Discovery of Novel Inhibitors of LpxC Displaying Potent in Vitro Activity against Gram-Negative Bacteria

Article abstract of DOI:10.1021/acs.jmedchem.9b01604

UDP-3-O-((R)-3-hydroxymyristoyl)-N-glucosamine deacetylase (LpxC) is as an attractive target for the discovery and development of novel antibacterial drugs to address the critical medical need created by multidrug resistant Gram-negative bacteria. By using a scaffold hopping approach on a known family of methylsulfone hydroxamate LpxC inhibitors, several hit series eliciting potent antibacterial activities against Enterobacteriaceae and Pseudomonas aeruginosa were identified. Subsequent hit-to-lead optimization, using cocrystal structures of inhibitors bound to Pseudomonas aeruginosa LpxC as guides, resulted in the discovery of multiple chemical series based on (i) isoindolin-1-ones, (ii) 4,5-dihydro-6H-thieno[2,3-c]pyrrol-6-ones, and (iii) 1,2-dihydro-3H-pyrrolo[1,2-c]imidazole-3-ones. Synthetic methods, antibacterial activities and relative binding affinities, as well as physicochemical properties that allowed compound prioritization are presented. Finally, in vivo properties of lead molecules which belong to the most promising pyrrolo-imidazolone series, such as 18d, are discussed.

Full text of DOI:10.1021/acs.jmedchem.9b01604

Subscriber access provided by WESTERN SYDNEY U
Article
Discovery of Novel Inhibitors of LpxC Displaying
Potent In Vitro Activity against Gram-Negative Bacteria
Jean-Philippe Surivet, Philippe Panchaud, Jean-Luc Specklin, Stefan Diethelm, Anne-Catherine Blumstein,
Jean-Christophe Gauvin, Loïc Jacob, Florence Masse, Gaëlle Mathieu, Azely Mirre, Christine Schmitt, Roland
Lange, Naomi Tidten-Luksch, Carmela Gnerre, Swen Seeland, Charlyse Herrmann, Peter Seiler, Michel
Enderlin-Paput, Aengus Mac Sweeney, Micha Wicki, Christian Hubschwerlen, Daniel Ritz, and Georg Rueedi
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b01604 • Publication Date (Web): 05 Dec 2019
Downloaded from pubs.acs.org on December 5, 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 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Discovery of novel inhibitors of LpxC displaying
potent in vitro activity against Gram-negative
bacteria
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
†
Authors: Jean-Philippe Surivet,* Philippe Panchaud, Jean-Luc Specklin, Stefan Diethelm, Anne-
Catherine Blumstein, Jean-Christophe Gauvin, Loïc Jacob, Florence Masse, Gaëlle Mathieu,
Azely Mirre, Christine Schmitt, Roland Lange, Naomi Tidten-Luksch, Carmela Gnerre, Swen
Seeland, Charlyse Herrmann, Peter Seiler, Michel Enderlin-Paput, Aengus Mac Sweeney, Micha
§
Wicki, Christian Hubschwerlen, Daniel Ritz and Georg Rueedi
Idorsia Pharmaceuticals Ltd, Hegenheimermattweg 91, CH-4123 Allschwil, Switzerland
KEYWORDS: Antibacterial, Gram-negative pathogens, hydroxamic acid, LpxC inhibitors,
scaffold hopping, structure-guided transformation.
ABSTRACT: UDP-3-O-((R)-3-hydroxymyristoyl)-N-glucosamine deacetylase (LpxC) is as an
attractive target for the discovery and development of novel antibacterial drugs to address the
critical medical need created by multi-drug resistant Gram-negative bacteria. Using a scaffold
hopping approach on a known family of methylsulfone hydroxamate LpxC inhibitors, several hit
series eliciting potent antibacterial activities against Enterobacteriaceae and Pseudomonas
aeruginosa were identified. Subsequent hit-to-lead optimization, using co-crystal structures of
inhibitors bound to Pseudomonas aeruginosa LpxC as guides, resulted in the discovery of multiple
chemical series based on i) isoindolin-1-ones, ii) 4,5-dihydro-6H-thieno[2,3-c]pyrrol-6-ones and
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 83
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
iii) 1,2-dihydro-3H-pyrrolo[1,2-c]imidazole-3-ones. Synthetic methods, antibacterial activities
and relative binding affinities, as well as physico-chemical properties that allowed compound
prioritization are presented. Finally, in vivo properties of lead molecules which belong to the most
promising pyrrolo-imidazolone series such as 18d, are discussed.
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 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
INTRODUCTION
The World Health Organization (WHO) classifies antimicrobial resistance (AMR) as one of the
biggest threats to global health, acknowledging that AMR steadily erodes the power of
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
antibiotherapy used to cure life-threatening infections. WHO has established a global priority
pathogens list where multidrug- and extensively drug-resistant (MDR and XDR) Acinetobacter,
Pseudomonas aeruginosa (Pae), Enterobacteriaceae strains (including Klebsiella pneumoniae
2
(Kpn), Escherichia coli (Eco), Serratia and Proteus) ranked high. These Gram-negative bacteria
(GNB) are not only refractory to recommended empiric treatment options (carbapenems and third-
generation cephalosporins), but are also becoming increasingly resistant to some of the last resort
3
drugs such as colistin, creating a critical medical need for novel, effective and safe antibacterial
agents. Public health experts as well as authorities have recognized the urgency of the situation
and economic incentives have been awarded to boost innovation.4
The success of incremental improvement of old antibacterial classes will soon come to an end
as it becomes increasingly difficult to circumvent cross-resistance issues. The level of innovation
in the field of anti-Gram-negative bacteria drugs has been low, and this can mainly be attributed
to the scarcity of drug targets, novel or validated. Expanding the spectrum of activity of drugs
effective against Gram-positive bacteria to include GNB has therefore recently gained a lot of
5
attention. In this context, efforts to identify rationally the processes that drive cytoplasmic drug
6
accumulation in Gram-negative bacteria have been reported. Over the last decade, only a handful
of novel targets amenable to the treatment of Gram-negative infections have been studied, resulting
7
in only a few chemical entities suitable for entry-into-man. Among them, LpxC is a biological
target that holds much promise and has captured the interest of many biopharmaceutical
8
companies, leading to the first Phase I trial of Achaogen's ACHN-975 1 (Figure 1). Motivated by
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 83
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
these successes, we sought to develop an independent program to identify a novel class of LpxC
inhibitors.
LpxC is a cytosolic metal-dependent enzyme found only in Gram-negative bacteria. It catalyzes
the first committed step in the biosynthesis of lipid A, a key component of the Gram-negative outer
membrane. Utilizing a single active-site metal ion, LpxC hydrolyses the 2-acetamide of UDP-(3-
O-((R)-3-hydroxymyristoyl)-N-acetylglucosamine to produce acetate and UDP-(3-O-((R)-3-
hydroxymyristoyl)-glucosamine (Figure 1).9-12
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
OH
HO
HO
OH
O
CO2-
HO
OH
OH
HO
O
HO
HO
HO
O
O
O
O
O
HO
HO
HO
CH3CO2-
CO2-
-
O
O
O
HN
2
H N
LpxA
ACP
UDP
UDP
O
O
Kdo2-Lipid A
HO
HO
2-
HN
3
O PO
UDP
O
LpxC
O
O
O
O
O
NH
UDP-GlucNAc
7
7
HO
O
O
HO
O
O
O
O
9
O
NH
OPO32-
O
O
HO
HO
NH2
O
N
OH
O
H
O
H
H
N
N
N
N
OH
OH
N
OH
H
H
O
O
O
O
Achaogen
ACHN-975 1
Merck
L-573,655 2
Chiron
CHIR-090 3
HO
N
MeO2S
N
H
NH2
N
O
MeO2S
H
H
OH
N
N
N
H
OH
OH
O
O
O
O
O
Novartis
4
O
Pfizer
5
Pfizer
6
3
N
O
Figure 1: LpxC, ubiquitously found in GNB, catalyzes the irreversible deacetylation of UDP-(3-
O-((R)-3-hydroxymyristoyl)-N-acetylglucosamine. Representative inhibitors of LpxC feature a
hydroxamate as a zinc-binding group (box).
Extensive X-ray crystallographic studies of LpxC homologues from several bacterial species
2
+
2+
revealed a unique -α-α- sandwich fold along with a unique metal (Zn or Fe )-binding motif
defining the catalytic site of the enzyme. These structural investigations further provided detailed
4
ACS Paragon Plus Environment

