922 after 9 h of
incubation and the culture was completely sterilized within this period. Overall, the species-dependent
degree of bacterial load reduction reflects, to a large extent, the fold potentiation of novobiocin by
compound 1 in checkerboard assay.
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
A.
B.
12
1
1
2
0
8
6
4
2
0
10
8
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
-1.87 Log CFU/mL
-
3.68 Log CFU/mL
6
4
2
0
-8.68 Log CFU/mL
-9.48 Log CFU/mL
0
3
6
9
12
Time (h)
15
18 21 24
0
3
6
9
12
Time (h)
15
18
21
24
C.
D.
1
2
0
8
6
4
2
0
12
10
8
1
>
-5 Log CFU/mL
6
> -5 Log CFU/mL
4
2
>
-3 Log CFU/mL
> -2 Log CFU/mL
12
Time (h)
0
0
3
6
9
12
Time (h)
15 18 21 24
0
3
6
9
15
18
21
24
Time (h)
Control
Novobiocin (32 µg/mL)
Novobiocin (32 µg/mL) + 1 (16 µg/mL)
Figure 3. Time-kill synergy graphs. The activities of novobiocin (32 µg/ml) in combination with
compound 1 (16 µg/ml, i.e. 7.1 µM) against (A) P. aeruginosa PAO1, (B) K. pneumoniae 116381, (C) A.
baumannii ATCC 17978, (D) E. coli ATCC 25922. Red bars and numbers indicate differences in
bacterial concentrations between the starting inoculum and drug combination at 24 h. Purple bars and
numbers indicate differences in bacterial concentrations between the combination and the most active
single agent at 24 h (9 h for E. coli). The dashed lines represent the lower limit of detection. Each data
point represents an average of three independent determinations.
18
ACS Paragon Plus Environment

Page 19 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
Outer Membrane Permeabilization Assay
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
To investigate whether compound 1 enhances the uptake of novobiocin by permeabilizing the
OM, we measured the intensity of fluorescence of the nonpolar probe 1-N-phenylnaphthylamine (NPN)
using NPN uptake assay.40 An intact OM will ordinarily prevent the uptake of NPN dye which
fluoresces strongly in phospholipid environments but only weakly in an aqueous environment. We
observed that tobramycin homodimer 1 permeabilizes the OM of P. aeruginosa PAO1 in a dose-
dependent manner, reminiscent of PMBN, while tobramycin alone showed weak fluorescence (Figure 4).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
6
PMBN is a known OM permeabilizer, and our results showed that compound 1 (at 16 µg/ml) caused
an increased fluorescence of NPN relative to PMBN (8 µg/ml) at similar micromolar concentrations
(Figure 4). This observation is consistent with data from checkerboard assay (Table 2).
3
2
2
00
50
00
1 (56 µM)
(28 µM)
1 (14 µM)
(7 µM)
1
150
100
50
0
1
PMBN (14 µM)
PMBN (7 µM)
TOB (14 µM)
0
60 120 180 240 300 360
Time (secs)
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Figure 4. Outer membrane permeabilization by compound 1, polymyxin B nonapeptide (PMBN) and
tobramycin (TOB) was determined by measuring the accumulation of 1-N-phenylnaphthylamine (NPN)
in P. aeruginosa PAO1 cells. Each data point is an average of four independent determinations ± SD.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Mechanism of Resistance Study
To investigate the possible mechanisms of resistance development to the combination, we
generated some A. baumannii mutants by exposing wild-type ATCC 17978 to sub-MICs of novobiocin,
alone and in combination with compound 1. A. baumannii was used for this study because all strains
exhibited a high level of susceptibility to novobiocin + compound 1 combination (Table 1).
