Full text of DOI:10.1039/d0md00174k

RSC
Medicinal Chemistry
View Article Online
View Journal
RESEARCH ARTICLE
Rational design, synthesis and testing of novel
tricyclic topoisomerase inhibitors for the
treatment of bacterial infections part 1†
Cite this: DOI: 10.1039/d0md00174k
*
R. Kirk,
A. Ratcliffe, G. Noonan, M. Uosis-Martin, D. Lyth, O. Bardell-Cox,
J. Massam, P. Schofield, S. Hindley, D. R. Jones, J. Maclean, A. Smith, V. Savage,
S. Mohmed, C. Charrier, A-M. Salisbury, E. Moyo, R. Metzger, N. Chalam-Judge,
*
J. Cheung, N. R. Stokes,
S. Best, M. Craighead, R. Armer and A. Huxley
The alarming reduction in drug effectiveness against bacterial infections has created an urgent need for the
development of new antibacterial agents that circumvent bacterial resistance mechanisms. We report here
a series of DNA gyrase and topoisomerase IV inhibitors that demonstrate potent activity against a range of
Gram-positive and selected Gram-negative organisms, including clinically-relevant and drug-resistant
strains. In part 1, we present a detailed structure activity relationship (SAR) analysis that led to the discovery
of our previously disclosed compound, REDX05931, which has a minimum inhibitory concentration (MIC)
of 0.06 μg mL−1 against fluoroquinolone-resistant Staphylococcus aureus. Although in vitro hERG and CYP
inhibition precluded further development, it validates a rational design approach to address this urgent
unmet medical need and provides a scaffold for further optimisation, which is presented in part 2.
Received 25th May 2020,
Accepted 7th August 2020
DOI: 10.1039/d0md00174k
rsc.li/medchem
Fluoroquinolones are a first-line antibacterial therapy used
to treat both Gram-negative and Gram-positive bacterial
1. Introduction
As reported by the World Health Organisation (WHO) and the
Centers for Disease Control and Prevention (CDC), there is a
clear and growing unmet need for novel antibacterial agents
to combat infections.1,2 The rapid spread of multi-drug
resistant (MDR) bacterial strains, coupled with the lack of
new antibiotics in clinical development, could soon result in
society entering a ‘post-antibiotic era’.3–5 For example, in a
recent hospital admission a patient presented with an acute
bacterial infection resistant to 26 different antibiotics,
including all aminoglycosides and polymyxins.6 Left
unaddressed, it is predicted that by 2050 deaths attributed to
bacterial infections could rise to 10 million per year, with an
estimated cost to the global economy of $100 trillion.7
infections. Fluoroquinolones act by inhibiting bacterial type II
topoisomerases, DNA gyrase and topoisomerase IV, whose
function are to maintain DNA topology during replication.10
DNA gyrase/topoisomerase IV are tetrameric enzymes that
consist of two GyrA (DNA gyrase)/ParC (topoisomerase IV)
subunits and two GyrB (DNA gyrase)/ParE (topoisomerase IV)
subunits. Fluoroquinolones inhibit the enzymes by creating a
cleaved complex between these bacterial enzymes and DNA.
From a combination of mechanistic studies, and crystal
structures of enzyme–fluoroquinolone complexes, a model for
fluoroquinolone binding has been established.11–14 A primary
interaction involves chelation by the β-keto acid moiety on the
fluoroquinolone to
a
non-catalytic Mg2+ ion, which is
Staphylococcus aureus (S. aureus) causes a range of illnesses,
from skin and soft tissue infections, to pneumonia, endocarditis
and sepsis. The CDC estimates that there were around 80 000 S.
aureus infections leading to 11 000 deaths in the US in 2011
alone. Despite a near decade-long decline as a result of
improved antibiotic stewardship and hospital hygiene,8 Public
Health England (PHE) has recently (2016/2017) reported an
8.9% rise in incidence of S. aureus bacteraemia of 8.9%.9
anchored to the active site within GyrA via a water-mediated
metal ion bridge with serine and an acidic residue15,16
(Fig. 1A). Human type II topoisomerases lack these key
residues, resulting in fluoroquinolone selectivity for bacterial
enzymes. The fluoroquinolone core also intercalates with the
bacterial DNA (Fig. 1B). This π-stacking interaction plays a key
role in determining fluoroquinolone potency13 which is likely
to be influenced by the distribution of electron density within
the fluoroquinolone core. The amine functionality at C7
(Fig. 1D) has no obvious contacts with the enzymes across
multiple species but may interact with Arg458 (S. aureus GyrA
nomenclature, Fig. 1C).
Redx Anti-Infectives Ltd, Alderley Park, Mereside, Macclesfield SK10 4TG, UK
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
d0md00174k
This journal is © The Royal Society of Chemistry 2020
RSC Med. Chem.

View Article Online
Research Article
RSC Medicinal Chemistry
Fig. 1 Moxifloxacin bound to S. aureus DNA gyrase, taken from PDB: 5CDQ. A) Moxifloxacin chelation to a non-catalytic magnesium ion, which in
turn has polar interactions with Ser84 and Glu88 through an expected water network, as observed in homologous crystal structures 5BTI through
to 5BTN. B) Intercalation of moxifloxacin with DNA. C) Nearby amino acid Arg458 at the C-7 position. D) 2D representation of moxifloxacin,
highlighting the C-7 position.
Fluoroquinolone resistance (FQR) is frequently driven by
point mutations in the quinolone resistance-determining
regions (QRDRs) of the DNA gyrase (encoded by GyrA and GyrB)
and topoisomerase IV (encoded by ParC and ParE; known as
GrlA and GrlB for S. aureus),17–20 resulting in target-based
resistance. One of the most widely-studied resistance
mechanisms is the mutation of the key serine and/or acidic
residues within the active site of GyrA, which disrupts the
Fig. 2 Chemical structures of A) zoliflodacin, B) gepotidacin, and C) EV-035. D) EV-035 orthogonal view, exemplifying the torsion angle between
the methyl/cyclopropyl group.
RSC Med. Chem.