Page 5 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
information on the binding mode of the deacetylation reaction product, as well as small molecule
1
3
inhibitors and other ligands. The key attributes that make LpxC an attractive drug target are its
essentiality for bacterial survival, its gene conservation across a spectrum of bacterial pathogens,
the absence of a human homologue and the availability of specific antibacterial inhibitors with
proven efficacy in animal models of infection. Inhibition of LpxC by small molecules represents
a new mode of antibacterial action with low chance of being compromised by pre-existing target-
specific mechanisms of resistance.14
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
Since the groundbreaking discovery at Merck that L-573,655 2 (Figure 1) exerted antibacterial
activity against Eco by inhibiting lipid A biosynthesis,15 all reported inhibitors of LpxC that
exhibited potent antibacterial activity against Enterobacteriaceae and Pae have shared two
structural elements: (1) a hydroxamic acid moiety as metal binding unit16, 17 and (2) a linear and
rigid lipophilic tail that interacts with the hydrophobic tunnel otherwise occupied by the fatty acyl
chain of the natural substrate (hereafter dubbed substrate-binding passage or passage).17-19
Families of LpxC inhibitors differ from one another mainly in the nature of the chemical scaffold
that links those two recurrent elements. At the time when our research program was initiated,
optimization efforts on threonyl-hydroxamate based LpxC inhibitors deriving from CHIR-090 3
had culminated in the identification of Phase I candidate 1 at Achaogen and lead compound 4 at
Novartis,20, 21 offering only limited opportunity for a fast-follower approach. However, the
concurrent disclosure of the structurally novel methylsulfone hydroxamate derivatives such as 5
and 6 with potent antibacterial activities,22, 23 made this motif an excellent starting point for our
investigations. Herein, we report the scaffold hopping approach that eventually led to the
identification of novel methylsulfone hydroxamate LpxC inhibitors articulated around isoindolin-
2
4
25
1
-one (core I), 1,2-dihydro-3H-pyrrolo[1,2-c]imidazole-3-one (core II) and 4,5-dihydro-6H-
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 83
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
thieno[2,3-c]pyrrol-6-one24 (core III) (Figure 2). Synthesis, crystallographic, biological and
pharmacological data of compounds generated during the subsequent hit-to-lead optimization
program are reported, with a focus on derivatives of the pyrrolo[1,2-c]imidazole-3-one series (core
II).
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
O
O
N
N
N
MeO S
2
H
N
(R)
OH
core I
isoindolin-1-one
core II
R
CORE
O
1,2-dihydro-3H-pyrrolo[1,2-c]imidazole-3-one
O
S
N
core III
4,5-dihydro-6H-thieno[2,3-c]pyrrol-6-one
Figure 2: Core structures of novel methylsulfone hydroxamate LpxC inhibitors based on bicyclic
heteroaromatic systems.
CHEMISTRY
Experimental procedures leading to compounds 8-14 are reported in the Supporting Information
(SI). All LpxC inhibitors 15-20 featuring either phenyl or alkyne-based side chains, were prepared
from terminal alkynes 29, 34 and 40.
6
ACS Paragon Plus Environment