Tobramycin-resistant and colistin-resistant mutants were also generated to investigate the mechanism
and pattern of resistance development. Emergence of resistance to novobiocin alone was fast and high
41
level, consistent with GyrB mutation, with a 64-fold change in MIC (from 8 µg/ml to 512 µg/ml) after
just 3 days and a 256-fold change in MIC (from 8 µg/ml to 2048 µg/ml) after 5 days (Figure 5). On the
other hand, a combination of novobiocin and compound 1 (7.1 µM) resulted in a slow and steady 2-fold
increase in MIC every generation, with a 4-fold increase in MIC (from 0.5 µg/ml to 2 µg/ml) after 3
days and a 16-fold change in MIC (from 0.5 to 8.0 µg/ml) after 5 days (Figure 5). Tobramycin alone
resulted in a 512-fold increase in MIC (from 1 to 512 µg/ml) after 7 days while colistin resulted in a
1024-fold change in MIC after 7 days (Figure 5).
20
ACS Paragon Plus Environment

Page 21 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
4
096
2048
024
12
256
28
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
1
6
4
32
1
6
8
4
2
1
Tobramcycin
Novobiocin (Mutant A)
Novobiocin + 1 (Mutant C)
Colistin (Mutant B)
0
.5
.25
0.125
0
0
1
2
3
4
5
6
7
8
9
Days
Figure 5. Emergence of resistance study. Resistance acquisition during serial passaging of A. baumannii
ATCC 17978 in the presence of sub-MIC levels of antibiotics. For combination study, compound 1 was
kept constant at a concentration of 7.1 µM throughout the experiment.
To gain insights into the emerging mechanism(s) of resistance to novobiocin/compound 1
combination, we assessed the susceptibilities of all generated mutants to different antibiotics (Table 4)
and performed checkerboard assay on the novobiocin-resistant and colistin-resistant mutants A and B,
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
respectively. Exposure of A. baumannii ATCC 17978 to sub-MIC levels of tobramycin homodimer 1 for
seven serial passages did not generate tobramycin-resistant phenotype, whereas tobramycin alone did,
suggesting that tobramycin homodimers do not bind to any critical target in the bacteria. Novobiocin-
resistant mutant A (Table 4) was found to be resistant to ciprofloxacin (MIC,16 µg/ml) and colistin
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(MIC, 8 µg/ml), suggesting gyrase mutation(s) as a primary mechanism of resistance but also LPS
modification and efflux. Compound 1 was able to further potentiate novobiocin against this highly
novobiocin-resistant mutant by 16-fold (Figure S3b). Surprisingly, the generated colistin-resistant
mutant B was hyper-susceptible to all antibiotics tested, consistent with a complete loss of LPS
4
2
production that resulted in collateral sensitivity to antibiotics, or/and mutations in the PmrAB two-
4
3
42
component system. The lipid A component of LPS is critical for the activity of colistin, and the
development of such high level resistance to colistin under clinical conditions might constitute a huge
fitness cost to the pathogen. PMBN and compound 1 did not potentiate novobiocin further against this
strain (Figure S3c), suggesting that LPS interaction is critical for OM destabilization by both compounds.
Resistance to novobiocin/compound 1 combination did not confer resistance to tobramycin,
ciprofloxacin, and colistin, but conferred low level resistance to novobiocin alone (Mutant C; Table 4),
suggesting that the combination did not trigger the production of aminoglycoside modifying enzymes
(AMEs), overexpression of efflux pumps and/or LPS modifications. By comparing the MICs of
novobiocin versus mutants A and C, it is evident that compound 1 suppressed the development of
resistance to novobiocin. Overall, the mechanism of resistance development to novobiocin/compound 1
combination is consistent with gyrase mutation, not LPS modification and AMEs production, and
compound 1 significantly delayed this process.
22
ACS Paragon Plus Environment

Page 23 of 51
Journal of Medicinal Chemistry
1
2
3
Table 4. Susceptibility profiles (MIC in µg/ml) of wild-type A. baumannii ATCC 17978 versus resistant
mutants generated from seven serial passages (Day 8) of exposure to sub-MICs novobiocin (Mutant A),
colistin (Mutant B) and novobiocin + 7.1 µM compound 1 (Mutant C).