This journal is © The Royal Society of Chemistry 2020

View Article Online
RSC Medicinal Chemistry
Research Article
water–Mg2+ ion bridge, thereby preventing fluoroquinolone
chelation.21–23 Recently, two new classes of bacterial
topoisomerase II inhibitors, which do not rely on this
interaction, have completed phase II clinical trials. The
spiropyrimidinetrione, zoliflodacin (Fig. 2A), interacts with DNA
gyrase and topoisomerase IV in
a similar pocket to
Fig. 4 Areas of derivatisation.
fluoroquinolones, and demonstrates activity against Gram-
positive and fastidious Gram-negative organisms.16,24–28 The
crystal structure of a closely-related compound, quinolone
pyrimidone trione-1 (QPT-1), provides compelling evidence that
had low thermodynamic solubility at physiological pH (3
μM). Without an obvious solubilising group, this property
alone would require extensive optimisation for progression.
For these reasons, we decided to conduct a medicinal
chemistry optimisation program to explore the SAR of this
series (Fig. 4), focusing on modifications of the core (X and
R1), R4 substitution (aromatic and saturated heterocycles), as
well as simple substitution at the oxazole 2-position (R3).
Since our initial publications,34,35 two papers on an
imidazopyrazinones (IPYs) series have been published which
have a similar scientific rationale to our novel tricyclic
topoisomerase inhibitor (NTTI) series. Although similar, the
IPY series did not demonstrate superior potency to the NTTI
series and displayed genotoxic liabilities.38,39 Herein, we
describe the results of our initial work, presenting the
chemical, in vitro biological and pharmacokinetic profiles of
our optimisation program.
this series binds in
a
unique fashion compared to
fluoroquinolones, relying upon interactions with DNA gyrase
Asp437 (S. aureus nomenclature) and DNA. A number of non-
fluoroquinolone ‘novel bacterial type II topoisomerase
inhibitors’ (NBTIs), (such as gepotidacin, Fig. 2B), have also
been reported in recent years, and are under active
development. The NBTI class works by binding to DNA gyrase
in a pocket close to, but distinct from, the fluoroquinolone
binding site,15,16,29 and intercalating into the DNA. As a result,
this class does not display cross-resistance with
fluoroquinolones.30,31 Another series of experimental bacterial
topoisomerase II inhibitors are the 2-pyridones (isomeric
quinolones), exemplified by EV-035 (Fig. 2C). This series of
compounds binds to the same pocket as the fluoroquinolones
with similar interactions, owing to their β-keto acid
functionality and 2-pyridone core. EV-035 displays broad-
spectrum in vitro activity against both wild type (WT) and
fluoroquinolone-resistant strains. This is ascribed to the
apparent increased wedge angle between the methyl/cyclopropyl
motifs32 (Fig. 2D) which is thought to be complementary to the
slightly distorted bound DNA in this region.33
2. Results and discussion
2.1 SAR of tricyclic core
Given the predicted mechanism-of-action, initial efforts
focused on simple modifications of the core to establish the
similarity of the series with the fluoroquinolones. It was
quickly identified that R1 = cyclopropane proved optimal for
antibacterial activity (Table 1). Microbiological activity was
An early-stage drug discovery project was undertaken to
identify novel antibacterials unaffected by FQR mechanisms,
leading to the discovery of REDX04139 (Fig. 3), which lacks
the β-keto-acid function essential for fluoroquinolone activity
and incorporates the aromatic C-7 substituent found in EV-
035. REDX04139 demonstrated excellent in vitro microbiology
against all tested variants of S. aureus, including
≤0.25 μM when
X = C–Me or C–Cl (REDX04139 and
REDX07368, respectively), but when X = C–H (REDX05166),
the microbiological activity increased to 2 μM (Table 1).
Incorporation of F at the R2 position is a known strategy to
improve traditional fluoroquinolone potency40,41 and whilst
there was a modest improvement in the X = C–H series
(REDX05166 vs. REDX05557), this was not seen in the X =
Me, R1 = Et series (REDX05237 vs. REDX05028). Additionally,
unlike its fluoroquinolone counterpart, enoxacin, X = N
(REDX05634) was inactive. Replacement of the H substituent
at the R3 position of the oxazole moiety with CH3
(REDX05282) was tolerated without significant impact on
potency versus the Gram-positive strains tested (Table 1).
Finally, the fused core (REDX06345, levofloxacin analogue)
had an MIC of 0.5 μg mL−1.
fluoroquinolone and multidrug-resistant strains, with
a
MIC90 of 0.25 μg mL−1. It also demonstrated an excellent
microbiological profile against a range of resistant Neisseria
gonorrhoeae strains with MICs ranging from 0.06 to 1 μg
mL−1.34,35 It is noteworthy to add, however, that there was no
measured activity against Escherichia coli (E. coli), possibly
due to the relatively high lipophilicity (measured log D = 2.96)
observed for REDX04139, which is known to disfavor
permeability into Gram-negative bacteria.36,37 REDX04139
The apparent beneficial effect of R1 = cyclopropane and X
= C–Me led us to evaluate the conformation of a difluoro
intermediate (Fig. 5A) via small molecule X-ray crystallography
(Fig. 5B). The results indicate that the R1 cyclopropyl
substituent is significantly displaced from the plane of the
core, with increased trigonal character of the amide nitrogen.
Fig. 3 Structure of REDX04139.
This journal is © The Royal Society of Chemistry 2020
RSC Med. Chem.

View Article Online
Research Article
RSC Medicinal Chemistry
Table 1 SAR of tricyclic core, minimum inhibitory concentrations (MIC)
versus S. aureus ATCC 29213 are given in μg mL−1
ring, which we hypothesised could strengthen the cation–π
interaction (Fig. 6). In keeping with this premise, mono
fluoro derivatives, (REDX05866 and REDX05412, ortho and
meta, respectively (Table 2)), both had reduced activities
compared to REDX04139, while the unsubstituted aniline
(REDX05510) showed
a
further reduction in activity.