Page 7 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Scheme C1a
O
O
b
MeO S
O
MeO S
2
2
OtBu
OH
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
OMe
N
N
MeO S
2
OtBu a
O
c
O
+
H N
(R)
2
O
22
O
23 X = Br
24 X = I
25
Br
X
I
21
O
MeO S
O
MeO S
2
H
N
2
NHOTHP
N
O
O
N
TMS
d
O
O
27
-OTHP
26
e
1
28 R = TMS
I
f
R1
1
29 R = H
a
Reagents and conditions: (a) NaBH(OAc) , AcOH, DCM, rt, 4h (b) NaI, CuI, trans-N,N'-
3
dimethylcyclohexane-1,2-diamine, dioxane, reflux, 3h (c) TFA, Et SiH, rt, 16h (d) NH OTHP,
3
2
EDC.HCl, HOBt, TEA, DMF, rt, 16h (e) 27, PdCl (PPh ) , CuI, TEA, THF, reflux, 1h (f) K CO ,
2
3 2
2
3
MeOH, rt, 1h (g) PPTS, EtOH, 70°C
As described in Scheme C1, the preparation of isoindolin-1-one alkyne 29 began with a reductive
amination involving known aldehyde 21 and chiral amine 22 of (R) absolute stereochemistry (for
preparation of amine 22, see SI), which after spontaneous lactamization afforded bromo-
isoindolinone 23. A bromine-iodine exchange was performed using NaI in presence of a copper
catalyst (obtained by mixing CuI and a trans-N,N'-dimethylcyclohexane-1,2-diamine) to afford
iodo-derivative 24. Acidic hydrolysis of t-butyl ester 24 using trifluoroacetic acid (TFA) in the
presence of triethylsilane yielded the corresponding free acid 25, which was further converted to
the THP-protected hydroxamic acid 26 by reaction with O-(tetrahydro-2H-pyran-2-
yl)hydroxylamine (NH OTHP) in presence of 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide
2
(
EDC), 1-hydroxybenzotriazole (HOBt) and triethylamine (TEA). Subsequent Sonogashira
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 83
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
26
coupling performed with trimethylsilylacetylene 27 in the presence of PdCl (PPh ) and CuI
2
3 2
afforded intermediate 28 as an equimolar mixture of THP anomers. Treatment with potassium
carbonate in methanol (MeOH) of TMS-protected 28 gave rise to terminal alkyne 29.
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 C2a
H
N
H
O
MeO S
2
O
N
a
OtBu
c, d, e
MeO S
b
2
N
OtBu
N
N
O
I
I
H
O
30
31
O
32
I
MeO S
2
NHOTHP
N
N
O
1
3
3 R = TMS
R
f
1
34 R = H
a
Reagents and conditions: (a) 22, 3Å molecular sieves, DCM, rt then NaBH , MeOH, 0°C (b) CDI,
4
NaH, THF, 0°C (c) TFA, Et SiH, DCM, 0°C to rt, 5h (d) THPONH , T P, DIPEA, THF, rt, 2h
3
2
3
(
e) TMS-CCH, PdCl (PPh ) , CuI, TEA, THF, rt, 1h (f) TBAF, THF, rt
2 3 2
As described in Scheme C2, the synthesis of pyrrolo-imidazolone intermediate 34 was realized
from commercially available pyrrole-aldehyde 30. A reductive amination between 30 and chiral
amine 22 gave rise to an alkylated amine 31, which upon treatment with carbonyl-diimidazole
(
CDI) in the presence of sodium hydride afforded the bicyclic compound 32. The t-butyl ester was
then transformed to the intermediate TMS-alkyne 33 via a hydrolysis-amidation-Sonogashira
coupling sequence using methods described previously in Scheme C1 except that
propylphosphonic anhydride (T3P) and N,N-diisopropylethylamine (DIPEA) were used as amide
coupling reagents. From intermediate 33, TMS-cleavage using tetrabutylammonium fluoride
(
TBAF) in tetrahydrofuran (THF) afforded the terminal alkyne 34 as an equimolar mixture of THP
anomers.
8
ACS Paragon Plus Environment

Page 9 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Scheme C3a
MeO S
O
MeO2S
O
O
2
OtBu
OH
N
a
N
c
d, e
S
O
S
O
S
O
R1
Br
36 R = H
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
35
b
38
3
7 R = Br
1
MeO S
O
2
NHOTHP
N
S
O
1
39 R = TMS
R1
f
1
40 R = H
a
Reagents and conditions: (a) 22, AcOH, DCM, rt, 4h (b) Br , aq AcOH, rt, 16h (c) 4N HCl,
2
dioxane, rt, 16h (d) THPONH , EDC, HOBt, TEA, DMF, rt, 16h (e) TMS-CCH, PdCl (PPh ) ,
2
2
3 2
CuI, TEA, THF, rt, 1h (f) TBAF, THF, rt
The synthesis of thienopyrrolone precursor 40, described in Scheme C3, started with a
cyclization process that took place between commercially available dialdehyde 35 and chiral
amine 22 in acetic acid, affording thienopyrrolone 36 as sole product. Subsequent regioselective
bromination using bromine in aqueous acetic acid gave rise to bromo-intermediate 37. After t-
butyl ester cleavage under acid conditions, the free acid 38 was coupled with NH OTHP and the
2
resulting THP-protected hydroxamic acid was converted to TMS-alkyne 39. Both transformations
employed methods similarly to those reported for the preparation of alkyne 29. The TMS-group
of intermediate 39 was cleaved using TBAF, leading to terminal alkyne 40, recovered as an
equimolar mixture of THP anomers.
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 83
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
Scheme C4a
O
I
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
MeO S
MeO S
2
H
2
H
N
R1
N
N
41-50
a, b
OTHP
OH
CORE
CORE
O
O
29 (core I)
O
15a,b, 17a-j or 19a-d
2
9, 34 or 40
R
N
N
Hal
MeO S
R2
Hal = Br or I
2
H
34 (core II)
51-60
N
OH
CORE
O
O
c, b
or c, d, b
or c, e, b
S
N
R
16a,b, 18a-j or 20a-f
40 (core III)
Chemical structures of groups R
HO
O
O
MsHN
HO
HO
HO
N
O
N
OH
NH2
OH
OH
HO
O
a
Reagents and Conditions: (a) 41-50, PdCl (PPh ) , CuI, TEA, THF, reflux (b) Method A: PPTS,
2
3 2
EtOH, 70°C; Method B: PPTS, aq dioxane, 70°C; Method C: aq HCl, MeCN, rt (c) 51-60, CuCl,
NH OH.HCl, BuNH , rt (d) TBAF, THF, rt, (e) K CO , MeOH, rt.
2
2
2
3
As shown in Scheme C4, the synthesis of hydroxamic acids 15a, b, 17a-j and 19a-d featuring a
phenyl-based side chains were generally performed via a Sonogashira cross-coupling reaction
involving one of the terminal alkynes 29, 34 or 40 and a 4-iodo-substituted phenyl derivative 41-
5
0. Standard coupling conditions (vide supra) were employed. Subsequently, a final THP-
deprotection using suitable acidic conditions afforded hydroxamic acids 15a, b, 17a-j and 19a-d
respectively. The synthesis of hydroxamic acids 16a, b, 18a-j and 20a-f featuring acetylene-based
side chains was achieved via a Cadiot-Chodkiewicz coupling reaction27 between one of the
terminal alkynes 29, 34 or 40 and halo-alkynes 51-60. Reaction conditions utilizing copper
chloride as a catalyst and hydroxylamine as a base were found suitable for all coupling
1
0
ACS Paragon Plus Environment