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Antibiotics
Wild-type
Mutant A Mutant B
Mutant C
Novobiocin
Tobramycin
Ciprofloxacin
Minocycline
Rifampicin
16
1
2048
1
<0.031
0.25
128
1
1
16
0.5
2
0.0625
<0.031
<0.031
1
1
0.125
2
0.125
2
Ceftazidime
Chloramphenicol
Colistin
16
32
32
8
16
64
8
64
0.063
0.5
1024
<0.031
0.031
16
Novobiocin + 1
128
Potentiation of Novobiocin by Compound 1 is dependent on Gyrase B Activity
A study has shown that synergistic relationship between novobiocin and polymyxin B is
independent of gyrase activity of novobiocin.44 To ascertain whether potentiation of novobiocin by
compound 1 is indeed based on the binding of novobiocin to the ATP-binding site in the ATPase subunit
of GyrB, we prepared novobiocin aglycone 6 (Scheme 2) lacking the L-noviose sugar of novobiocin,
4
5
which is known to make important contacts with GyrB. Consistent with a loss of gyrase activity,
potent antibacterial activity was completely lost in compound 6, even against Gram-positive organisms
that are susceptible to novobiocin (Table S9). Novobiocin aglycone 6 has been reported to lose in vitro
4
4
supercoiling activity against A. baumannii gyrase by at least 260-fold, relative to novobiocin. We
thereafter evaluated whether any synergistic relationship exists between compounds 1 and 6 against P.
aeruginosa PAO1, A. baumannii ATCC 17978, and E. coli ATCC 25922. Tobramycin homodimer 1
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
potentiates novobiocin against all of these strains (Table 1) but there was no potentiation of compound 6
against any of these pathogens (FICI > 1) (Table S10a). On the contrary, compound 6 was found to
potentiate colistin against A. baumannii ATCC 17978 and E. coli ATCC 25922 (Table S10b), similar to
a previous report.44 This suggests that whereas the synergistic relationship between colistin and
novobiocin (or compound 6) might not be gyrase-dependent, the synergistic relationship between
tobramycin homodimers and novobiocin is DNA-gyrase dependent.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
In Vivo Efficacy in Galleria mellonella Larvae
To expand the robustness of the data generated using in vitro assays, we performed preliminary
in vivo investigation on the ability of novobiocin and compound 1 to protect G. mellonella moths from
two different MDR A. baumannii infections. Despite its well-described limitations, the use of G.
mellonella larvae as preliminary in vivo model system to study host-pathogen interactions, virulence,
toxicity, and efficacy of novel compounds has been demonstrated with clinically used therapeutic
agents.46–49 We found that the larvae tolerated 200 mg/kg each of compound 1 and novobiocin alone,
and a 100 + 100 mg/kg combination of both, for more than 4 days, whereas exposure to 100 mg/kg of
colistin (single dose administration) resulted in 90 % mortality after 4 days (Figure S1c). A multiple, but
not single, dose administration of 25 mg/kg novobiocin + 50 mg/kg PMBN had previously been
2
6
reported to protect mice challenged with MDR P. aeruginosa or K. pneumoniae. In pilot studies, we
investigated the in vivo therapeutic potential of novobiocin + compound 1 (25 + 25 and 50 + 50) mg/kg
in the infection model. Untreated larvae challenged with MDR A. baumannii died within 24 h of
infection. A single dose administration of 100 mg/kg each of novobiocin alone or compound 1 alone
also resulted in 100 % mortality of infected larvae after 24 h. However, a single dose administration of
novobiocin + compound 1 (50 + 50 mg/kg) protected ~80 % of the larvae after 24 h for more than 96 h
24
ACS Paragon Plus Environment

Page 25 of 51
Journal of Medicinal Chemistry
1
2
3
(Figure 6). An approximate equimolar concentration of novobiocin + PMBN (50 + 25 mg/kg) only
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
protected ~35 % of the larvae from the same infection (Figure 6).