Introduction of cyano group (REDX07910) sustained
a
activity, whilst reducing plasma protein binding (PPB)
(Table 2).
Compound
X
R1
R2
R3
MIC
Nitrogen-containing six-membered heterocycles at R4
were less active than the parent molecule REDX04139.
2-Amino pyrimidine (REDX05774) displayed MICs against
methicillin-sensitive S. aureus (MSSA) ATCC 29213 and
Streptococcus pneumoniae (S. pneumoniae) ATCC 49619 of
1 and 4 μg mL−1, respectively. The pyrazine (REDX06615)
and pyridazine (REDX06062) analogues however, were
much less potent, with both displaying activities no
better than 64 μg mL−1. It should be noted that these
derivatives were also much less stable to chemical
manipulation, with observations of hydrolytic cleavage of
the oxazole moiety.
Ciprofloxacin
REDX04139
REDX05028
REDX05166
REDX05237
REDX05282
REDX05557
REDX05634
REDX06345
REDX07368
—
—
Cprop
Et
Cprop
Et
Cprop
Cprop
Cprop
Fc
—
H
H
H
F
H
F
F
—
H
H
H
H
Me
H
H
H
H
0.25
0.03
2
2
1
0.03
0.25
128
0.5
0.25
C–Me
C–Me
C–H
C–Me
C–Me
C–H
N
Fc
C–Cl
H
H
Cprop
The torsion angle between the C–Me and N-cyclopropyl
groups was 28° – which is in keeping with previously reported
torsion angle data for the aforementioned 2-pyridone, EV-035
(ref. 32 and 42) (Fig. 2D). The implication of this torsion angle
suggest that the C–Me may have a steric interaction with the
ortho-F of the difluoroaniline from REDX04139. However,
there is no observed impact of this potential interaction in
our SAR as outlined in this manuscript.
To further our understanding, we also quantified the
inhibition of DNA supercoiling and DNA decatenation,
mediated by DNA gyrase and topoisomerase IV, respectively.
Previously-reported
compounds
(REDX04139
and
REDX05931) are highly-active against both enzyme targets
indicating a dual targeting mechanism of action.34
Next, we evaluated the role of the aniline. Removal of
the nitrogen maintained antibacterial potency (REDX05510
vs. REDX05867). Conversion to the difluoro benzoic acid
derivative (REDX06189) demonstrated a considerable loss
of activity, which could be due to the inherent
incompatibility of the acidic environment within the active
site, specifically Glu477, but could equally be due to loss
in permeability and/or increased efflux. Analogues that
retained a H-bond donor functionality, but located further
from the core, such as benzylic alcohols and amines
(REDX06627 and REDX05853) had MICs of ≤0.5 μg mL−1
against S. aureus ATCC 29213 (Table 2). Despite the
metabolic risks associated with these unsubstituted
benzylic analogues, they provided a good indication that
extending this portion of the molecule might be well-
tolerated. Fused heterocycles designed to explore this
hypothesis, such as the isoindoline analogue (REDX06305)
offered an excellent antibacterial activity profile against
2.2 SAR of aromatic substitution at the R4 position
We then explored the SAR of the aromatic R4 substituent
(Table 2). Using S. aureus as a model (PDB 5CDQ), we
predicted that the original incorporation of the
difluoroaniline group provided two points of interaction at
the active site of S. aureus DNA gyrase/topoisomerase IV,
via H-bonding of the aniline nitrogen with an adjacent
Glu477 residue and
a cation–π interaction between a
highly conserved Arg458 residue and the R4-aryl group.
Incorporation of electron modulating nitrogen and fluorine
substituents can attenuate the electron density on the aryl
MSSA, but unfortunately also proved to have
metabolic turnover (data not shown). Other fully aromatic
systems, including benzoxadiazole (REDX06110),
a high
benzimidazole (REDX06530), and indazole (REDX05504)
had good microbiological profiles and appear to offer a
favourable π-stacking interaction with the conserved amino
acid Arg458,43 and with DNA bases. However, these
compounds all had thermodynamic solubilities of <9 mg
L−1, log D7.4 values of >3 and were >95% bound in
human plasma which precluded them from further
analysis. Introduction of an amine in the 2-position of the
Fig. 5 A) 2-D diagram of crystallised structure B) orthogonal view,
illustrating torsion angle between methyl/cyclopropyl.
RSC Med. Chem.
This journal is © The Royal Society of Chemistry 2020

View Article Online
RSC Medicinal Chemistry
Research Article
Table 2 Selected microbiological and in vitro DMPK data of R4 SAR. MICs given in μg mL−1 for S. aureus ATCC 29213 (MSSA) and S. pneumoniae ATCC
49619 – data are based on at least two experimental replicates, shown as modal averages. Mean values are provided for PPB, log D7.4 and
thermodynamic solubility (TD Sol)
MIC S.
aureus
MIC S.
pneumoniae
Cytotoxicity
PPB (human)%
free
Compound
REDX04139
R4
(HepG2, CC50 μg mL−1
)
Measured log D7.4
TD Sol (mg L−1
)
0.03
1
>128
1.2
2.96
<2.5
REDX05412
REDX05504
0.12
2
>128
>16
n.d.
2.4
n.d.
3.29
n.d.
1.1
0.004
0.5
REDX05510
REDX05774
REDX05853
0.5
1
4
n.d.
>32
>32
3.8
2.59
1.32
3.14
n.d.
15
4
n.d.
2.6
0.06
0.25
1.9
REDX05866
0.12
1
16
3.0
3.05
1.1
REDX05867
REDX06062
1
4
n.d.
n.d.
n.d.
n.d.
1.46
n.d.
6.8
64
>128
>30
REDX06110
REDX06189
0.25
64
4
>32
3
3.03
n.d.
<9
>128
n.d.
n.d.
n.d.
REDX06305
REDX06530
0.12
0.25
0.5
>64
>16
23.5
4.7
0.75
3.03
208.6
0.25
<9
REDX06615
>64
>64
n.d.
n.d.
n.d.
n.d.
This journal is © The Royal Society of Chemistry 2020
RSC Med. Chem.