Page 11 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
experiments. THP-deprotection under acidic conditions afforded hydroxamic acids 16a, b, 18a-j
2
and 20a-f. Depending on the nature of the group R present in coupling partners 51-60, an
intermediate deprotection step (TBAF in THF when a silyl protecting group is present or potassium
carbonate in MeOH in the case of an ester) was required prior to the acid-mediated THP-
deprotection. The detailed synthesis of iodophenyl derivatives 41-50 and halo alkynes 51-60 are
available in the SI.
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
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 83
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
Scaffold hopping. With the methylsulfone hydroxamate inhibitor family as a starting point for
scaffold hopping, we analyzed the published X-ray co-crystal structures of Pfizer's inhibitors 5 and
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
6
in complex with Pae LpxC (PDB accession code 3U1Y and 3UHM). While the methylsulfone
hydroxamate headgroups of 5 and 6 establish crucial H-bonding interactions with the catalytic
metal and its immediate environment, the aromatic components, i.e. phenyl or pyridone cores of
these inhibitors, together with their respective side lipophilic tails appeared to be only involved in
van der Waals interactions with the LpxC protein. Furthermore, the central aromatic cores, seated
at the entrance of the substrate-binding passage, played an essential role in maintaining lipophilic
side chains at the right distance from the hydroxamate anchor and in directing these tails toward
the hydrophobic passage. Therefore, the search for replacements of the central scaffold was
2
3
initiated using pyridone 7 as a starting point. Three design criteria had to be fulfilled. A novel
core scaffold had to (1) be planar, (2) provide two exit vectors projecting in opposite direction,
and (3) be structurally dissimilar from the pyridone ring. The oxazol-2-one ring met these
constraints (Figure 3), although both exit vectors around the 5-membered core formed a marked
obtuse angle (~139°), in clear contrast to the one observed in the original 6-membered pyridone
ring (~180°). As a side aromatic ring, we chose the 2-fluoro-4-methoxyphenyl residue, which had
previously provided superior results in the 2-pyridone series. (Figure 3).23 The absolute
stereochemistry of the chiral methylsulfone hydroxamate headgroup was set to (R), following
Pfizer’s findings.22
1
2
ACS Paragon Plus Environment

Page 13 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
central
aromatic core MeO S
2
MeO S
H
2
H
N
N
Scaffold
Hopping
N
(R)
OH
O
N
(R)
OH
O
O
O
O
O
F
~
180°
7
O
F
methylsulfone
hydroxamate
head group
~
139°
(R)-8
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
side aromatic
ring
Figure 3: Scaffold hopping approach from pyridone 7 leading to novel oxazol-2-one 8. The
numbers indicated below the dash-lined boxes represent the angle at which both exit vectors of the
core intersect.
Oxazol-2-one (R)-8 and (S)-8 were prepared as reported in Supporting Information. The in vitro
antibacterial activity of compounds (R)-8 and (S)-8 were determined against a panel of clinically
relevant Gram-negative pathogens including wild-type Eco ATCC 25922 and Pae ATCC 27853
28
strains, and Kpn T 6474, an MDR strain. Wild-type (wt) Pae PAO1 and an isogenic efflux mutant
(
PA2616, (mexAB-OprM)), were also employed to assess the impact of compound efflux
processes on the antibacterial activity. Binding affinities to Eco and Pae LpxC enzymes were
estimated by differential scanning fluorimetry (DSF) and are expressed as thermal shifts (Tm in
K).29
Table 1: Antibacterial activity against three Gram-negative pathogens and binding affinity for Eco
and Pae LpxC enzymes of compounds 7, (R)-8 and (S)-8.
MIC (g/mL)a
Tm (K)b

Cpds
7
Eco
Kpn
0.25
Pae
0.5
PAO1
PA2616
Shift
Eco
7.8
Pae
≤0.06
0.5
8
≤0.06
≥8
11.8
(R)-8
8
16
8
1
8
5.2
1.9
9.9
3.7
(S)-8
>32
>32
>32
>32
16
NM
aEco ATCC 25922 (wt); Kpn T 6474 (MDR); Pae ATCC 27853 (wt); Pae PAO1 (wt); Pae PA2616, laboratory
(mexAB-OprM) mutant with PAO1 background; MIC efflux shift: calculated as the ratio of MIC PAO1/MIC
PA2616. DSF on isolated enzyme, determined at 25 µM compound concentration. NM: not meaningful.

b
1
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 83
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
As indicated in Table 1, MICs (minimal inhibitory concentration(s), [g/mL]) of both
enantiomers of oxazol-2-one 8 measured against wild-type Gram-negative bacterial strains were
significantly apart, allowing to assign the absolute configuration of the most potent one to (R).22
Nonetheless, (R)-8 elicited moderate to weak MICs (≥ 8 g/mL), much higher than that of
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
pyridone 7. Binding affinities of compound (R)-8 to Eco and Pae LpxC, determined at 25 M
compound concentration, were lower than those of 7 (Table 1), coinciding well with the observed
differences in antibacterial activity. Compounds 7 and (R)- 8 were more potent against the efflux-
defective Pae strain PA2616 than against its wild-type (wt) PAO1 parent. MIC PAO1/MIC
PA2616 ratios (henceforth MIC efflux shifts) calculated for compounds 7 and (R)-8 were found to
be ≥ 8-fold, suggesting that both compounds were efflux pump substrates. Despite its moderate
binding affinity to Pae LpxC, an X-ray co-crystal structure of oxazol-2-one 8 bound to Pae LpxC
could be obtained (Figure 4).30
1
4
ACS Paragon Plus Environment