To ascertain that the combination therapy does not elevate toxicity against eukaryotic cells, we
assessed the toxicity of a combination of compound 1 and novobiocin against HEK293 and HepG2 cell
lines and found the combination to be non-cytotoxic (Figure S1d). These preliminary data provide a
baseline for future studies in mice.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
A) 100
PBS only
80
AB LAC-4
Novobiocin (100 mg/kg)
Compound 1 (100 mg/kg)
6
4
2
0
0
0
0
Novobiocin + Compound 1 (25 + 25 mg/kg)
Novobiocin + Compound 1 (50 + 50 mg/kg)
Novobiocin + PMBN (50 + 25 mg/kg)
0
12
24
Time (h)
36
4 98 0 96
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
B) 100
PBS only
AB 92247
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
80
Novobiocin (100 mg/kg)
Compound 1 (100 mg/kg)
6
4
0
0
Novobiocin + Compound 1 (25 + 25 mg/kg)
Novobiocin + Compound 1 (50 + 50 mg/kg)
Novobiocin + PMBN (50 + 25 mg/kg)
20
0
0
12
24
Time (h)
36
4 98 0 96
Figure 6. Preliminary efficacy studies of mono- and combination therapies in MDR Acinetobacter
baumannii (AB)-challenged Galleria mellonella wax worms. A) AB LAC-4, a multidrug-resistant
hypervirulent strain isolated from an hospital outbreak in LA county, California. B) AB 92247, a
colistin-resistant MDR clinical isolate. Survivability of the larvae was scored every 6 h for 96 h.
CONCLUSION
The problem of antimicrobial resistance is forcing a re-think and re-evaluation of our old
antibiotics, with a view to repurposing them to combat emerging threats. Antibiotic adjuvants are
becoming increasingly attractive as a strategy to address the immediate needs of effective treatment
options against MDR pathogens. We report the development of a broad-spectrum homodimeric
tobramycin adjuvant that is more potent than the gold standard potentiator molecule, PMBN, and
confers potent bactericidal activity on novobiocin against MDR/XDR Gram-negative bacteria. A.
26
ACS Paragon Plus Environment

Page 27 of 51
Journal of Medicinal Chemistry
1
2
3
baumannii isolates exhibit a high level of susceptibility to this combination, allowing the use of lower
doses of novobiocin that could consequently improve tolerability and safety. Our data and others26
suggest that intrinsic, but not acquired, resistance is the predominant mechanism of resistance to
novobiocin in Gram-negative bacteria. Emergence of resistance to novobiocin alone by A. baumannii
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
1
was spontaneous and conferred resistance to ciprofloxacin (consistent with target mutation) whereas a
combination of novobiocin and compound 1 suppressed this process and did not induce cross-resistance
to other antibiotics. Preliminary in vivo evaluation in G. mellonella wax moths did not reveal any toxic
interaction between the combination (colistin alone was toxic), but rather suggests a therapeutic
potential for this combination. Future work will involve efficacy studies in animals and optimization of
dosing regimens for each constituent molecule.