View Article Online
Research Article
RSC Medicinal Chemistry
Table 2 (continued)
MIC S.
aureus
MIC S.
pneumoniae
Cytotoxicity
PPB (human)%
free
Compound
REDX06627
R4
(HepG2, CC50 μg mL−1
)
Measured log D7.4
TD Sol (mg L−1
n.d.
)
0.5
2
>32
26
n.d.
REDX07125
REDX07910
0.25
0.06
0.25
0.25
>32
>4
5.2
4.2
2.52
2.94
16
n.d.
5,6 bicyclic systems such as REDX07125 maintained
potency but delivered a high frequency of resistance (FoR)
of 1.11 × 10−7 (at 4 × MIC) against S. aureus ATCC 29213.
Comparatively, the FoR of REDX04139 against the same
strain was <3.8 × 10−10 (at 4 × MIC). Compounds in this
series generally displayed low cytotoxic against the HepG2
cell line, with CC50 values >16 μg mL−1. REDX07910 was
found to be the only outlier with an CC50 of >4 μg mL−1.
However, our preliminary kinetic solubility assessment of
this compound showed a solubility of 6 μM which may
account for this result.
potency against S. aureus ATCC 29213 as REDX04139
(Table 3). We predicted the methylene amine could offer a
potential interaction with the neighbouring serine residue
via the NH2 group, whilst potentially addressing the
physicochemical issues associated with REDX04139.
REDX05931 displayed low MICs vs. Gram-positive and
fastidious Gram-negative bacteria, including MDR strains
of N. gonorrhoeae.34 This compound was then progressed
to
a murine model of N. gonorrhoeae infection with
encouraging results.35 However, due to inhibition of
several CYP enzymes (Table 3) and a potential interaction
with the hERG ion channel, REDX05931 was not
developed further.
A range of other methylene amine
2.3 Modifications at the R4 position
derivatives were prepared. Two are exemplified in Table 3.
Replacement of the difluoroaniline with an indazole
(REDX07517) maintained potency but did not improve the
CYP inhibition or hERG profiles. The aza-indazole
(REDX07473), however, showed higher IC50 values for both
CYP and hERG inhibition. However, the potency of this
compound against FQ-resistant strains was deemed
insufficient to warrant further studies.
Having shown scope to improve both potency and
physicochemical properties through substitutions at the R4
position, we next explored the 2-position of the oxazole
(R3, Fig. 4). Introduction of
(REDX05931) in this position maintained the same
a
methylene amine
3. Conclusion
We have described the rational design and synthesis of
a range of novel bacterial topoisomerase inhibitors with
excellent in vitro activity against wild-type and FQR
Gram-positive strains. This work contests the well
reported SAR that β-keto acids are essential for
inhibiting DNA gyrase/topoisomerase IV and offers
a
subtle strategy to improve the resistance profiles of
bacterial topoisomerase targeting therapies. It is
interesting to note that in spite of the hypothesised
interaction of the methylene amino group of compounds
REDX05931, REDX07473 and REDX7517 with Ser84, the
Fig. 6 Proposed C-7 aryl substituent interactions of REDX04139 in
S. aureus DNA gyrase, modelled using Forge™ from Cresset and
PDB 5CDQ.
RSC Med. Chem.
This journal is © The Royal Society of Chemistry 2020

View Article Online
RSC Medicinal Chemistry
Research Article
Table 3 Microbiology and toxicity profiles of REDX05931 derivatives. MICs given in μg mL−1; data are based on at least two experimental replicates,
shown as model averages
REDX05931
REDX07473
REDX07517
Comparative data
REDX04139
0.03
Ciprofloxacin
0.25
S. aureus ATCC 29213
(MSSA)
0.06
0.25
0.06
S. aureus NRS74 (MRSA)
S. aureus VRS8 (MRSA)
CYP IC50 (μM)a
0.06
0.5
0.06
4
2
2
0.5
2
128
32
—
>30, >30, >30, >30,
>30
3% inhibition at 100 μM 7.9
>128
<2.5
7.5, 0.3, 0.5, 12.9,
11.3
30.0, 0.8, 2.9, 22.6,
22.8
45.08
>128
12.3, 0.2, 0.8, 1.3,
12.7
6.2
∼32
hERG IC50 (μM)
—
—
Cytotoxicity HepG2 CC50
>64
36
(μg mL−1
)
TD Sol (mg L−1
)
618
10
—
a
CYP inhibition values in order as follows: 1A2, 2C9, 2C19, 3A4, 2D6.
activity of this series of compounds is only diminished
seven fold vs. MRSA NRS74, with target relevant
mutations of GyrA (S84L, E409L) and GrlA (S80F). In
contrast, ciprofloxacin has a 500-fold drop-off in activity.
The implication being that a) the proposed interaction
is not essential to activity, b) the loss of activity may
be due to an alternative mechanism(s), or c) the
proposed MoA is incorrect. To improve the drug-like
properties of this series, the solubility, CYP inhibition
and hERG inhibition profiles need to be addressed. In
gave the key anhydride (6). At this point in the synthesis,
we either carried out condensation with ethyl
nitroacetate48 (route A, Scheme 1a, in THF at 60 °C) to
give hydroxy nitro intermediate (7), which, after
reduction of the nitro-group and cyclisation using an
ortho-formate,49 generated the tricyclic core, or in a more
direct fashion (route B, Scheme 1a) reacted anhydride (6)
with ethyl isocyanoacetate to form the tricycle.50 In both
a
a
cases the desired aryl group was added in
a Suzuki
coupling. Substitution of the oxazole 2-position to give
2-methyl derivatives was achieved via cyclisation of amino
hydroxyl intermediate (8) with trimethyl orthoacetate. The
synthesis was then completed as above using a Suzuki
arylation with appropriate aryl boronic acid or pinacol
boronic esters. Alternatively, the tricyclic core (9a/b) can
undergo Miyura borylation, followed by reaction with the
desired aryl halide (REDX06062).
part
2,
we
discuss
the
merits
of
saturated
heterocycloalkyls at the R4 position (Fig. 4). Optimisation
of the saturated amine heterocycle at this position, has
in the past led to improvements to the antibacterial
spectrum of activity, covering Gram-positive, Gram-
negative and anaerobic bacteria, whilst also improving
in vitro pharmacokinetic and physicochemical properties
including half-life and solubility.44
The R1
= ethyl (Fig. 4) derivative was successfully
synthesised using the procedure detailed in Scheme 1b via
alkylation of the commercially available isatin (12) with ethyl
iodide. This was then followed by the procedure detailed in
Scheme 1a to give the desired product.