Page 15 of 83
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
Figure 4: Oxazol-2-one 8 (white carbon atoms) bound in the active site of Pae LpxC (gray ribbon)
(1.9 Å) (PDB code: 6I46). Key interactions of oxazol-2-one 8 with metal (blue sphere) and key
amino acid side chains (gray carbon atoms) are displayed as dotted lines. Electron density of the
ligand is shown as a blue mesh (2Fo-Fc density contoured at 1 sigma). While rac-8 was used for
co-crystallization experiments with Pae LpxC, only active enantiomer (R)-8 was observed.30
The methylsulfone hydroxamate head group of oxazol-2-one (R)-8 is engaged in several metal-
or hydrogen-bonding interactions, either with the Zn atom or amino acid side chains (Glu77,
Thr190, Lys238, and His264), similar to those previously reported for methylsulfone hydroxamate
inhibitors 5 and 6 (Figure 4). The sulfone methyl was positioned within a small hydrophobic
pocket, pointing towards the phenyl ring of Phe191. As observed in case of the pyridone core, the
oxazol-2-one oxygen pointed towards the side chain of Met62. The aromatic tail projected into the
11
substrate-binding passage, i.e. the inner open space of the LpxC insert II motif. The side aromatic
1
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 83
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
ring and the oxazol-2-one were arranged in a nearly coplanar fashion, forming a dihedral angle of
around -10° (37° in case of pyridone 6). The low affinity of inhibitor (R)-8 could be explained by
the smaller angle at which both exit vectors present on its oxazol-2-one core intersect compared to
the one of the pyridone ring. Moreover, the comparison between the X-ray structure of pyridone 7
bound to Pae LpxC generated with the one of compound (R)-8, showed that the insert II α-helix
moved by 2.2 Å towards pyridone 7, probably leading to its tighter fit in LpxC compared to oxazol-
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
-one (R)-8. Concomitant with the characterization of oxazol-2-one (R)-8, several heterocyclic
methylsulfone hydroxamate LpxC inhibitors with promising antibacterial activities were
3
1
published. The scaffold hopping approach thus shifted towards the exploration of bicyclic ring
systems as core motifs to maximally diverge from those 5-membered heterocyclic templates while
taking advantage of our preliminary work. The expectation was that larger bicyclic ring systems
would better occupy the vacant space within the substrate-binding passage, hence be less prone to
thermal motion and therefore more potent at inhibiting LpxC.
From compound (R)-8, different types of bicyclic cores were designed by fusing one additional
aromatic ring to the oxazol-2-one motif. The carbonyl group was found beneficial to attain drug-
32
like properties, leading to the exploration of either benzoxazolone rings (path a) or isoindolin-1-
one-type rings (path b), as depicted in Figure 5.
1
6
ACS Paragon Plus Environment

Page 17 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Ar
MeO S
2
path a
H
Ar
path b
N
N
X
N
O
(R)
OH
N
O
O
O
O
O
O
side aromatic
ring (Ar)
F
~140°
(R)-8
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
ring
contraction
Ar
N
O
X
N
N
N
Ar
N
Ar
S
Ar
O
X
O
O
O
~
140°
9
14
~185°
12 (X = CH)
3 (X = N)
10 (X = CH)
11 (X = N)
~
175°
~150°
1
Figure 5: Paths followed to generate novel bicyclic cores from oxazol-2-one (R)-8. Side aromatic
ring (Ar) remained constant (Ar = 2-fluoro-4-methoxyphenyl). The numbers indicated below the
dash-lined boxes represent the angle at which both exit vectors of the core intersect. The orientation
of Ar relative to the central core was not considered.
The design was further refined integrating the hypothesis that both exit vectors of the central core
intersect with an angle close to 180° (as for pyridones 6 and 7). In the case of benzoxazolone and
isoindolin-1-one rings, both exit vectors intersect with angles that are markedly obtuse (around
140°-150°). In order to restore an overall linear geometry, the phenyl ring was replaced by a five-
membered aromatic ring, giving rise to substituted 4,5-dihydro-6H-thieno[2,3-c]pyrrol-6-one
(
hereafter dubbed thienopyrrolones), 1,2-dihydro-3H-pyrrolo[1,2-c]imidazol-3-one (or pyrrolo-
imidazolones, X = CH) and 6,7-dihydro-5H-imidazo[1,5-a]imidazol-5-one ring systems (or
imidazo-imidazolones, X = N) for which the vectors form an angle closer to 180° (See Figure 5).
Seven representative derivatives featuring these designed cores were prepared as reported in SI.
Isoindolin-1-one 10,24 pyrrolo-imidazolone 1225 and thienopyrrolone 1424 exhibited potent
antibacterial activities against Eco and Kpn strains (MICs ≤ 0.25 g/mL) and moderate activity
against Pae strains with MICs ≤ 2 g/mL (Table 2). Binding affinities of compounds 10, 12 and
1
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 83
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
4, determined by DSF (Tm recorded on isolated enzyme at 25 µM compound concentration),
were comparable on Eco (Tm = 8-9 K) and Pae (Tm = 11-12 K) enzymes, and comparable with
Tm values obtained with pyridone 7. MIC efflux shifts determined for these three derivatives

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
were about 4 to 8-fold, similarly to that of pyridone 7. These efflux shifts indicated that these LpxC
inhibitors were to some extent substrates of the Pae MexAB or OprM-dependent efflux systems.
Benzoxazolone 933 exhibited weaker antibacterial activity against Gram-negative strains than
derivatives 10, 12 and 14 although its binding affinities to both Eco and Pae LpxC (Tm = 8.2 and
11.0 K, respectively) were comparable. The large MIC efflux shift of benzoxazolone 9 (≥ 64-fold)
suggested that the weaker antibacterial activity against Pae resulted from extensive efflux.
Derivatives 11 and 13 were inactive against Kpn and Pae strains, and their MICs against Eco were
at least 32-fold higher than those of their direct analogues 10 and 12. The poor antibacterial profile
of compounds 11 and 13 could be explained by their lower binding affinities to Eco (Tm =5-6 K)
and Pae (Tm~7 K) enzymes, suggesting a negative effect of additional nitrogen atoms present in
the annulated aromatic ring. For compounds displaying MICs ≤ 4 g/mL against Pae¸ solubility at
pH 7 and distribution coefficient (logD ) were determined. Compound 12 appeared to be less
7
.4
soluble at pH 7 and more lipophilic than pyridone 7 (logD7.4 1.8 and 1.1 for compounds 12 and 7,
respectively). Compounds 10 and 14 were poorly soluble at pH 7 (~5 g/mL) and displayed high
lipophilicity (logD7.4 ~2) (Table 2). At pH 9, all compounds exhibited higher solubilities as the
hydroxamate moiety is predominantly ionized (measured pKas of hydroxamic acids usually range
34
between 7.3 and 7.7). However, compound 12 remained less soluble than pyridone 7 as this pH.
1
8
ACS Paragon Plus Environment