EXPERIMENTAL SECTION
Chemistry. All chemicals and reagents were purchased from Sigma-Aldrich (Oakville, ON, Canada)
except tobramycin that was purchased from AK Scientific Inc. (CA, USA). The chemicals were all used
without further purification. Air and moisture-sensitive reactions were performed with dry solvents
under nitrogen atmosphere. Thin-layer chromatography (TLC) plates were visualized by staining within
ninhydrin solution in n-butanol. Yields refer to chromatography-purified homogenous materials, except
otherwise stated. H and 13C NMR spectra were recorded on Bruker AMX-300 and AMX-500
spectrometers (Germany) as solutions and reported in the order of chemical shifts (δ) in ppm relative to
the indicated solvent, multiplicity (s, singlet; d, doublet; t, triplet and m, multiplet), number of protons,
and coupling constants (J) in hertz (Hz). ESI-MS and MALDI-TOF MS analyses were performed on
Varian 500-MS ion trap mass spectrometer (USA) and Bruker Daltonics Ultraflextreme MALDI
1
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
TOF/TOF mass spectrometer (Germany), respectively. Purity of final compounds, as determined by
elemental analysis, was >95 % (Table S11).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
General Procedure A: Global Amine Deprotection (Removal of Boc Protecting Groups) for
Preparation of Compounds 1–3. Solution of Boc-protected compounds 12a–c in DCM (2.0 mL) were
treated with trifluoroacetic acid (2.0 mL), stirred at RT for 1 h and concentrated under low vacuo. 2%
methanol in diethylether (2.0 mL) was then added, stirred gently for 1 min and the solvent carefully
decanted to give off-white solid compounds. The crude products were subsequently purified by reverse-
phase flash chromatography (eluted with 100% deionized water) to afford analytically pure compounds
1–3 (80 – 91 %) as off-white TFA salt solid compounds.
Tobramycin Homodimers 1–3. Final compounds 1–3 were prepared by global amine
deprotection of compounds 12a-c, according to general procedure A.
1
Tobramycin Homodimer 1. Yield (80 %). H NMR (500 MHz, D O) δ 7.75 (s, 1H, CH of
2
triazole), 5.19 (d, J = 2.5 Hz, 1H, anomeric H-1’), 5.17 (d, J = 2.5 Hz, 1H, anomeric H-1’), 4.97 (d, J =
2
.7 Hz, 2H, anomeric H-1’’), 4.30 (t, J = 7.0 Hz, 2H, N-CH of linker), 4.13 – 4.06 (m, 2H, H-5’), 3.96
2
(
t, J = 9.7 Hz, 2H, H-5’’), 3.78 – 3.49 (m, 22H, H-5, H-6, H-4, H-2’, H-4’, H-2’’, H-4’’, H-6’’, O-CH2
of linker 2), 3.44 – 3.31 (m, 6H, H-1, H-3, H-3’’), 3.25 – 3.17 (m, 2H, H-6’), 3.14 – 3.08 (m, 2H, H-
’), 2.61 (m, 2H, CH of linker), 2.36 (dt, J = 12.9, 4.3 Hz, 2H, H-2), 2.09 – 2.00 (m, 4H, H-3’), 1.84 –
6
2
1
3
1.71 (m, 4H, H-2, CH of linker), 1.57 – 1.41 (m, 6H, CH of linker 3), 1.19 – 1.16 (m, 2H). C NMR
2
2
(
125 MHz, D O) δ 163.28 (TFA), 163.00 (TFA), 162.72 (TFA), 162.44 (TFA), 124.16 (CH of triazole),
2
1