4. Chemistry
4.1 Derivatisation of core
Our synthetic route began with 2-bromo-3-methylaniline
and introduction of cyclopropane moiety as an initial step
using (1-ethoxycyclopropoxy)trimethylsilane45 followed by
borane reduction to give (3). Acylation with oxalyl chloride
followed by intra-molecular Friedel–Crafts acylation,46
using aluminium trichloride, gave isatin (4). This was then
reacted with hydrogen peroxide to provide the anthranilic
acid (5).47 Treatment of (5) with triphosgene or phosgene
4.2 Chemistry of X substitution
To determine the contribution of the X= methyl
substituent (Fig. 4) to the observed anti-bacterial effect,
we initially synthesised
X = C–H with R2 = fluorine
derivatives, as depicted in Scheme 2. 1,4-diBromo,2,5-
difluorobenzene (14) was treated with butyl lithium at
This journal is © The Royal Society of Chemistry 2020
RSC Med. Chem.

View Article Online
Research Article
RSC Medicinal Chemistry
Scheme 1 1a Reagents and conditions: (a) (1-ethoxycyclopropoxy)-trimethyl-silane, AcOH, MeOH, rt-70 °C, 16 h, quant. (b) BH3. THF, THF, 0–70
°C, 18 h. (c) (1) Oxalyl chloride, Et2O, 40 °C; (2) AlCl3, DCM, 0 °C – rt, 21% (d) H2O2, NaOH O °C – rt, 2 h, 99% (e) triphosgene, rt, 3 h. (f) Ethyl
nitroacetate, THF, 60 °C, 16 h, 27% (g) sodium hydrosulfite, EtOH, H2O, 77 °C, 1 h. (h) Triethyl orthoformate, reflux, 2 h or trimethyl orthoacetate,
cat HCl, reflux, 4 h, 40% (i) aryl pinacol ester, Pd(dppf)Cl2, Cs2CO3, DME, H2O, MW, 120 °C, 20 min. (j) (1) ethyl isocyanoacetate, THF, 60 °C, 16 h;
(2) NaH, DMF, 100 °C, 1 h, 35% (k) bis(pinacolato)diboron, Pd(dppf)Cl2, K2CO3, DME, 75–90 °C, 16–72 h, 85% (l) aryl halide, Pd(dppf)Cl2, Cs2CO3,
DME, H2O, MW, 120 °C, 20 min. 1b Reagents and conditions: (a) Et-I, K2CO3, DMF, 100 °C, 1 h, 88%.
−78 °C and carboxylated with carbon dioxide.51 The
resulting acid (15) was converted to acid chloride and
then to the oxazole carboxylic acid ethyl ester (16) via
treatment with ethyl isocyanoacetate and triethylamine.52
Saponification with lithium hydroxide afforded (17) which
was followed by acid chloride formation (oxalyl chloride,
catalytic DMF) and addition of cyclopropylamine to give
amide (18).
SNAr cyclisation by displacement of fluorine using
potassium carbonate as base a nd 18-crown-6 at 140 °C gave
the desired fluoro oxazole (19), which was then derivatised
via Suzuki cross coupling (Scheme 2).
Saponification, amide coupling and cyclisation gave
intermediate (24), which was arylated via Suzuki cross
coupling, as described previously. The (S)-morpholino
fused core was synthesised in
a similar fashion from
2,3-difluoro-4-bromo benzoic acid (25), using 2-(S)-
aminopropan-3-ol as the amine coupling partner to give
compound (26). This was followed by
a double ring
closure (firstly, the quinolone ring forms then morpholine
ring via sequential SNAr fluorine displacement53). The
resulting tetracyclic core (27) was then derivatised using
standard Suzuki cross coupling conditions.
R2 and X unsubstituted derivatives were synthesised
using an alternative route, starting with Chan–Lam
alkylation54 of 6-bromo indole (28) using cyclopropane
boronic acid. NBS–DMSO oxidation55 produced the isatin
(30) and after treatment of (30) with hydrogen peroxide
and sodium hydroxide, the anthranilic acid (31) was
formed. A one-pot procedure using triphosgene followed
X = C–chloro core was synthesised in an analogous
fashion to Scheme 2, using
a commercially available
carboxylic acid (20, Scheme 3). 8-Aza-tricyclic derivatives
were synthesised in a similar manner using 2,6-dichloro-5-
fluoronicotinic acid (22) and methyl isonitrile carboxylate,
to give the oxazole methyl ester derivative (23).
RSC Med. Chem.
This journal is © The Royal Society of Chemistry 2020

View Article Online
RSC Medicinal Chemistry
Research Article
Scheme 2 Reagents and conditions: (a) n-BuLi, CO2, THF, −78 °C, 97% (b) (1) oxalyl chloride, DMF, 0 °C – rt, 2–16 h; (2) ethyl isocyanoacetate,
THF, NEt3, 0 °C – rt 16 h, 89% (c) (1) LiOH, 1,4-dioxane, rt, 16 h, 85% (d) (1) oxalyl chloride, DMF, DCM, 0 °C – rt, 2 h; (2) cyclopropylamine, DCM, rt,
83% (e) K2CO3, 18-crown-6, DMSO, MW, 140 °C, 90 min, 26% (f) aryl pinacol ester, Pd(dppf)Cl2, Cs2CO3, DME, H2O, MW, 120 °C, 20 min.
