Imidazopyrazinones (IPYs): Non-Quinolone Bacterial Topoisomerase Inhibitors Showing Partial Cross-Resistance with Quinolones

Article abstract of DOI:10.1021/acs.jmedchem.7b01892

In our quest for new antibiotics able to address the growing threat of multidrug resistant infections caused by Gram-negative bacteria, we have investigated an unprecedented series of non-quinolone bacterial topoisomerase inhibitors from the Sanofi patrim

Full text of DOI:10.1021/acs.jmedchem.7b01892

Subscriber access provided by - Access paid by the | UCSB Libraries
Imidazopyrazinones (IPYs): non-quinolone bacterial topoisomerase
inhibitors showing partial cross-resistance with quinolones
Frédéric Jeannot, Thomas Taillier, Pierre Despeyroux, Stephane Renard, Astrid
Rey, Michaël Mourez, Christoph Poeverlein, Imène Khichane, Marc-Antoine Perrin,
Stéphanie Versluys, Robert A. Stavenger, Jianzhong Huang, Thomas Germe,
Anthony Maxwell, Sha Cao, Douglas L. Huseby, Diarmaid Hughes, and Eric Bacqué
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.7b01892 • Publication Date (Web): 29 Mar 2018
Downloaded from http://pubs.acs.org on March 31, 2018
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 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Imidazopyrazinones (IPYs): nonꢀquinolone
bacterial topoisomerase inhibitors showing
partial crossꢀresistance with quinolones
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
†
†
†
†
†
Frédéric Jeannot, Thomas Taillier, Pierre Despeyroux, Stéphane Renard, Astrid Rey,
†
‡
¥
¥
Michaël Mourez, Christoph Poeverlein, Imène Khichane, MarcꢀAntoine Perrin, Stéphanie
§
┴
┴
Ⱶ
Ⱶ
Versluys, Robert A. Stavenger, Jianzhong Huang, Thomas Germe, Anthony Maxwell, Sha
#
#
#
,†
Cao, Douglas L. Huseby, Diarmaid Hughes, and Eric Bacqué*
†
Sanofi R&D, Therapeutic Area Infectious Diseases, 1541 Avenue Marcel Mérieux, 69280
Marcy L’Etoile, France
‡
SanofiꢀAventis Deutschland GmbH, R&D, Integrated Drug Discovery, Industriepark
Hoechst, 65926 Frankfurt am Main, Germany
¥
Sanofi R&D LGCR, Analytical Sciences, 13 Quai Jules Guesde, 94400 Vitry sur Seine,
France
§
Evotec France, 195 route d’Espagne, BP 13669, 31036 Toulouse Cedex 1, France
┴
Antibacterial DPU, GlaxoSmithKline, 1250 Collegeville Rd., Collegeville, Pennsylvania,
PA19426, USA
Ⱶ
Department of Biological Chemistry, John Innes Centre Norwich Research Park, Norwich
NR4 7UH, UK
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 59
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
#
Uppsala University, Department of Medical Biochemistry and Microbiology, Biomedical
Center (Box 582), Uppsala Sꢀ751 23, Sweden
ABSTRACT
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
In our quest for new antibiotics able to address the growing threat of multiꢀdrug resistant
infections caused by Gramꢀnegative bacteria, we have investigated an unprecedented series of
nonꢀquinolone bacterial topoisomerase inhibitors from the Sanofi patrimony, named IPYs for
ImidazoPYrazinones, as part of the Innovative Medicines Initiative (IMI) European Gram
Negative Antibacterial Engine (ENABLE) organization. Hybridization of these historical
compounds with the quinazolinediones, a known series of topoisomerase inhibitors, led us to
a novel series of tricyclic IPYs that demonstrated potential for broad spectrum activity, in vivo
efficacy and a good developability profile, although later profiling revealed a genotoxicity
risk. Resistance studies revealed partial crossꢀresistance with fluoroquinolones (FQs)
suggesting that IPYs bind to the same region of bacterial topoisomerases as FQs and interact
with at least some of the keys residues involved in FQ binding.
INTRODUCTION
Bacterial topoisomerases (BTs) are critical enzymes for the control of DNA topological state
1
within bacteria and are essential for replication and transcription . Exploitation of the two
related type II BTs, DNA gyrase and topoisomerase IV, as antibacterial targets, has led to
1
2
different classes of antibiotics such as the highly popular fluoroquinolones (FQs) and the
3
less successful coumarins (Figure 1). Considering the paucity of new validated antibacterial
targets, the difficulties to discover new scaffolds active against Gramꢀnegative bacteria, and
4
the increasing levels of resistance to FQs , BTs remain attractive targets in the fight against
resistant bacteria, provided crossꢀresistance to FQs can be avoided. In line with this analysis,
recent years have witnessed a continuous flow of publications of novel BTs inhibitors (BTIs),
ACS Paragon Plus Environment

Page 3 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
5
structurally related or not to quinolones (Figure 1), binding to the different subꢀunits of the
two BTs and acting as either cleavageꢀcomplex stabilizing agents (like FQs) or catalytic
8
ꢀ10
inhibitors (like coumarins). The most advanced compounds that have emerged from these
endeavors have progressed to clinical trials but none has reached market to date.
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
Cleavage-complex stabilizing agents
O
O
O
O
O
F
F
F
XH
O
OH
N
NH
R1
R1
N
N
R
S
R'
N
R
N
N
R
R2
R2
Fluoroquinolones (FQs)2
Isothiazoloquinolones6
R'= NR R or Ar/Het
Quinazolinediones (QZDs)7
1
2
X= NH or O
HN
N
O
O
S
H
N
O
N
N
O
N
N
S
OH
O
NH
F
O
N
HO
O
O
O
O
N
N
N
O
O
H
N
F
Gepotidacin (GSK2140944)8
NXL1019
ETX091410
3
Catalytic inhibibitors (GyrB/ParE binders)
O
O
O
O
O
H
N
N
O
HN
N
H
N
N
F
O
O
OH
H
N
O
O
OH
N
OH
N
O
N
N
NH2
HN
N
H N
2
OH
Novobiocin (Coumarin class)
VXc-486
GP-4
Figure 1. Structures of various Bacterial Topoisomerase Inhibitors (BTIs)
As part of our efforts to identify new antibacterials active on Gramꢀnegative bacteria, we have
given a high priority to the discovery of a nextꢀgeneration of nonꢀquinolone bacterial
topoisomerase inhibitors. With this objective in mind, we decided to revisit a proprietary
series of tetracyclic imidazopyrazinones (1), nicknamed IPYs (Figure 2). This series,
structurally unrelated to quinolones, had been identified, in the mid 1990’s, by RhôneꢀPoulenc
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 59
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
1
Rorer, using the SOS Chromotest (a cellꢀbased SOS induction assay in E. coli), an assay that
1
2
has also delivered other DNA replication inhibitors including the RhôneꢀPoulenc Rorer hit
9
that was to be ultimately optimized into NXL101 . At that time, IPYs had been demonstrated
to be topoisomerase inhibitors, stabilizing the gyraseꢀDNA cleavage complex with some
analogs displaying potent antibacterial activities, albeit only against Gramꢀpositive bacteria.
The optimization program that was conducted on the initial hit failed to deliver a clinical
candidate due to modest in vivo efficacy and poor exposures in mice for the most potent
analogs. Furthermore, a brief program exploring a simplified bicyclic IPY subꢀseries (Figure
2) did not yield compounds with reasonable in vitro antibacterial activities. As a consequence
of these different failures, the project was finally discontinued and the IPY series was never
patented nor published.
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
Alkyl/Aromatic/Hetero-Aromatic
N
O
N
N
N
O
N
N
Ri
Ri
Alkyl/c-Alkyl/Aromatic
Tetracyclic IPYs 1
Alkyl/c-Alkyl/Aromatic
Bicyclic IPYs 2
Figure 2. Structures of the different historical IPY subꢀseries
In this paper, we report our recent investigations of the IPY series aimed at identifying novel
BTIs active on Gramꢀnegative bacteria. We describe the steps that led us from the initial
tetracyclic parent IPYs to the design of a new, promising tricyclic IPY hit that displayed
activities against effluxꢀcompetent Gramꢀnegative bacteria. We present the exploration
program developed around the initial hit, the antibacterial and ADMET profiles of the best
analogs as well as the in vivo properties (tolerability, efficacy, and PK) of selected analogs
and the results of the resistance studies that unfortunately unveiled the partial crossꢀresistance
of the series with FQs. This work was mainly conducted as part of a collaboration between
ACS Paragon Plus Environment