Page 19 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
MeO S
2
H
N
F
OH
CORE
O
MeO
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
Table 2: Antibacterial activities, binding affinities and physico-chemical properties of compounds
-14.
7
Solubility
g/mL)
pH 7 pH 9
MIC (g/mL)a
Tm (K)b
Compounds
logD
(
7
.4
N°
CORE
Eco
Kpn
0.25
Pae
PAO1
PA2616
Shift Eco
Pae
O
7
N
≤0.06
0.5
0.25
≤0.06
≥4x
7.8
11.8
1.1
130
692
9
N
0.5
4
8
2
4
2
≤0.06
≥64x
8.2
9.1
11.0
11.0
2.2
1.9
7
5
66
22
O
O
O
1
0
≤0.06 ≤0.06
0.25
8x
N
O
N
1
1c
32
0.12
4
>32
0.25
16
32
1
>32
0.5
16
0.12
8
NM
4x
5.6
8.1
6.0
9.3
7.0
11.5
6.9
ND
37
ND
5
N
O
1
2
N
N
1.8
2.1
302
N
O
13
>32
1
>32
0.5
NM
4x
N
N
O
S
1
4
≤0.06
0.25
0.12
12.4
36
N
aEco ATCC 25922 (wt); Kpn T 6474 (MDR); Pae ATCC 27853 (wt); Pae PAO1 (wt); Pae PA2616, laboratory
(mexAB-OprM) mutant with PAO1 background; MIC efflux shift: calculated as the ratio of MIC PAO1/MIC PA2616.
DSF on isolated enzyme, determined at 25 µM compound concentration. tested as a racemate. ND: not determined;
NM: not meaningful

b
c
1
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 83
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
Hit-to-Lead Optimization. The next stages towards identifying an LpxC inhibitor suitable for
the treatment of GNB infections consisted of chemical exploration of the hits generated by scaffold
hopping (i.e. hits 10, 12 and 14), and then selection of the most suitable series for further lead
optimization. Antibacterial potency and spectrum, as well as aqueous solubility were selected as
main parameters to guide the hit-to-lead optimization.
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
Upon examination of the co-crystal structures of isoindolin-1-one 10 and pyrrolo-imidazolone
1
2 (Figures S-6 and S-7 in SI, respectively) bound to Pae LpxC,30 identical electrostatic
interactions of the methylsulfone hydroxamate head group of both entities to the catalytic site were
22
found to those previously described. As expected, both bicyclic cores of compounds 10 and 12
are located at the entrance of the hydrophobic substrate-binding passage, projecting the side
aromatic rings into the space defined by LpxC insert II. The superimposition of LpxC structures
bound to oxazol-2-one (R)-8 (binding affinity to Pae LpxC Tm: 9.9 K) and 10 (Tm: 11.6 K)
showed that the tighter complex formed with isoindolin-1-one 10 arose mainly from stronger
interactions between the α-helix of insert II and the inhibitor (Figure 6). The entire α-helix was
pulled closer towards the inhibitor by 1.9 Å (distance measured between Cα atoms of Ile197,
relatively to an all-Cα-fit of both structures with a Cα-RMSD of 0.239 Å) and therefore closed the
hydrophobic tunnel to a tighter fit around the molecule compared to what had been observed for
compound (R)-8.
2
0
ACS Paragon Plus Environment

Page 21 of 83
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
Figure 6: Superimposition of oxazol-2-one (R)-8 (white carbons) and isoindolin-1-one 10 (cyan
carbons) bound to Pae LpxC. The α-helix of insert II (and Ile197 side chain, colored in cyan)
moved closer to the inhibitor in case of 10.
1
9
From these observations and previous SAR studies conducted on CHIR-090 3 analogues, we
anticipated that the composition of the lipophilic tail that spans the hydrophobic passage would
greatly influence the binding affinity and therefore the antibacterial activity of LpxC inhibitors.
The lipophilic tails of the disclosed LpxC inhibitors with antibacterial activity can be divided in
three segments (A, B and R), each occupying well-defined binding regions in or outside the
substrate-binding passage of LpxC enzymes (Figure 7a). The substructure element directly linked
to the bicyclic core, henceforth dubbed segment A, is located within the narrowest part of insert II
hydrophobic passage. In the threonyl-hydroxamate series, broad and enhanced antibiotic activity
could be achieved for inhibitors featuring a small acetylenic unit (van der Waals (vdW) volume:
3
35
3
17
3
8Å per C H unit) in place of a larger phenyl ring (vdW volume: 80.3Å per C H unit). In
2 2 6 6
2
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 83
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
order to similarly maximize vdW interactions with LpxC insert II, an acetylenic unit was set as a
3
6
recurrent segment A for the remainder of the present study. The second segment of the lipophilic
tail (chemical substructure element B, Figure 7a) protrudes from the hydrophobic passage and
resides at the interphase of the distal end of the hydrophobic passage and the surface of the enzyme
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
(see Figure 7b). Only two previously exemplified hydrophobic groups (ACHN-975 1 and CHIR-
0
90 3; see Figure 1) varying in volume and length were initially considered: a phenyl ring (2.83Å)
and an acetylene unit (1.22Å). The third substructure element R (Figures 7a,b), positioned at the
distal end of the inhibitor is partly or totally exposed to solvent, thus providing an excellent means
to increase chemical diversity and to modulate physicochemical properties of inhibitors. Efficient
synthetic access to designed LpxC inhibitors could be achieved by coupling reactions involving
appropriate partners carrying side chains (defined as the combination of structural elements B and
R) and three terminal alkyne building blocks 29, 34 and 40, differing from the nature of their core
I, II and III respectively (see Scheme C4). To initiate a study on the effects of segments B and R
on antibacterial activity, a few archetypical side chains (SC1-4) were selected as they were easily
accessible from commercially available building blocks (including CHIR090 3 morpholine-based
side chain SC1). For all compounds reported in the following tables, antibacterial activities on
wild-type Eco and Pae, MDR Kpn and PAO1 strains, as well as binding affinities (thermal shift,