Scheme 3 Reagents and conditions: (a) (1) LiOH, 1,4-dioxane, 16 h, 94%; (2) cyclopropylamine, DIPEA, HATU, DCM, rt, 16 h, 74% (3) NaH, DMF, 0
°C – rt, 16 h, 90% (b) K2CO3, 18-crown-6, DMSO, 140 °C, 40 min, 45%.
by treatment with ethyl isocyanoacetate and sodium
hydride gave the tricyclic intermediate (32); which was
then arylated under Suzuki cross coupling conditions
(Scheme 4).
brominate at the 6 position of the isatin, this was bis-
brominated with NBS to give the (35) which was then
selectively mono-debrominated with the use of copper and
propionic acid.57 The regioselectivity of the mono-
debrominated product (36) was confirmed by NMR.
Compound (36) was alkylated with ethyl iodide in the
presence of K2CO3 to give isatin (37). This was then reacted
with hydrogen peroxide/sodium hydroxide to provide the
anthranilic acid (38). Treatment of (38) with triphosgene
followed by NEt3 and ethyl isocyanoacetate formed the
oxazole (39). Cyclisation to the tricyclic core (40) was achieved
4.3 Chemistry of REDX05237
Synthesis of REDX05237 began with the formation of the
isatin (34) from 4-fluoro-2-methyl-aniline (33) and chloral
hydrate in HCl, followed by the addition of hydroxylamine
hydrochloride.56 As we were unable to selectively mono-
This journal is © The Royal Society of Chemistry 2020
RSC Med. Chem.

View Article Online
Research Article
RSC Medicinal Chemistry
Scheme 4 Reagents and conditions: (a) cyclopropylboronic acid, Na2CO3, 2,2′-bipyridyl, Cu(OAc)2, DCE, 70 °C, 5 h, 73% (b) NBS, DMSO, 60 °C, 18
h. (c) NaOH, H2O2, 0 °C – rt, 2 h, 99% (d) (1) triphosgene, THF, rt, 3 h; (2) ethyl isocyanoacetate, NaH, DMF, 100 °C, 2 h. (e) Aryl pinacol ester,
Pd(dppf)Cl2, Cs2CO3, DME, H2O, MW, 70 °C, 2 h.
with use of NaH in DMF, which was then derivatised via
Suzuki cross coupling to give REDX05237 (Scheme 5).
underwent Suzuki arylation with 2,5-difluoro-4-(4,4,5,5-
tetramethyl-1,3,2-dioxaborolan-2-yl)aniline to form the azide
precursor (43). Intermediate (43) was then subject to
standard Staudinger reduction conditions to form
REDX05931. To synthesise REDX07473 and REDX07517,
intermediate (42) was reduced with PPh3 (Staudinger) to
give the 2′-methylene amine intermediate (44), which was
then BOC protected (45) and functionalised with an aryl
4.4 Chemistry of 2′-methylene amine derivatives
Synthesis
of
the
2′-methylene
amine
derivatives
(REDX05931, REDX07473 and REDX07517) began with the
ring closure of the amino hydroxy intermediate (8) with
2-chloro-1,1,1-trimethoxyethane in the presence of boron
trifluoride diethyl etherate to form the ‘2′-methylene
chloro’ intermediate (41). Sodium azide was then used to
displace the chlorine at room temperature to form
intermediate (42) which, for the synthesis of REDX05931
pinicol ester (46). Finally,
a
BOC deprotection of
intermediate(s) (46) with TFA in DCM furnished the
desired products as TFA salts, which were then be
neutralised/purified by column chromatography to elicit
the final compounds as free bases (Scheme 6).
Scheme 5 Reagents and conditions: (a) (1) chloral hydrate, Na2SO4, HCl followed by hydroxylamine hydrochloride, rt – reflux; (2) H2SO4, 50–80
°C, 65% (b) NBS, H2SO4, 0 °C – rt, 6 h. (c) Cu, propionic acid, 130 °C, 83% (d) ethyl iodide, K2CO3, DMF, 100 °C, 1 h. (e) NaOH, H2O2, rt, 2 h, 87% (f)
(1) triphosgene, THF, rt, 3 h; (2) ethyl isocyanoacetate, NEt3, THF, 60 °C, 18 h, 84% (g) NaH, DMF, 100 °C, 1 h, 21% (h) aryl pinacol ester, Pd(dppf)
Cl2, Cs2CO3, DME, H2O, MW, 70 °C, 2 h.
RSC Med. Chem.
This journal is © The Royal Society of Chemistry 2020

View Article Online
RSC Medicinal Chemistry
Research Article
Scheme 6 Reagents and conditions: (a) 2-chloro-1,1,1-trimethoxyethane, boron trifluoride diethyl etherate, DCM, 0 °C – rt, 44% (b) NaN3, DMSO,
rt, 16 h, 86% (c) 2,5-difluoro-4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)aniline, Pd(dppf)Cl2, Cs2CO3, DME, H2O, MW, 120 °C, 20 min, 47% (d)
PPh3, THF/H2O, 75 °C, 2 h, 86% (e) PPh3, THF/H2O, 75 °C, 1 h, 87% (f) di-tert-butyl dicarbonate, NEt3, DCM, rt, 2 h, 74% (g) aryl pinacol ester,
Pd(dppf)Cl2, Cs2CO3, DME, H2O, MW, 70 °C, 2 h (h) trifluoroacetic acid, DCM, 2 h.
Funding
3 D. A. Tadesse, S. Zhao, E. Tong, S. Ayers, A. Singh, M. J.
Bartholomew and P. F. McDermott, Emerging Infect. Dis.,
2012, 18, 741–749.
4 F. MacKenzie, K. Forbes, T. Dorai-John, S. Amyes and I.
Gould, Lancet, 2017, 350, 783.
This study was funded by Redx Pharma Plc in collaboration
with the Royal Liverpool and Broadgreen University Hospitals
NHS Trust.
5 F. L. de Sá Cavalcanti, A. C. S. Almeida, M. A. Vilela, M. A. de
Morais Junior, M. M. C. de Morais and T. C. Leal-Balbino,
Diagn. Microbiol. Infect. Dis., 2013, 77, 250–251.