Page 5 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
GSK and Sanofi, within the European IMI New Drugs for Bad Bugs (ND4BB) consortium
1
3
ENABLE .
RESULTS AND DISCUSSION
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
Chemistry. Compounds 1a and 1b were prepared as shown in Scheme 1. Reaction of
14
indanone 5 with methylamine led to the tetracyclic IPY 6 that was directly converted into 1a
by a Knoevenagelꢀlike reaction with 4ꢀacetamido benzaldehyde. Compound 1b was obtained
15
in one step by Nꢀmethylation of the known tetracyclic IPY 7 .
a
Scheme 1. Synthesis of the historical tetracyclic IPYs 1a and 1b
O
O
HN
HN
O
N
O
CHO
a
N
N
N
N
N
O
N
N
b
O
O
5
6
1a
N
N
c
N
O
N
N
O
N
N
H
N
7
1b
a
2
Reagents and conditions: (a) MeNH , EtOH, AcOH, reflux; (b) Piperidine, MeOH, reflux; (c) NaH,
MeI, DMF, 70 °C.
Compound 2a was synthesized according to Scheme 2. Imidazole carboxylate 8 was Nꢀ
alkylated to afford compound 9. Reaction with Bredereck’s reagent led to intermediate 10 that
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 59
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
was reacted with methylamine to afford the bicyclic IPY 11. Palladiumꢀcatalyzed transfer
hydrogenation followed by acylation successively delivered compounds 12 and 2a.
a
Scheme 2. Synthesis of the historical bicyclic IPY 2a
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 N
2
O N
2
O N
O N
2
Br
2
N
O
O
b
N
c
N
N
N
N
N
N
O
N
H
a
O
N
O
O
O
8
9
10
11
H
N
d
H N
2
e
N
N
N
O
N
N
N
O
Cl
O
12
2a
a
Reagents and conditions: (a) Acetonitrile (MeCN), K
reagent), toluene, 110°C, 20 h; (c) MeNH , MeOH, AcOH, 85°C, 30 min; (d) Pd/C (cat.), MeOH,
HCO H, EtOH, 65°C, 4 h; then concentrated HCl; (e) 2ꢀ(4ꢀClPh)CH COCl, DCM, pyridine, RT.
2 3 2 2
CO , RT; (b) HC(NMe ) OtBu (Bredereck’s
2
2
2
The synthetic routes to compounds 3aꢀc and 3eꢀi are shown in Scheme 3. 5ꢀbromoꢀ2,4ꢀ
difluoroaniline and (1ꢀethoxycyclopropoxy)trimethylsilane were first reacted to afford hemiꢀ
aminal 13. The crude reaction mixture was then reacted with sodium borohydride, in the
presence of trifluoroboron diethyl etherate, to afford aniline 14 that was then acylated by
various alkyl imidazo or triazolo carboxylates, in the presence of trimethylaluminium, to
afford amides 15aꢀc. Cyclisation of these intermediates into compounds 16aꢀc was triggered
by reaction with 1,8ꢀDiazabicyclo[5.4.0]undecꢀ7ꢀene (DBU), in DMF. Starting from 16aꢀc,
we prepared the targeted compounds relying either on Suzuki coupling reactions (compounds
3eꢀi) or Buchwald–Hartwig aminations (compounds 17aꢀc). In the later cases, compounds 3aꢀ
c were obtained by removing the tꢀbutyloxy carbonyl protecting group under acidic
conditions. In the cases of triazolo compounds 3b and 3f, the ambiguity on the identity of the
generated isomers was cleared by an XꢀRay structure (see Figure S1 in Supporting
ACS Paragon Plus Environment

Page 7 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Information) of 3f that demonstrated that cyclisation into 16b had occurred through the
nitrogen at position 1 of the 1,2,4ꢀtriazolo ring.
a
Scheme 3. Synthesis of the novel tricyclic IPYs 3a, 3b, 3c, 3e, 3f, 3g and 3i
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
O
O
Si
F
F
F
F
F
F
b
F
F
c
O
H
Br
NH
O
Br
NH
N
Br
NH2
a
Br
N
W
Y
X
1
4
15 a-c
13
W
N
X
Y
F
W
N
X
Y
W
N
X
Y
O
F
F
O
d
N
N
O
f
e
N
H
N
N
O
Br
N
O
H N
2
17 a-c
16 a-c
3
a-c
a W=X=CH, Y=N
b W=Y=N, X=CH
c W=Y=CH, X=N
W
X
Y
F
N
g
R
N
O
e W=X=CH, Y=N, R=pyridin-4-yl
f W=Y=N, X=CH, R=pyridin-4-yl
g W=Y=CH, X=N, R=pyridin-4-yl
h W=X=CH, Y=N, R=pyridin-3-yl
i W=X=CH, Y=N, R=pyridin-2-yl
3
e-i
a
Reagents and conditions: (a) acetic acid, MeOH, reflux 5 h; (b) BF3ꢀ(C
2
H )
5 2
4
O, NaBH , THF
5
°C to reflux, 4 h; (c) ethyl or methyl imidazole or triazole carboxylate, AlMe
3
(2M in
toluene), Et N, toluene, reflux 2 h; (d) DBU, DMF, 120°C, 3 h; (e) (S)ꢀtertꢀbutyl pyrrolidinꢀ3ꢀ
3
ylcarbamate, 2,2’ꢀbis(diphenylphosphino)ꢀ1,1’ꢀbinaphtyle (BINAP), Pd(OAc) , Cs CO
2
2
3,
toluene, microwaves, 120°C, 20 h; (f) TFA, DCM, RT, 12 h; (g) pyridinylꢀ4ꢀboronic acid,
Pd(PPh , K CO , MeOH/Toluene (1:2), microwaves, 120°C, 1 h or pyridinꢀ3ꢀyl boronic
acid, Pd(PPh , Cs CO , water/1,4ꢀdioxane (1:2), 100°C, 4 h or 2ꢀ(tributylstannyl)pyridine,
Pd(PPh , water/1,4ꢀdioxane (1:2), microwaves, 150°C, 2 h.
3
)
4
2
3
3
)
4
2
3
3 4
)
The synthesis of compound 3d is shown in Scheme 4. Compound 18 was obtained by reaction
of cyanamide, triethyl orthoformate and ethyl 2ꢀ(benzylamino)acetate in the presence of
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 59
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
sodium methoxide, followed by in situ transꢀesterification of the expected ethyl ether into the
methyl ester. Nitrosation of the primary amine of 18 afforded the corresponding diazonium
salt that was then converted, in the presence of potassium iodide, into the iodoꢀimidazole 19.
Suzuki coupling then led to ester 20 that was next hydrolyzed into acid 21 which, in turn, was
transformed into amide 22 by coupling with cylopropylamine. Cyclisation of the later
intermediate was carried out in the presence of DBU to yield compound 23. Subsequent
Buchwald–Hartwig amination and deprotection of the tꢀbutyloxy carbonyl protecting group
successively provided compounds 24 and 25. 3d was finally obtained following
hydrogenolysis of the benzylic protecting group still present in compound 25.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
Scheme 4. Synthesis of the novel tricyclic IPYs 3d
O
O
N
a
b
c
N
N
N
N
d
H N
N
F
2
N
H N
O
I
O
2
O
O
Br
F
18
19
20
R
N
N
N
N
O
N
N
f
F
N
O
i
NH
O
g
F
F
F
Br
N
O
N
N
N
N
R'
H N
2
Br
F
2
1 R=OH
e
22 R=HNcyclopropyl
2
3
24 R' = NHboc
5 R'= NH2
3d
h
2
a
Reagents and conditions: (a) HC(OEt) , reflux, 2 h and then ethyl 2ꢀ(benzylamino)acetate,
3
Et
difluorophenyl)boronic acid, Pd(PPh
d) NaOH 6M, THF/water (1:1), RT, 1 h; (e) cyclopropylamine, Oꢀ(7ꢀazabenzotriazolꢀ1ꢀyl)ꢀ
N,N,N′,N′ꢀtetramethyluronium hexafluoro phosphate (HATU), DIEA, THF, RT, 12 h; (f)
2
2
O, NaOMe, RT, 1 h; (b) NaNO , HCl, KI, water, 0°C to reflux, 24 h; (c) (4ꢀbromoꢀ2,5ꢀ
3
)
4
, K CO , MeOH/PhMe (1:2), microwaves, 1 h, 120°C;
2
3
(
DBU, DMF 120°C, 24 h; (g) (S)ꢀtertꢀbutyl pyrrolidinꢀ3ꢀylcarbamate, BINAP, Pd(OAc)
Cs CO3, toluene, microwaves, 150°C, 5 h; (h) HCl 2M in Et O, RT, 12 h; (i) H CO NH
Pd/C, MeOH, reflux, 6 h.
2
,
,
2
2
2
2
3
ACS Paragon Plus Environment