Tm in K) to Eco and Pae LpxC enzymes were determined. For all compounds displaying
MIC ≤ 2 µg/mL against Eco, logD7.4 and solubilities at pH 4, 7 and 9 were also measured from
DMSO stock solutions.
2
2
ACS Paragon Plus Environment

Page 23 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
O
O
MeO S
2
H
N
O
N
N
OH
S
R
B
CORE
SC1
SC2
SC4
core I
N
OH
O
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
SIDE
CHAINS
O
core III
O
N
N
OH
SC3
OH
core II
Table 3: Antibacterial activities, binding affinities and physico-chemical properties of
compounds 15-20a, b.
MIC (g/mL)a
Tm (K) b
Eco Pae
Solubility (g/mL)
Compounds
logD
7.4
Side
chain
pH
7
N°
Core
Eco
0.12
0.25
16
Kpn
0.5
0.5
8
Pae
2
PAO1
Shift
4x
4
9
1
5a
SC1
2
8
4
2
1
1
4
2
1
1
4
2
11.0 13.8 1.8
9.5 12.8 2.3
7.0 11.6 ND
8.5 12.6 1.2
>890
148
>890
1
5b
SC2
SC3
SC4
SC1
SC2
SC3
SC4
SC1
SC2
SC3
SC4
4
16x
16x
8x
4
7
136
I
1
6a
6b
7a
17b
8a
8b
19a
9b
0a
20b
8
ND
97
1
1
2
2
71
802
70
564
621
556
1
≤0.06
0.12
1
0.25
0.25
2
1
>16x 11.9 13.6 1.6
68
0.5
2
8x
16x
8x
11.8 13.7 1.9
8.8 11.3 0.3
9.9 12.3 0.9
13.0 15.5 1.8
13.0 15.5 2.1
7.4 12.2 0.5
9.9 13.6 1.2
110
II
1
>740 >740 >740
1
0.5
1
1
398
>900
69
>710 >710
≤0.06
0.12
2
0.12
0.25
4
1
8x
155
83
571
613
1
0.5
4
8x
III
2
16x
8x
>760 >760 >760
>740 >740 >740
0.5
2
1
a
Eco ATCC 25922 (wt); Kpn T 6474 (MDR); Pae ATCC 27853 (wt); Pae PAO1 (wt); MIC efflux shift: calculated
as the ratio of MIC PAO1/MIC PA2616. DSF on isolated enzyme, determined at 25 µM compound
concentration. ND: not determined.
b
All derivatives featuring side chains SC1 and SC2 (B = phenyl) displayed good binding affinity
to Eco LpxC (Tm ranging from 9.5 to 13 K), translating to good antibacterial activity against Eco
and Kpn strains with MICs ≤ 0.5 µg/mL (Table 3). Derivatives 17a, b and 19a, b elicited good
2
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 83
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
MICs against Pae as well (0.5-1 µg/mL range), correlating with good binding affinities to Pae
LpxC (Tm ranging from 13.6 to 15.5 K). While featuring different cores, pairs of derivatives
17a/19a and 17b/19b displayed identical antibacterial profiles against GNBs, even though larger
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

Tm values (+1.9 K) were obtained for thienopyrrolones on Eco and Pae LpxC enzymes. These
larger Tm values of thienopyrrolones 19a and 19b tracked with their higher lipophilicity (higher
logD ) (Table 3). Groups R present in SC1 and SC2, had a negligible effect on biochemical and
7
.4
antibacterial activities when comparing 17a to 17b and 19a to 19b. Isoindolin-1-one derivatives
15a, b elicited slightly higher MICs (2-fold elevation) against Eco, Kpn and Pae strains than their
analogues featuring either core II or III. The weaker antibacterial activity of isoindolin-1-ones 15a,
b may be due to weaker binding to LpxC enzymes of these compounds compared to their
analoguous inhibitors 17a, b or 19a, b (featuring cores II or III) (Table 3). This may result from
the marked difference that exists between the orientation of the two exit vectors around core I
relative to cores II and III (see Figure 4). No efforts to further rationalize this observation were
undertaken. Solubilities of compounds 15a and 19a featuring morpholine-containing side chain
SC1 were higher within the assessed pH range than those of derivatives 15b and 19b featuring
alcohol-based side chain SC2. Pyrrolo-imidazolone derivatives 17a and 17b featuring side chains
SC1 and SC2, respectively shared similar solubility profiles. Morpholine-derivatives 15a, 17a and
19a had consistently lower logD7.4 than alcohols 15b, 17b and 19b.
For derivatives 16a, b, 18a, b and 20a, b featuring alkyne-based side chains SC3 and SC4 (B =
acetylene), MICs determined against Eco and Kpn were within a wider range (0.5-16 g/mL) than
observed for analogues 15a, b, 17a, b and 19a, b (B = phenyl ring). Derivatives 16a, 18a, and 20a
featuring the side chain SC3 displayed weaker antibacterial activities with MICs usually
≥
2 µg/mL. Binding affinities of these derivatives to Eco (Tm ranging from 7 to 8.8 K) and Pae
2
4
ACS Paragon Plus Environment