Conflicts of interest
All authors were or are employees of Redx Pharma and may
hold shares or share options in the company.
6 C. Lei, Notes from the Field: Pan-Resistant New Delhi Metallo-
Beta-Lactamase-Producing Klebsiella pneumoniae
— Washoe
County, Nevada, 2016, https://www.cdc.gov/mmwr/volumes/66/
wr/mm6601a7.htm?s_cid=mm6601a7_w#suggestedcitation.
7 J. O'Neil, Antimicrobial Resistance: Tackling a crisis for the
health and wealth of nations, 2017.
8 M. A. Borg, M. Hulscher, E. A. Scicluna, J. Richards, J.-M.
Azanowsky, D. Xuereb, A. Huis, M. L. Moro, H. C. Maltezou
and U. Frank, J. Hosp. Infect., 2017, 87, 203–211.
9 Research and analysis HPR volume 11 issue 24: news (14 July),
https://www.gov.uk/government/publications/health-protection-
report-volume-11-2017/hpr-volume-11-issue-24-news-24-july.
10 A. D. Bates and A. Maxwell, in DNA Topology, 2005, ISBN
978–0–19-850655-3.
Acknowledgements
We are grateful for the provision of bacterial strains by
the Network on Antimicrobial Resistance in Staphylococcus
aureus (now part of BEI Resources, Manassas, USA) and to
Inspiralis Ltd (Norwich, UK) for performing the DNA
supercoiling, decatenation and cleavage assays. We would
like
to
acknowledge
SYN|THESIS
MedChem
for
synthesizing the first exemplar of this series and Harry
Adams of Sheffield University for X-ray analysis of
intermediate A.
11 K. J. Aldred, R. J. Kerns and N. Osheroff, Biochemistry,
2014, 53, 1565–1574.
Notes and references
1 P. Association, The Guardian, 2017.
12 A. Fabrega, S. Madurga, E. Giralt and J. Vila, Microb.
Biotechnol., 2009, 2, 40–61.
13 D. C. Hooper and G. A. Jacoby, Cold Spring Harbor Perspect.
Med., 2016, 6(9), a025320.
2 Biggest Threats | Antibiotic/Antimicrobial Resistance | CDC,
https://www.cdc.gov/drugresistance/biggest_threats.html,
(accessed May 11, 2017).
This journal is © The Royal Society of Chemistry 2020
RSC Med. Chem.

View Article Online
Research Article
RSC Medicinal Chemistry
14 M. A. Kohanski, D. J. Dwyer and J. J. Collins, Nat. Rev.
Microbiol., 2010, 8, 423–435.
M. Uosis-Martin, G. Noonan, S. Mohmed, S. A. Best, A. J.
Ratcliffe and N. R. Stokes, J. Antimicrob. Chemother.,
2016, 71, 1905–1913.
15 B. D. Bax, P. F. Chan, D. S. Eggleston, A. Fosberry, D. R.
Gentry, F. Gorrec, I. Giordano, M. M. Hann, A. Hennessy, M.
Hibbs, J. Huang, E. Jones, J. Jones, K. K. Brown, C. J. Lewis,
E. W. May, M. R. Saunders, O. Singh, C. E. Spitzfaden, C.
Shen, A. Shillings, A. J. Theobald, A. Wohlkonig, N. D.
Pearson and M. N. Gwynn, Nature, 2010, 466, 935–940.
16 B. D. Bax, G. Murshudov, A. Maxwell and T. Germe, J. Mol.
Biol., 2019, 431, 3427–3449.
17 H. Yoshida, M. Bogaki, M. Nakamura, L. M. Yamanaka and
S. Nakamura, Antimicrob. Agents Chemother., 1991, 35,
1647–1650.
18 S. Conrad, M. Oethinger, K. Kaifel, G. Klotz, R. Marre and
W. V. Kern, J. Antimicrob. Chemother., 1996, 38, 443–455.
19 P. Heisig, Antimicrob. Agents Chemother., 1996, 40, 879–885.
20 D. M. Breines, S. Ouabdesselam, E. Y. Ng, J. Tankovic, S.
Shah, C. J. Soussy and D. C. Hooper, Antimicrob. Agents
Chemother., 1997, 41, 175–179.
35 V. J. Savage, C. Charrier, A.-M. Salisbury, H. Box, N. Chaffer-
Malam, A. Huxley, R. Kirk, G. M. Noonan, S. Mohmed and
M. W. Craighead, Antimicrob. Agents Chemother., 2016, 60,
5592–5594.
36 D. G. Brown, T. L. May-Dracka, M. M. Gagnon and R.
Tommasi, J. Med. Chem., 2014, 57, 10144–10161.
37 M. A. Scorciapino, S. Acosta-Gutierrez, D. Benkerrou, T.
D'Agostino, G. Malloci, S. Samanta, I. Bodrenko and M.
Ceccarelli, J. Phys.: Condens. Matter, 2017, 29, 113001.
38 T. Germe, J. Vörös, F. Jeannot, T. Taillier, R. A. Stavenger, E.
Bacqué, A. Maxwell and B. D. Bax, Nucleic Acids Res.,
2018, 46, 4114–4128.
39 F. Jeannot, T. Taillier, P. Despeyroux, S. Renard, A. Rey, M.
Mourez, C. Poeverlein, I. Khichane, M.-A. Perrin, S. Versluys,
R. A. Stavenger, J. Huang, T. Germe, A. Maxwell, S. Cao, D. L.
Huseby, D. Hughes and E. Bacqué, J. Med. Chem., 2018, 61,
3565–3581.
21 K. J. Aldred, S. A. McPherson, P. Wang, R. J. Kerns, D. E.
Graves, C. L. Turnbough and N. Osheroff, Biochemistry,
2012, 51, 370–381.
40 V. E. Anderson and N. Osheroff, Curr. Pharm. Des., 2001, 7,
337–353.
22 S. K. Morgan-Linnell, L. Becnel Boyd, D. Steffen and L.
Zechiedrich, Antimicrob. Agents Chemother., 2009, 53,
235–241.