Page 9 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a
Scheme 5. Synthesis of compound 4
OH
HN
HN
N
N
HN
N
a
b
c
O
O
N
O
O
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
26
27
28
4
a
Reagents and conditions: (a) MeNH
2
, AcONa, reflux, 20 h; (b) AlCl
c) 4ꢀpyridyl carboxaldehyde, piperidine, DMSO, 100°C, 16 h.
3
, xylene, 100°C, 10 min;
(
16
Compound 4 has been prepared as shown in Scheme 5. Compound 26 was reacted with
methylamine to afford the tricyclic 2ꢀpyridone 27. Cleavage of the benzylic side chain by
reaction with aluminum trichloride led to 28 that was next converted, upon Knoevenagelꢀlike
condensation with 4ꢀpyridyl carboxaldehyde, into compound 4.
Evaluation of historical IPYs (compounds 1 and 2). We started our investigations with
profiling a selection of historical tetracyclic IPYs. Across a primary panel of representative
Gramꢀpositive and Gramꢀnegative strains, we confirmed that compounds such as 1a and 1b
displayed good activities (Table 1) on Gramꢀpositive bacteria such as Staphylococcus aureus
(
see in Table 1, the MIC of 1a on the Methicillinꢀresistant Staphylococcus aureus (MRSA)
strain at 0. 25 µg/ml) but no activities on the effluxꢀcompetent Escherichia coli (E. coli) or
Pseudomonas aeruginosa (P. aeruginosa) strains. For the E. coli strains, this lack of activity
was accounted for by a strong efflux rather than by a lack of affinity for the E. coli
topoisomerases or by poor influx. This conclusion was suggested, as illustrated for 1a, by the
very low MIC on the effluxꢀdeficient E. coli MG1655 ꢁtolC and by the lack of MIC shift on
17a
the hyperꢀpermeable (UDPꢀ3ꢀOꢀacylꢀNꢀacetylglucosamine deacetylase (lpxC) mutant) D22
E. coli strain. More generally, all along our work, we used MICs on the pumpꢀdeleted E. coli
strains as surrogates to estimate target affinity, assuming that only efflux limited net
permeability of the IPYs in wildꢀtype (wt) E. coli. In the case of P. aeruginosa, the reason for
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 59
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
7b
the lack of activity against the wt and the partially effluxꢀdeficient PAO750 strains could
not be easily ascertained. It could be due to efflux by other pumps (not deleted in PAO750),
poor permeability or even poor target binding.
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
For 1a, at 8ꢀfold its MIC, the frequency of resistance (FoR) of the MRSA strain
ꢀ7
ATCC33592 was poor (> 10 ). Cytotoxicity values on the HepG2 cell line, a human
hepatocellular carcinoma cell line, were variable across the series but often quite significant
(
e.g. TC = 8.4 µM for 1a). Regarding the ADME and physicoꢀchemical profile (data not
50
shown), the tetracyclic IPYs such as 1a generally displayed poor solubility in water across a
variety of pHs, poor permeability in Cacoꢀ2 cells, metabolic instability in the presence of
murine or human microsomes and high chemical instability in a DMSO solution in the
presence of thiols such as mercaptoꢀethanol or glutathione (more than 50% degradation within
one hour), whereas it was stable in a pH 7.4 buffer solution. These different issues were
consistent with our historical failure to optimize the series and reach good in vivo efficacy.
They were thought to stem from a combination of high hydrophobicity (clogD at pH 7.4 = ~4
for 1a and 1b), high planarity and the presence of the reactive exoꢀmethylene moiety prone to
act as a Michael acceptor. In addition, compounds such as 1a displayed significant hERG
inhibition (typically >50% inhibition at 10 µM). Overall, there was little hope to optimize the
tetracyclic IPY subseries away from the high hydrophobicity that is generally detrimental to
18
activities on Gramꢀnegative bacteria .
We next turned our attention to the bicyclic IPY subseries (compounds 2) that appeared
more drugꢀlike than the tetracyclic IPYs 1. Historically, the few compounds prepared in this
subꢀseries had shown poor activities on Gramꢀpositive bacteria, with MICs for Staphylococci
strains generally above 32 µg/ml (as for 2a in Table 1), despite cꢀlogP/cꢀlogD at pH 7.4
values compatible with activities against Gramꢀpositive (clogP = 3 in the case of 2a). As
previously, we first profiled a small selection of historical bicyclic IPYs and confirmed that
ACS Paragon Plus Environment