17.92 (TFA) 115.54 (TFA), 101.27 (C-1’’), 101.17 (C-1’’), 92.76 (C-1’), 82.09 (C-4’’), 81.99 (C-4’’),
81.80 (C-5), 81.77 (C-5), 77.07 (C-5’’), 76.96 (C-5’’), 76.14 (C-5’), 75.99 (C-5’), 73.21 (C-4), 73.18 (C-
4
), 73.06 (O-CH of linker), 72.51 (O-CH of linker), 68.55 (C-2’’), 64.86 (C-6), 64.80 (C-6), 63.10 (C-
2
2
28
ACS Paragon Plus Environment

Page 29 of 51
Journal of Medicinal Chemistry
1
2
3
4
4
’), 59.35 (C-6’’), 59.24 (C-6’’), 54.82 (C-3’’), 54.44, 50.86 (CH of linker), 49.67 (C-1), 49.61 (C-1),
2
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
8.36 (C-3), 47.27 (C-2’), 47.24 (C-2’), 38.36 (C-6’), 38.32 (C-6’), 28.89 (CH of linker), 27.98 (C-3’),
2
27.70 (C-2), 26.39 (CH of linker), 25.99 (CH of linker), 24.97 (CH of linker), 24.12 (CH of linker),
2
2
2
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
+
+
1
7.79, 16.32. MALDI: m/e calcd for C H N O H , 1112.653; found 1112.650 [M + H]
46 89 13 18
1
Tobramycin Homodimer 2. Yield (85 %). H NMR (500 MHz, D2O) δ 7.68 (s, 1H, CH of
triazole), 5.21 (d, J = 2.6 Hz, 1H, anomeric H-1’), 5.20 (d, J = 2.6 Hz, 1H, anomeric H-1’), 4.99 (d, J =
.5 Hz, 2H, anomeric H-1’’), 4.24 (t, J = 7.1 Hz, 2H, N-CH of linker), 4.14 – 4.07 (m, 2H, H-5’), 3.97
3
2
(td, J = 9.8, 4.8 Hz, 2H, H-5’’), 3.79 – 3.50 (m, 22H, H-5, H-6, H-4, H-2’, H-4’, H-2’’, H-4’’, H-6’’, O-
CH of linker 2), 3.46 – 3.34 (m, 6H, H-1, H-3, H-3’’), 3.27 – 3.19 (m, 2H, H-6’), 3.18 – 3.08 (m, 2H,
2
H-6’), 2.65 – 2.54 (m, 2H, CH of linker), 2.37 (dt, J = 12.8, 4.4 Hz, 2H, H-2), 2.14 – 1.97 (m, 4H, H-
2
3
’), 1.82 – 1.69 (m, 4H, H-2, CH of linker), 1.60 – 1.41 (m, 6H, CH of linker 3), 1.21 – 1.11 (m, 4H,
2
2
1
3
CH of linker 2). C NMR (125 MHz, D O) δ 163.37 (TFA), 163.09 (TFA), 162.81 (TFA), 162.53
2
2
(
TFA), 123.62 (CH of triazole), 119.96 (TFA), 117.67 (TFA), 115.35 (TFA), 113.04 (TFA), 101.32 (C-
’’), 101.30 (C-1’’), 92.76 (C-1’), 81.96 (C-4’’), 81.84 (C-5), 76.93 (C-5’’), 76.88 (C-5’’), 75.99 (C-5’),
75.89 (C-5’), 73.49 (O-CH of linker), 73.18 (C-4), 73.13 (O-CH of linker), 68.55 (C-2’’), 64.79 (C-6),
1
2
2
6
4
4.78 (C-6), 63.14 (C-4’), 63.10 (C-4’), 59.23 (C-6’’), 54.82 (C-3’’), 50.59 (CH of linker), 49.69 (C-1),
2
8.37 (C-3), 47.29 (C-2’), 47.25 (C-2’), 38.41 (C-6’), 38.34 (C-6’), 29.41 (CH of linker), 29.28 (CH of
2
2
linker), 28.85 (CH of linker), 28.05 (C-3’), 27.98 (C-3’), 27.71 (C-2), 25.83 (CH of linker), 25.19 (CH
2
2
2
+
of linker), 24.65 (CH of linker), 24.31 (CH of linker). MALDI: m/e calcd for C H N O H ,
2
2
48 93 13 18
1
140.684; found 1140.695 [M + H]+
1
Tobramycin Homodimer 3. Yield (91 %). H NMR (500 MHz, D O) δ 7.73 (s, 1H, CH of
2
triazole), 5.21 (d, J = 2.6 Hz, 1H, anomeric H-1’), 5.20 (d, J = 2.6 Hz, 1H, anomeric H-1’), 4.98 (d, J =
3
.4 Hz, 2H, anomeric H-1’’)), 4.25 (t, J = 7.0 Hz, 2H, N-CH of linker), 4.10 (dt, J = 8.4, 3.8 Hz, 2H, H-
2
29
ACS Paragon Plus Environment