Page 25 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(
Tm ranging from 11.3 to 12.2 K) LpxC enzymes were lower than those of analogues featuring
side chain SC4. The comparison of antibacterial activities recorded for pairs of derivatives 15b-
16b, 17b-18b and 19b-20b featuring side chains SC2 and SC4 (differing only from the nature of
segment B) showed that MICs against Eco and Kpn strains were about 4-fold higher for analogues
featuring a bis-acetylenic tail (wherein B = acetylene). Similarly, binding affinities (ΔTm)
recorded for compounds 16b, 18b and 20b to Eco LpxC were lower than those featuring side chain
SC2. For the Pae strain, no significantly different antibacterial activities could be observed for
compounds differing only in the nature of segment B. Bis-acetylenic derivatives were also
sensitive to efflux as indicated by MIC shifts (8 and 16-fold). While the preliminary antibacterial
activity profiles of bis-acetylenic derivatives determined against three representative GNB strains
seemed slightly inferior to those of phenyl-containing analogues, their solubilities, determined
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
between pH 4 and 9, were consistently higher, matching their lower lipophilicity (lower logD7.4
values). With the exception of derivative 15a, isoindolin-1-ones were less soluble at pH 7 than
their direct analogues based on cores II and III. Because of their inferior antibacterial and solubility
profiles, the isoindolin-1-one series was deprioritized.
2
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 83
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)
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 7: a) and b) Views of inhibitor 17a bound to P. aeruginosa LpxC (2.0 Å) (PDB code: 6I49).
The lipophilic tail is subdivided into three segments A, B and R. From the acetylenic unit A, the
side chain resulting from the combination of segment B (here a phenyl ring) and segment R, here
an N-methyl-morpholine group protruding out of the substrate-binding passage of Pae LpxC.
Several conformations/positions for the morpholine ring of 17a are possible, the depicted one
serving as an example. Partial ligand electron density is represented as a blue mesh (2Fo-Fc density
contoured at 1 sigma).
The binding mode of compound 17a to Pae LpxC was determined by X-ray crystallography.30
The hydroxamate-based headgroup of 17a was bound to zinc in an identical mode to that of
previous analogues. As illustrated in Figure 7a, the acetylenic unit (segment A) of compound 17a
occupied the hydrophobic passage, while the phenyl ring (segment B) was located close to the
surface of the protein, at the exit to the hydrophobic passage. The morpholine ring of 17a was
located on the surface of the enzyme, fully exposed to solvent, turnable and not visible in the
electron density map (see Figure 7b). The analysis of the crystal structure of 17a revealed one
2
6
ACS Paragon Plus Environment

Page 27 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
significant difference in comparison to the structures of hits 8, 10, and 12. While the cores of
compounds 8, 10, and 12 were coplanar with the peptide unit connecting Phe191 and Gly192,
forming a π–π stacking interaction, the core of 17a was inclined and therefore not able to stack on
these peptide bonds (Figure 8a). However, another π–π stacking interaction could be observed for
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
7a involving the phenyl moiety and the peptide bond between Gly209 and Ser210 at the outer
part of the hydrophobic tunnel, close to the protein surface (Figure 8b). This additional interaction
could compensate for the former loss of the stacking, leading to decent binding affinity (Eco Tm
~11.9 K and Pae Tm ~13.9 K for 17a).
a)
b)
Figure 8: a) Superimposition of cores of compound 17a (green carbons) and compounds 8, 10,
12, and 18d (all grey carbons). Phe191 and Gly192 peptide bonds on right hand side. b) View of
the B phenyl moiety of 17a and its stacking over the peptide bond between Gly209 and Ser210.
1
2
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 83
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
Figure 9: Superimposition of inhibitor 17a (orange carbons) with CHIR-090 3 (gray carbons)
bound to Pae LpxC (PDB code: 5VWM).
The superimposition of 17a and CHIR-090 3 structures (PDB accession code 5VWM) showed
that the core of compound 17a resided at the entrance of the hydrophobic passage, while the phenyl
core of CHIR-090 3 was seated deeper inside the tunnel (Figure 9). As the morpholine ring of 17a
was pushed 0.6 Å further outwards we synthesized shorter bis-acetylenic derivatives to take
20
advantage of the shallow hydrophobic pocket filled by the morpholine ring of CHIR-090 3. Pairs
of derivatives differing only in the nature of the structural element B (either a phenyl or and
acetylenic unit) were thus produced. Diverse R-groups were attached according to previously
reported favorable properties such as substituted cyclopropyl rings (featured in ACHN-975 and
3
7
analogues). The most readily accessible pyrrolo-imidazolone series was selected to conduct the
study.
2
8
ACS Paragon Plus Environment

Page 29 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
O
N
R
B
N
SO Me
2
(
R)
NH
OH
B1
B2
O
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
7a, c-j
18a, c-j
Chemical structures of groups R
HO
O
O
MsHN
R8
N
O
HO
N
HO
HO
NH HCl OH
HO
R3
2.
O
OH
R1
R2
R4
R5
R6
R7
R9
Table 4: Antibacterial activities, binding affinities and physico-chemical properties of novel LpxC
inhibitors 17a, c-g and 18a, c-g
MIC (g/mL) a
Tm (K) b
Solubility (g/mL)
Compounds
R
log
pH
D
7.4
N°
17c
B
Eco
Kpn
Pae
PAO1
Shift Eco Pae
4
7
9
B
B
B
1
0.12
1
0.5
2
0.5
2
1
4
1
16x
16x
≥8x
10.8 13.0 1.2
8.8 11.3 0.3
11.9 13.9 1.6
195
>740
802
>820
>740
68
>820
>740
621
R1
R2
1
1
8a
7a
2
1
0.12
0.25
1
18c
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
2
2
0.25
1
2
1
0.5
1
4x
4x
8x
9.6 11.8 0.7
11.8 13.6 2.0
9.9 12.5 0.9
>780
5
>780
7
>780
56
1
1
7d
8d
0.12
0.25
0.5
0.5
R3
R4
R5
R6
R7
R8
R9
149
73
384
112
602
24
553
568
552
170
628
52
1
1
7e
8e
≤0.06 ≤0.06 0.5
0.12 0.25 0.5
≤0.06 ≤0.06 0.5
1
≥16x 12.1 13.4 2.0
≥8x 10.8 13.6 1.1
≥16x 12.1 13.4 1.4
0.5
1
600
11
1
1
7f
8f
0.12
0.12
0.25
0.5
8
0.25 0.25
0.5
0.5
0.5
1
≥8x
4x
11.2 13.2 1.3
11.3 12.6 1.6
10.0 11.6 0.7
11.1 12.7 0.5
7.3 10.3 ND
11.2 13.2 1.0
8.0 10.8 ND
11.8 14.2 1.7
9.4 12.9 0.9
600
589
571
167
588
32
17g
8g
17h
8h
17i
8i
0.25
0.25
1
0.5
0.5
1
1
≥8x
>8x
16x
8x
418
605
ND
3
536
660
1
8
4
8
0.25
2
0.5
4
1
1
1
56
1
4
1
4x
ND
15
1
1
7j
8j
0.25
0.5
0.5
1
2
2
8x
<1
251
2
2
8x
641
785
>880
a
Eco ATCC 25922 (wt); Kpn T 6474 (MDR); Pae ATCC 27853 (wt); Pae PAO1 (wt); MIC efflux shift: calculated
as the ratio of MIC PAO1/MIC PA2616. DSF on isolated enzyme, determined at 25 µM compound concentration.
ND: not determined
b
2
9
ACS Paragon Plus Environment