Page 11 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
these compounds were inactive against our panel of Gramꢀpositive and effluxꢀcompetent
Gramꢀnegative strains despite significant inhibition of wt E. coli gyrase in the DNA
supercoiling assay (e.g. IC50 = 6.35 µM for 2a, compared to ciprofloxacin (Cip) IC50 = 0.31
µM). Interestingly, compound 2a displayed poor MIC against the permeable D22 strain but
reasonable activities on the pumpꢀdeleted E. coli strain, suggesting significant target
inhibition and, effluxꢀ rather than influxꢀlimited penetration, hence potential to obtain
activities on wt E. coli, provided efflux could be overcome. In addition, the E. coli ꢀtolC
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
ꢀ9
strain showed a satisfactory frequency of resistance (< 5.7 10 ), when challenged by 2a at
xMIC (4 µg/ml). Based on this attractive preliminary data and with the objective of
4
enlarging the spectrum toward Gramꢀnegative bacteria, we produced several libraries of
bicyclic IPYs displaying an amide, a urea, a carbamate or a sulfonamide instead of the
phenylacetamido moiety of 2a. Unfortunately, the only analogs that displayed activity were
carboxamides very similar to 2a and these were still only active on the efflux pumpꢀdeleted
strains (data and structures not shown), which led us to discontinue efforts on this series.
Design and antibacterial properties of a new series of tricyclic IPYs. Noting that there
was potential structural similarity between the IPYs and the Pfizer quinazolinedione
7
topoisomerase inhibitors (QZD; Figure 1), we theorized that design elements from the QZDs
2
could be beneficial to the IPYs. Even if the hydrogen bond donors OH or NH of the QZDs
were exchanged for a hydrogen bond acceptor in our IPYs, the rest of the pyrazinone moiety,
and in particular the N1ꢀalkyl group and the C2ꢀcarbonyl, superimposed nicely with those of
the QZD scaffold. To get a perfect fit with the QZDs, the indene part of the tetracyclic IPYs
had to be obviously exchanged for a fused benzo moiety bearing the fluoro and the amino
substituents of the QZDs. This reasoning led us to hybridize the tetracyclic IPYs and the
QZDs, as shown in Figure 3, to obtain a new series of tricyclic IPYs.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 59
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
O
N
R
F
NH , OH
F
N
N
N
O
2
N
N
O
Hybridization
N
+
R1
R1
N
O
N
N
R2
R1
R2
QZDs
Tetracyclic IPYs (1)
Novel tricyclic IPYs
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
7
Figure 3. Hybridization of the tetracyclic IPYs with Quinazolinediones (QZDs)
We first synthesized compound 3a fixing the amino group to (S)ꢀ3ꢀamino pyrrolidino (one
7
of the best substituents from the QZD series ) to maximize our chance to generate an active
compound. Lending credit to our design hypothesis, compound 3a was found active, albeit
less so than QZD1, showing modest MICs on the effluxꢀcompetent Gramꢀnegative and the
MRSA strain (Table1). Comparing MICs between effluxꢀcompetent, effluxꢀdefective, and
hyperꢀpermeable strains indicated strong efflux and no significant problem of influx, at least
in E. coli. In addition, the very low MICs on the two pumpꢀdeleted E. coli and P. aeruginosa
strains suggested good inhibition of the corresponding topoisomerases, assuming
topoisomerase inhibitionꢀmediated activities. In line with the latter hypothesis, in a wild type
E. coli gyrase DNA supercoiling assay, an IC50 was measured for 3a at 1.19 µM (0.42 µM for
Cip). Moreover, a clear and selective doseꢀresponse effect on DNA replication was confirmed
for this compound in an E. coli macromolecular synthesis assay (Scheme S1).
We immediately prepared a series of analogs of 3a with modified imidazo moieties
(
compounds 3a to 3d) to probe the influence of this ring (Table 1). As for 3a (clogD pH 7.4 =
0
.3), all these compounds displayed clogD at pH7.4 compatible with activities against Gramꢀ
15
negative (ꢀ0.3 to ꢀ1.1; Table1). Among this set of “imidazoꢀlike” analogs, the triazolo
homolog 3b had similar activities compared to 3a on E. coli and MRSA and showed higher
MICs on the P. aeruginosa strains.
ACS Paragon Plus Environment

Page 13 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 1. MICs of representative IPYs on a panel of Gramꢀpositive and Gramꢀnegative
strains
logD
pH
7.4
MIC (µg/ml)
b
c
d
e
f
g
h
Cpd Structure
Ec
Ec
Ec
Ec
Pa
Pa
Sa
a
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
H
N
1a
4.1
3.8
3
>64
>64
>32
>64
>64
>32
<0.125
<0.125
1
>64
>64
32
>64
>64
>32
>64
>64
>32
0.25
O
N
N
N
O
i
N
1b
>64
N
N
N
O
Cl
2
3
a
a
>32
O
N
N
N
H
O
N
F
N
N
N
0.3
ꢀ0.3
ꢀ1.1
4
8
<0.125
<0.125
1
2
1
32
>64
>64
64
<0.125
8
4
H
2
N
N
O
N
3b
F
N
N
2
4
1
1
H
2
N
N
N
O
N
3c
F
N
N
16
16
16
32
64
32
16
16
8
16
16
4
H
2
N
N
O
N
3d
F
NH ꢀ0.3
0.5
0.5
8
H
2
N
N
N
O
F
N
N
N
3.3
<0.125
>64
3e
O
N
N
F
N
N
3
f
2.6
1.9
32
64
>64
>64
<0.125
0.25
32
64
>64
>64
32
32
16
64
N
O
N
N
F
N
3g
N
O
N
F
F
N
N
O
3
h
3.4
3.3
16
32
<0.125
4
8
>64
>64
4
64
N
N
N
N
N
>64
>64
>64
>64
>64
3
i
N
O
O
N
QZ
D1
F
N
NH
2
ꢀ1.8
1
1
<0.125
0.5
32
<0.125
2
H
2
N
N
O
O
O
F
Cip
OH
ꢀ0.7 <0.125 <0.125 <0.125 <0.125 0.25 <0.125
0.125
N
N
HN
a
b
c
LogD at pH 7.4 as calculated by ACD9; E. coli ATCC35218 (ESBL); E. coli MG1655
d
e
17a
(
K12); E.coli MG1655 ꢀtolC (pumpꢀdeleted); E. coli D22
(hyperpermeable LpxC
(PAO1 with ꢁmexABꢀoprM
f
g
17b
mutant); P. aeruginosa PAO1; P. aeruginosa PAO750
ꢁmexCDꢀoprJ ꢁ mexEFꢀoprN ꢁmexXY ꢁopmH ꢁmexJK ꢁpscC); S. aureus ATCC33592
(MRSA); Unreliable result due to poor solubility; Cip = Ciprofloxacin.
h
i
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 59
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
Compounds 3c and 3d had modest MICs on the MRSA strain similar to that of 3a whereas
they were both less active on the Gramꢀnegative strains than 3a, may be as a result of
decreased permeability. Of note, the lower activities of 3d, a compound designed to be the
closest bioꢀisostere of the quinolones and quinazolinediones, suggested that the analogy with
quinazolinediones was less straightforward than initially envisioned.
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
We then tested 3a and 3b against a small secondary panel of Gramꢀnegative) strains (Table
S2) to provide a more complete picture of the breadth of activities of these compounds against
Gramꢀnegative bacteria. Similar or slightly better activities were confirmed on the E. coli and
P. aeruginosa strains while the effluxꢀcompetent Acinetobacter baumannii (A. baumannii)
and Klebsiella pneumoniae (K. pneumoniae) strains were less sensitive (MICs = 32 or > 64
µg/ml). MICs on the corresponding pumpꢀdefective strains suggested potential on K.
pneumoniae strains (MICs = 0.5 µg/ml), provided efflux could be overcome, but less so on A.
baumannii strains (MICs = 4ꢀ8 µg/ml). This difference may reflect different affinity of the
test compounds for the BTs of these two species or differences in penetration across species.
We also exchanged the pyrrolidino substituent for a pyridine ring, as reported in the
6
isothiazoloquinolone BTI series . 3e, the direct analog of 3a (Table 1), showed a similar MIC
on the MRSA strain. 3e also displayed activity on the E. coli ꢀtolC strain, as good as 3a
suggesting similar affinity for the topoisomerases upon substitution of the amino pyrrolidino
substituent present in 3a for the 4ꢀpyridyl ring of 3e. This deduction was confirmed by the
result of the DNA wild type E. coli gyrase supercoiling assay: 3e showed an IC50 of 2.5 µM,
very close to that of 3a (IC = 1.19 µM) but significantly less potent than that of Cip (IC =
5
0
50
0.42 µM). Unfortunately, good target affinity failed to translate into good MICs on the effluxꢀ
competent E. coli strains, possibly due to increased efflux or decreased permeability.
Compared to 3e, compounds 3f to 3i were much less active on the MRSA strain. Moreover, 3f
ACS Paragon Plus Environment