23 V. E. Anderson, R. P. Zaniewski, F. S. Kaczmarek, T. D. Gootz
and N. Osheroff, J. Biol. Chem., 1999, 274, 35927–35932.
24 M. D. Huband, M. A. Cohen, M. Zurack, D. L. Hanna, L. A.
Skerlos, M. C. Sulavik, G. W. Gibson, J. W. Gage, E.
Ellsworth, M. A. Stier and S. J. Gracheck, Antimicrob. Agents
Chemother., 2007, 51, 1191–1201.
25 M. Unemo, J. Ringlander, C. Wiggins, H. Fredlund, S.
Jacobsson and M. Cole, Antimicrob. Agents Chemother.,
2015, 59, 5220–5225.
26 D. J. Biedenbach, M. D. Huband, M. Hackel, B. L. M. de
Jonge, D. F. Sahm and P. A. Bradford, Antimicrob. Agents
Chemother., 2015, 59, 6053–6063.
41 T. D. M. Pham, Z. M. Ziora and M. A. T. Blaskovich, Med.
Chem. Commun., 2019, 10, 1719–1739.
42 Y. Kuramoto, Y. Ohshita, J. Yoshida, A. Yazaki, M. Shiro and
T. Koike, J. Med. Chem., 2003, 46, 1905–1917.
43 P. F. Chan, V. Srikannathasan, J. Huang, H. Cui, A. P.
Fosberry, M. Gu, M. M. Hann, M. Hibbs, P. Homes, K.
Ingraham, J. Pizzollo, C. Shen, A. J. Shillings, C. E.
Spitzfaden, R. Tanner, A. J. Theobald, R. A. Stavenger, B. D.
Bax and M. N. Gwynn, Nat. Commun., 2015, 6, 10048.
44 C. M. Oliphant and G. Green, Am. Fam. Physician, 2002, 65,
455.
45 G. Grewal, E. Hennessy, V. Kamhi, D. Li, P. Lyne, V. Oza,
J. C. Saeh, Q. Su and B. Yang, Heterocyclyc Sulfonamides
Having Edg-I Antagonistic Activity, WO2008056150(A1), 2008.
46 Z. Liu, Q. Lei, Y. Li, L. Xiong, H. Song and Q. Wang, J. Agric.
Food Chem., 2011, 59, 12543–12549.
27 M. J. Pucci, S. D. Podos, J. A. Thanassi, M. J. Leggio, B. J.
Bradbury and M. Deshpande, Antimicrob. Agents Chemother.,
2011, 55, 2860–2871.
47 Y. Wang, M. Zhang, S. Cao, H. Lin, M. Gao and Z. Li, Synth.
Commun., 2012, 42, 2715–2727.
28 R. K. Flamm, C. Vojtko, D. T. Chu, Q. Li, J. Beyer, D. Hensey,
N. Ramer, J. J. Clement and S. K. Tanaka, Antimicrob. Agents
Chemother., 1995, 39, 964–970.
48 F. Suzuki, T. Kuroda, Y. Nakasato, H. Manabe, K. Ohmori, S.
Kitamura, S. Ichikawa and T. Ohno, J. Med. Chem., 1992, 35,
4045–4053.
29 W. So, J. L. Crandon and D. P. Nicolau, Antimicrob. Agents
Chemother., 2015, 59, 4956–4961.
49 K. R. Kunz, E. W. Taylor, H. M. Hutton and B. J. Blackburn,
Org. Prep. Proced. Int., 1990, 22, 613–618.
30 S. B. Singh, Bioorg. Med. Chem. Lett., 2014, 24, 3683–3689.
31 E. G. Gibson, B. Bax, P. F. Chan and N. Osheroff, ACS Infect.
Dis., 2019, 5, 570–581.
32 J. Heim, Safe drugs for bad bugs, 2012, http://docplayer.net/
143901303-Safe-drugs-for-bad-bugs.html.
50 M. Suzuki, K. Matsumoto, M. Miyoshi, N. Yoneda and R.
Ishida, Chem. Pharm. Bull., 1977, 25, 2602–2607.
51 A. Ratcliffe, A. Huxley, D. Lyth, G. Noonan, R. Kirk, M. Uosis-
Martin and N. Stokes, Antibacterial Compounds,
WO2015155549(A1), 2015.
33 A. Wohlkonig, P. F. Chan, A. P. Fosberry, P. Homes, J.
Huang, M. Kranz, V. R. Leydon, T. J. Miles, N. D. Pearson,
R. L. Perera, A. J. Shillings, M. N. Gwynn and B. D. Bax, Nat.
Struct. Mol. Biol, 2010, 17, 1152.
52 A. Huxley, R. Kirk, A. Ratcliffe and D. Lyth, Antibacterial
Compounds, WO2017046605(A1), 2017.
53 H. Egawa, T. Miyamoto and J.-I. Matsumoto, Chem. Pharm.
Bull., 1986, 34, 4098–4102.
34 V. J. Savage, C. Charrier, A.-M. Salisbury, E. Moyo, H.
Forward, N. Chaffer-Malam, R. Metzger, A. Huxley, R. Kirk,
54 D. M. T. Chan, K. L. Monaco, R.-P. Wang and M. P. Winters,
Tetrahedron Lett., 1998, 39, 2933–2936.
RSC Med. Chem.
This journal is © The Royal Society of Chemistry 2020

View Article Online
RSC Medicinal Chemistry
Research Article
55 J. Tatsugi, T. Zhiwei and Y. Izawa, ARKIVOC, 2001, i, 67–73,
DOI: 10.3998/ark.5550190.0002.105.
56 Design, synthesis, and QSAR study of novel 2-(2,3-dioxo-
as cytotoxic agents
|
SpringerLink, https://link.springer.
com/article/10.1007/s00044-010-9361-y, (accessed July 27, 2020).
57 H. L. Rayle and R. W. Stephens, Method for preparing
aromatic compounds, US5886210(A), 1999.
2,3-dihydro-1H-indol-1-yl)-N-phenylacetamide
derivatives
This journal is © The Royal Society of Chemistry 2020
RSC Med. Chem.