Page 15 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
and 3g only showed, at best, modest MICs on the effluxꢀcompetent E. coli strains probably
due to either increased efflux (for 3f) or reduced affinity for the target (for 3g). Interestingly,
compound 3h and 3i showed contrasting properties: while 3h retained similar antibacterial
activities on the effluxꢀcompetent and the
inactive (MIC > 64 µg/ml) on the effluxꢀcompetent E. coli strains and displayed modest
activity on the tolC E. coli strain. These differences suggested efficient interactions of the
ꢀtolC E. coli strains compared to 3e, 3i was totally
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
ꢀ
pyridyl nitrogens of 3e and 3h with the E. coli topoisomerases that are lost when the nitrogen
atom is shifted to the 2ꢀposition of the pyridyl moiety. Moreover, in comparison to 3a,
compounds 3e to 3i all showed a lack of activity on the effluxꢀcompetent P. aeruginosa
PAO1 and much higher MICs on the pumpꢀdeleted P. aeruginosa PAO750, suggesting in all
cases a large drop of affinity for the targets with respect to 3a and stressing the importance of
the aminoꢀpyrrolidine moiety of 3a for good affinity for the P. aeruginosa topoisomerases.
Finally, we looked at the serum effect across the new tricyclic IPY series. In the presence of
either 50% calf fetal serum or 50% mouse serum (Table S3), there were no significant MIC
shifts for 3aꢀe, whereas 1a displayed a strong serum effect (MIC shift >32 fold). These
results, consistent with the hydrophobicity of these different compounds, boded well for in
vivo efficacy (see below) of the most potent tricyclic IPYs.
Developability profiling of the new tricyclic IPYs. At this stage, we profiled our best
tricyclic IPYs in various ADMET and physicoꢀchemical assays in order to determine the
developability potential of the series (Table 2). Overall, the profile of the 8ꢀamino derivatives
3a and 3b was good. In addition to their small molecular weights (~330), these compounds
displayed good aqueous solubility even at pH 7.4, low metabolism in microsomes across
species, stability in rodent plasma, minimal CYP3A4 inhibition and even potential for oral
absorption (see 3b). In terms of early toxicity signals, 3a and 3b were devoid of cytotoxicity
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 59
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
in HepG2 cells and only weakly inhibited the hERG channel (IC s > 30 µM). In addition, an
50
in vitro pharmacology assessment of 3b showed minimal issues with only two targets (M1
and M3) out of a panel of 33 enzymes, receptors and transporters inhibited at 10 µM, in the
50ꢀ60% range. The only serious alert that was detected in this profiling was genotoxicity in a
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
micronucleus test (MNT) assay, with or without S9 metabolic activation. This result may be
linked to the mechanism of action of the IPYs (topoisomerase inhibition) and is a usual
1
9
concern in topoisomerase inhibitor programs .
Table 2. Physicoꢀchemistry and ADMET profile of representative IPYs
Compound
Aqueous solubility @ pH 7.4
Cacoꢀ2 Papp (nm.s )
Microsome lability h/m/r (%)
3a
3b
926 µg/ml
32
3e
10 µg/ml
216
13/81/68
15/13 µM
>30 µM
Stable at 4 h
>30 µM
a
c
1132 µg/ml
4
ꢀ
1 b
6/8/11
0/8/8
d
CYP3A4 IC50 Me /T
>30/>30 µM
>30 µM
Stable at 4 h
>30 µM
>30/>30 µM
>30 µM
Stable at 4 h
>30 µM
hERG IC50
f
Plasma stability (m, r)
g
HepG2 (TC50
)
h
MNT w/wo metabolic activation Positive w/wo Positive w/wo Positive w/wo
Thermodynamic solubility. Apparent permeability; results above 20 nm.s suggest good
oral absorption in humans. Percentage metabolized by human/mouse/rat liver microsomes
following 20 min of incubation at 5 µM in the presence of 1mM NADPH. Inhibition in
a
b
ꢀ1
c
d
human liver microsomes of the CYP3A4ꢀmediated metabolism of ether midazolam (M) or
e
testosterone (T). Inhibition of the hERG channel transfected in CHO cells in an automatic
f
patch clamp format, at 20°C. Measured in mouse and rat plasma after 4 h of incubation of 1
µM of the test compounds at 37 °C; the test compounds were considered stable at 4 h if the
g
percentage of stability was within the 80 ꢀ 120% range. Cytotoxicity in the HepG2 cell line
h
after 40 h of incubation. L5178Y mouse lymphoma cells incubated with increasing
concentrations of test articles for 23 h (without S9 fraction) or 3 h (in the presence of S9).
The profile of 3e was less satisfactory than that of 3a and 3b but still acceptable at this
early stage. Probably due to its higher hydrophobicity (clogD pH 7.4 = 3.3), Cacoꢀ2
permeability was excellent and aqueous solubility was much lower while metabolic stability
in the presence of rodent microsomes was poorer and CYP3A4 inhibition was stronger. As for
3a and 3b, 3e was found to be positive in the MNT assay, with or without S9 metabolic
ACS Paragon Plus Environment

Page 17 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
activation. This general genotoxicity alert was perceived as a setback in the otherwise
promising developability profile of the IPYs.
In vivo profiling of the new tricyclic IPYs. Considering the good in vitro ADME
properties of the best tricyclic IPYs, we decided to progress compounds 3b to in vivo mice
studies in order to get an assessment of the tolerability and efficacy of the series, two key
hurdles for any new scaffold active on Gramꢀnegative bacteria. Compound 3b was well
tolerated at 30 mg/kg following iv bolus administration. In a septicemia model using E. coli
ATCC35218, 3b (MIC = 4 µg/ml) showed no significant effect on survival and on bacterial
burden, following iv administration of 2 x15 mg/kg (Table 3). At 2 x30 mg/kg, compound 3b
reduced bacterial burden in the blood (Table 3) but with a limited impact on survival at 96 h.
Higher doses could not be tested because of solubility limitations in the simple aqueous
formulation used.
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
a
Table 3. In vivo efficacy of compound 3b in an E. coli ATCC35218 mouse septicemia
b
model
Survival @ 96 h
n = 10
Log10 CFU/ml of blood
ꢀ logCFU
(
mice/group)
0%
(n = 5 mice/group)
Vehicle
7.6
1.9
6.1
4.5
ꢀ
c
Tienam® ꢀ 15 mg/kg/d
90%
5.8
d
3b – 2 x15 mg/kg
0%
1.6
e
3
b – 2 x30 mg/kg
30%
3.2
a
b
Formulated in sterile water; Infection ip at T
0 0 0
, iv administration of 3b at T +1 h and T + 3
c
h, blood sampling for CFU count at T
.0002 compared with vehicle group; p = 0.0432 compared with vehicle group; p = 0.0233
0
d
+ 4 h and follow up of mice up to T + 96 h. p =
0
e
0
compared with vehicle group.
To shed light on these results, we first measured mouse protein binding by rapid
equilibrium dialysis and found it to be moderate at 62%. We then compared this to a single
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 59
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
dose PK data in mice (iv; 30 mg/kg of 3b formulated in sterile water; see Table S6). Exposure
in plasma (AUC0ꢀlast = 7200 ng.h/ml) was reasonable but halfꢀlife was short (1 h) and the time
during which the free concentration of 3b remained above or close to MIC (4 µg/ml) was no
more than a few minutes (unbound concentration in plasma at 5 min = 2.5 µg/ml; mean value
for 3 animals), which may explain the poor efficacy of 3b in the septicemia model, at tested
doses. Of note, tissue concentrations in lung, liver and kidney were much higher than those
recorded in plasma. For instance, in all sampled tissues, assuming similar protein binding as
in plasma, unbound concentrations remained above MIC for ~2 h, suggesting that efficacy in
tissue infection models might be better, at comparable doses, than that seen in the septicemia
model. However, this study was never conducted due to the resistance issue associated to the
series (see below).
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
Resistance studies. In parallel with the in vitro and in vivo profiling discussed above, we
moved to resistance and crossꢀresistance studies. In order to confirm the good frequency of
resistance of the E. coli ꢀtolC with 2a, we first measured frequencies of resistance for various
wildꢀtype effluxꢀcompetent Gramꢀnegative strains challenged by 3a and 3e (see Table 4).
After 16 h and 48 h of culture in the presence of either test compounds, E. coli, K.
pneumoniae and A. baumannii wildꢀtype strains gave modest to good FoR rates, ranging from
ꢀ7
ꢀ9
~
10 (A. baumanii 48 h) to ~ 10 (E. coli). Results were similar between 3a and 3e for the E.
ꢀ9
coli strain (≤ 3.6 10 ) while there was a oneꢀlog difference in favor of 3a for the K.
ꢀ9
ꢀ8
pneumoniae strain (≤ 2.8 10 compared to 3.9 10 ). On the other hand, for both compounds,
the pumpꢀdeleted ( tolC) and permeable (lpxCꢀmutant) E. coli strains unexpectedly
displayed, for unclear reasons, significantly higher FoR than the more relevant, wt E. coli
ꢀ
ꢀ7
ꢀ9
strains (~10 compared to ~10 ). This difference of behavior couldn’t be confirmed for P.
aeruginosa since MIC levels allowed us to test only the pumpꢀdeleted strain. However, for the
ꢀ7
ꢀ8
latter, FoR were also in the 10 to 10 range for both compounds.
ACS Paragon Plus Environment

Page 19 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 4. Frequencies of resistance of various Gramꢀnegative bacteria at concentrations
(Conc) of 4xMICs of 3a or 3e
Cpd
Species
E. coli
Genotype
Conc.
16 h
48 h
(µg/ml) frequency frequency
ꢀ
9
ꢀ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
ATCC 25922 WT parent
ATCC 25922 ꢀtolC
CGSC 5163, D22 lpxC mutant
32
1
≤ 3.6 10
5.3 10
3.6 10
ꢀ7
E. coli
ꢀ
ꢀ
7
ꢀ7
ꢀ7
E. coli
P. aeruginosa
K. pneumoniae ATCC 13883 WT
E. coli
E. coli
E. coli
P. aeruginosa
K. pneumoniae ATCC 13883 WT
A. baumannii ATCC 19606 WT
8
4
64
32
0.5
2
8
64
64
2.6 10
1.6 10
≤ 2.8 10
≤ 3.6 10
5.3 10
2.6 10
3.7 10
3.1 10
4.5 10
6.7 10
3.4 10
≤ 2.8 10
≤ 3.6 10
ꢀ
4.5 10
8.2 10
3.9 10
4.5 10
3a
ꢀ8
PAO750 Effluxꢀdefective
ꢀ
9
9
ꢀ9
ꢀ9
ꢀ
ATCC 25922 WT parent
ATCC 25922 ꢀtolC
CGSC 5163, D22 lpxC mutant
PAO750 Effluxꢀdefective
ꢀ
7
ꢀ7
ꢀ8
ꢀ7
3e
ꢀ8
ꢀ
8
ꢀ8
ꢀ7
ꢀ
9
We further characterized some of the mutant strains selected in these resistance
experiments (see details in Table S4). We sequenced the whole genomes of five resistant
strains selected from E. coli ATCC25922 (one mutant selected by 3a) and the ꢀtolC E. coli
strain (2 mutants each for 3a and 3e) and looked at the corresponding MICꢀfold increases.
Mutations could be identified only in two out of the three mutant strains selected by 3a and
were found to have no connection with topoisomerases. In one case (ꢀtolC E. coli), resistance
was linked to efflux regulation and resulted in a 32ꢀfold MIC shift. In the other case (E. coli
wildꢀtype ATCC25922), a point mutation was identified in histidine tRNA synthetase hisS
(V334E) and was not associated with any MIC shift. For the mutant where no mutation could
be identified (selected in ꢀtolC E. coli), the MIC shift was 8ꢀfold.
We next turned our attention to crossꢀresistance of the IPYs to FQs and to nonꢀquinolone
8
bacterial topoisomerase inhibitors related to NXL101 . When tested against a panel of
isogenic S. aureus strains engineered to be resistant or hyperꢀsusceptible to such
20
compounds , the MICs of 3a, 3b, 3c but also those of 1a showed no or much smaller changes
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 59
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
8
with respect to the parent strain compared to NXL101 (Table S5), suggesting that IPYs
interact with gyrase and topo IV at a different site in comparison to compounds such as
NXL101. We then looked at the activity of 3a on a panel of S. aureus, K. pneumoniae and E.
21
coli strains engineered to be resistant to FQs (Table 5). To our surprise, 3a showed crossꢀ
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
resistance with FQs. However, the levels of crossꢀresistance remained low to moderate (fold
shifts ranging from 2 to 64) even for those strains that were resistant to highly resistant to Cip
(
fold shifts from >16 to 2000). Contrary to our initial hypothesis, 3a behaved more like FQs
than QZDs since these later compounds have been reported to display no crossꢀresistance with
7
FQs . Results in Table 5 also suggested that 3a targets both gyrase and topo IV. Furthermore,
similar to FQs, topo IV seems to be the primary target in S. aureus whereas gyrase appears as
the primary target in K. pneumoniae.
Table 5. MICs of 3a and of Cip on a panel of effluxꢀcompetent S. aureus, K. pneumoniae
and E. coli strains engineered to be resistant to FQs
MIC (µg/ml)
Fold shift vs wt strain
Strain
3a
Cip
3a
Cip
S. aureus
RN4220 wt parent
RN4220 GyrA D83N
4
4
16
4
8
16
0.25
0.25
1
0.25
1ꢀ2
64
ꢀ
ꢀ
1
4
1
4ꢀ8
256
1
4
1
2
4
RN4220 GyrA D83N ParC D79N
RN4220 GyrB D437N
RN4220 ParC S80F
RN4220 ParC S80F GyrA S84L
K. pneumoniae
1
161486 wt parent
161486 GyrA S83I
1161486 GyrA S83I ParC S80I
161486 GyrA S83F D87V ParC S80I
16
64
128
64
<0.063
ꢀ
ꢀ
1
1
8
8
4
8
4
>16
>128
>128
1
E. coli
MG1655 wt parent
MG1655 GyrA S83L D87N ParC S80I
8
512
0.016
32
ꢀ
64
ꢀ
2000
ACS Paragon Plus Environment

Page 21 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 6. MICs of selected IPYs and Cip on one FQꢀsensitive and one highly FQꢀresistant
E. coli ꢀtolC clinical isolates
MIC (µg/ml)
W4753 ꢁtolC FQ (GyrA S83L
S
R
7
623 ꢁtolC FQ
(gyrase wt)
0.002
Fold
shift
5000
32
16
32
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
Compound
D87N ParC S80I ParE S458A)
a
Cip
10
1a
0.063
0.016
0.125
2
0.25
4
1b
2a
3a
3e
4
0.125
0.063
1
2
0.5
2
16
8
2
a
Average of two experiments.
To confirm these results, we screened a series of representative IPYs on two E. coli clinical
isolates, engineered to harbor a tolC deletion that make them effluxꢀdeficient (Table 6). The
S
R
first strain is FQꢀsensitive (FQ ) while the second one is highly resistant to FQs (FQ ), due to
an accumulation of mutations in GyrA, ParC and ParE. The ratio (or fold shift) between the
MICs of these two strains was used as a surrogate to estimate the level of crossꢀresistance
with FQs. Under these conditions, Cip demonstrated an impressive 5000ꢀfold shift. On the
other hand, compound 3a displayed a more modest 16ꢀfold shift, confirming partial crossꢀ
resistance with FQs. Similar fold shifts were seen for 3e as well as for 1a, 1b and 2a.
The observed crossꢀresistance suggested that IPYs, occupy the same binding pockets in
topoisomerases as quinolones. Unfortunately, contrary to quinazolinediones that also occupy
the FQꢀbinding pocket, the molecular interactions of IPYs with their targets apparently
involve residues that have a key role in the binding of quinolones, thereby conferring crossꢀ
resistance to FQs if mutated. Interestingly, upon screening historical compounds related to the
parent IPYs on the same two above strains, we discovered that some “nonꢀimidazo” analogs
showed much reduced fold shifts, as low as 1ꢀ2 fold, although generally at the expense of the
activities on the pumpꢀdeleted E. coli strain, suggesting lower affinity for the topoisomerase
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 59
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
targets. Compound 4 (Scheme 5) was a typical such example with a 2ꢀfold shift, a modest
S
MIC (1 µg/ml) on the wt gyrase 7623 ꢁtolC Q strain and a lack of activity on the MG1655
effluxꢀcompetent strain (Table 6). The comparison of the fold shifts and the MICs on the wt
gyraseꢀexpressing strain for 4 and 1b, two compounds that differ only by the imidazo moiety,
suggests that an important part of the affinity for the targets as well as the crossꢀresistance of
the bicyclic, tricyclic and tetracyclic IPYs with FQs could be due to this heterocyclic moiety.
As a consequence of this hypothesis, we speculate that crossꢀresistance of IPYs might be
abolished by substituting the imidazo moiety for simpler substituents (as in compound 4).
This latter conclusion echoes a recent patent that claims 2ꢀquinolones with simple 3 and 4
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2
substituents related to those seen in compound 4 . These compounds display antibacterial
22
activities on FQꢀsensitive and resistant strains .
CONCLUSIONS
Starting from a proprietary series of bacterial topoisomerase inhibitors, we have designed
by hybridization with the known quinazolinedione series a series of tricyclic
imidazopyrazinones that inhibited bacterial topoisomerases and displayed promising
preliminary activities against Gramꢀnegative bacteria, acceptable frequencies of resistance
and in vivo efficacy with concomitant tolerability. During our profiling, we discovered that
this new series, as well as the historical parent compounds, displayed partial crossꢀresistance
with fluoroquinolones and that the crossꢀresistance could be abolished by removing the
imidazo moiety, although this led to much weaker inhibition. Considering the crossꢀresistance
issue, our inability to generate improved 8ꢀamino analogues (results not shown) and the MNT
alerts across the IPY series, we halted our program at this stage due to the low likelihood of
successfully delivering a development candidate, illustrating some of the challenges faced by
23
antibacterial discovery. In a companion paper , we report the results of our studies regarding
the mechanism of action of the IPYs and Xꢀray crystallography data confirming that IPYs
ACS Paragon Plus Environment

Page 23 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
bind to the same binding pocket as FQs. Finally, our results suggest that the two core rings of
the quinolones could be either rescaffolded to resemble the tricyclic core of the tetracyclic
parent IPY series or extended toward the direction occupied by the exoꢀarylidene or the Nꢀ
phenylacetamido moieties present respectively in the tetracyclic IPYs 1 and in the bicyclic
IPYs such as 2a.
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
EXPERIMENTAL SECTION
Chemistry. The nomenclature of the compounds is based on ACDLABS software, version
11.01. All solvents and reagents obtained from commercial sources were used without further
TM
purification. The microwave oven was a Biotage Initiator
2.0 (400 W, 2450 MHz).
Reactions were monitored by thin layer chromatography (TLC) or by LCꢀMS using an
Agilent 1100 or 1260 Infinity (UV detection at 254 nm or 220 nm) relying on a low resonance
electrospray mode (ESI). Thin layer chromatography was carried out on Merck silica gel 60
F254 preꢀcoated glass backed plates and the visualization was realized under UV light (254
nm) and/or by staining with aqueous solution of H O and Ce(SO followed
3
Mo12
O
40P x H
2
4 2
)
by gentle heating with a heat gun. Purifications were performed using preꢀpacked Redisep
silica gel cartridges (20−40 ꢂm) or by flash column chromatography using silica gel 60.
HPLC purifications were performed with a Waters AutoPurification HPLCꢀMS system or a
Gilson PLC2020 using RPꢀ18 columns with water/TFA (at respectively 0.1% or 0.05%) as
the mobile phase A and MeCN (plus TFA 0.035% in the case of the Gilson HPLC) as the
mobile phase B. The product containing fractions were collected and freezeꢀdried to yield the
1
13
final products. The H and C NMR spectra were performed, in DMSOꢀd or methanolꢀd , at
6
4
13
303K, either on a Bruker Avance I operating at a proton frequency of 600 MHz and a Cꢀ
carbon frequency of 151 MHz or on a Bruker Avance III spectrometers operating at a proton
13
frequency of 500 MHz and a Cꢀcarbon frequency of 126 MHz. Chemical shifts (δ, expressed
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 59
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
in ppm) were referenced to the solvent signals
(
2.50 ppm and 39.52 ppm). Coupling constants
(J) are given in hertz. The mass spectra were recorded on a Waters UPLCꢀSQD instrument
(
mode electrospray ES+ and ES−) using a RPꢀ18 column. Purities of the final compounds are
95.0% and were measured by liquid chromatography coupled to high resolution mass
≥
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
spectrometry, using UV detection at 220 nm. High resolution mass spectrometry was
performed on an Agilent 6230 TOF LCꢀMS in combination with an Agilent 1290 Infinity II
HPLC or on an LCT Premier XE Waters TOF spectrometer in combination with a Waters
Acquity UPLC for the LC part. Following the purification step, solid compounds
(intermediate and targeted compounds) were generally not crystallized and used/characterized
as such.
Nꢀ{4ꢀ[(E)ꢀ(5ꢀmethylꢀ4ꢀoxoꢀ4,5ꢀdihydroꢀ10Hꢀimidazo[1,2ꢀa]indeno[1,2ꢀe]pyrazinꢀ10ꢀ
ylidene)methyl]phenyl}acetamide (1a). To a solution of 6 (1.74 g, 7.3 mmol) and 4ꢀ
acetamido benzaldehyde (1.63 g, 10 mmol) in methanol (50 ml) under nitrogen bubbling, was
added piperidine (1 ml, 10 mmol). The reaction mixture was refluxed for 1.5 h. The resulting
orange precipitate was then filtered and washed with methanol. The crude solid was
crystallized from DMF (50 ml) to afford the expected compound 1a as an orange solid (2 g,
1
7
7
2
1
1%); mp 200 °C. H NMR (600 MHz) δ 10.16 (s, 1H), 8.66 (d, J = 1 Hz, 1H), 8.06 (s, 1H),
.95 (d, J = 7.8 Hz, 1H), 7.73 (d, J = 8.5 Hz, 2H), 7.66 (d, J = 1 Hz, 1H), 7.54 (d, J = 8.5 Hz,
H), 7.46 (d, J = 7.8 Hz, 1H), 7.31ꢀ7.35 (m, 1H), 7.11ꢀ7.16 (m, 1H), 3.97 (s, 3H), 2.89 (s,
1
3
H), 2.10 (s, 3H). C NMR (151 MHz) δ 169.1, 153.4, 140.4, 137.2, 134.2, 133.8, 133.3,
131.9, 131.1, 130.8, 130.7, 128.9, 128.7, 127.2, 123.8, 122.1, 119.2, 116.6, 116.1, 31.6, 24.6.
+
HRMS (ESIꢀTOF) m/z: calcd for C H N O [M+H] , 383.1508; found, 383.1496.
2
3
18
4
2
(10E)ꢀ5ꢀmethylꢀ10ꢀ[(pyridinꢀ4ꢀyl)methylidene]ꢀ5,10ꢀdihydroꢀ4Hꢀimidazo[1,2ꢀ
a]indeno[1,2ꢀe]pyrazinꢀ4ꢀone (1b). To suspension of (10E)ꢀ10ꢀ[(pyridinꢀ4ꢀ
a
ACS Paragon Plus Environment

Page 25 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
5
yl)methylidene]ꢀ5,10ꢀdihydroꢀ4Hꢀimidazo[1,2ꢀa]indeno[1,2ꢀe]pyrazinꢀ4ꢀone 7 (1 g, 3.2
mmol) in DMF (30 ml) was added sodium hydride (50% dispersion) (0.154 g, 3.2 mmol). The
reaction mixture was first heated at 100°C until a clear solution was obtained and then
allowed to cool down to room temperature. Methyl iodide (0.2 ml, 3.2 mmol) in DMF (1 ml)
was then added. The resulting suspension was heated at 70°C for 1.5 h and then poured on a
water ice mixture. The precipitate was washed successively with water and hot ethanol. The
crude solid was then crystallized twice from DMF to afford the title compound 1b as a solid
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
(0.28 g, 28%); mp 302°C. H NMR (600 MHz) δ 8.70ꢀ8.76 (m, 2H), 8.66 (d, J = 1.2 Hz, 1H),
8.02 (s, 1H), 7.96 (d, J = 7.9 Hz, 1H), 7.68 (d, J = 1.2 Hz, 1H), 7.53ꢀ7.59 (m, 2H), 7.35 (dt, J
13
=
1.3, 7.6 Hz, 1H), 7.15ꢀ7.19 (m, 1H), 7.10ꢀ7.15 (m, 1H), 3.96 (s, 3H). C NMR (151 MHz)
δ 153.4, 150.5, 144.6, 137.1, 134.3, 134.2, 133.0, 132.4, 130.8, 129.6, 128.1, 127.6, 124.2,
+
1
24.2, 122.5, 116.6, 115.5, 31.6. HRMS (ESIꢀTOF) m/z: calcd for C H N O [M+H] ,
2
0
14
4
3
27.1246; found, 327.1248.
2ꢀ(4ꢀchlorophenyl)ꢀNꢀ[4ꢀ(7ꢀmethylꢀ8ꢀoxoꢀimidazo[1,2ꢀa]pyrazinꢀ5ꢀyl)phenyl]
acetamide (2a). To a solution of 2ꢀ(4ꢀchlorophenyl)acetic acid (70 mg, 0.41 mmol, 1.2
equiv.) in DCM (20 mL) was added oxalyl chloride (0.11 mL, 1.23 mmol, 3.00 equiv.) and a
drop of DMF. After stirring overnight at room temperature, the reaction mixture was
concentrated in vacuo and the resulting acid chloride was used without further purification.
To a solution of the hydrochloride salt of 12 (94 mg, 0.34 mmol, 1 equiv.) in DCM (10 mL)
and pyridine (0.3 mL) was added the freshly prepared 2ꢀ(4ꢀchlorophenyl)acetic acid chloride
dissolved in DCM (5 mL) and the reaction mixture was then stirred overnight. The reaction
mixture was concentrated in vacuo. The residue was dissolved in MeOH (3 mL), filtered and
purified by preparative HPLC chromatography (RPꢀ18 column, 15 minutes gradient 10ꢀ85%
acetonitrile, retention time 7.3 min) to yield 2a (88 mg, 0.22 mmol; 66% yield) as a colorless
1
3
solid after freeze drying. H NMR (500 MHz): δ 10.45 (s, 1 H), 7.78 (d, J = 8.7 Hz, 2 H),
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 59
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
3
7
.72 (s, 1 H), 7.66 (s, 1 H), 7.55 (d, J = 8.7 Hz, 2 H), 7.41ꢀ7.36 (m, 4 H), 7.33 (s, 1 H), 3.70
1
3
(s, 2 H), 3.51 (s, 3 H). C NMR (126 MHz): δ 169.08, 151.86, 140.13, 136.09, 134.77,
1
31.33, 131.04, 130.53, 129.18, 128.24, 124.70, 121.02, 119.36, 118.46, 115.93, 42.45, 35.15.
+
HRMS (ESIꢀTOF) m/z: calcd for C21
H
18ClN
4
O
2
[M+H] , 393.1113; found, 393.1116.
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
(
S)ꢀ7ꢀ(3ꢀaminopyrrolidinꢀ1ꢀyl)ꢀ5ꢀcyclopropylꢀ8ꢀfluoroimidazo[1,2ꢀa]quinoxalinꢀ4(5H)ꢀ
one (3a). Trifluoroacetic acid (0.5 mL) was added to a solution of 17a (110 mg, 0.257 mmol),
in DCM (2 mL). The mixture was stirred for 12 h at room temperature and then concentrated.
The residue was washed with a mixture of EtOH and diꢀisopropyl ether (9/1). The ensuing
precipitate was filtered, washed twice with diꢀisopropyl ether and then dried under reduce
pressure to afford the trifluoroacetate salt of 3a as a white powder (74 mg, 0.168 mmol; 65%
1
yield). H NMR (600 MHz) δ 8.36 (d, J = 1.1 Hz, 1H), 8.14 (br s, 3H), 8.05 (d, J = 13.8 Hz,
1
H), 7.52 (d, J = 1.1 Hz, 1H), 7.04 (d, J = 8.4 Hz, 1H), 3.97 (br s, 1H), 3.73 (ddd, J = 2.4, 6.4,
0.6 Hz, 1H), 3.61ꢀ3.69 (m, 1H), 3.56 (td, J = 2.9, 10.6 Hz, 1H), 3.40ꢀ3.48 (m, 1H), 2.91ꢀ3.02
1
(
m, 1H), 2.28ꢀ2.38 (m, 1H), 2.07 (tdd, J = 5.0, 7.8, 12.8 Hz, 1H), 1.20ꢀ1.38 (m, 2H), 0.72ꢀ
13
0.91 (m, 2H). C NMR (151 MHz)+TFA δ ppm 158.33 (q, J = 30.9 Hz, 1 C), 153.85 (s, 1 C),
1
1
1
47.50 (d, J = 238.7 Hz, 1 C), 136.11 (s, 1 C), 134.92 (d, J = 10.6 Hz, 1 C), 133.06 (s, 1 C),
28.38 (s, 1 C), 115.84 (s, 1 C), 113.54 (d, J = 10.3 Hz, 1 C), 105.11 (d, J = 27.6 Hz, 1 C),
03.76 (d, J = 5.0 Hz, 1 C), 53.77 (d, J = 6.6 Hz, 1 C), 49.97 (s, 1 C), 47.86 (d, J = 4.0 Hz, 1
C), 29.30 (s, 1 C), 26.16 (s, 1 C), 10.50 (d, J = 9.0 Hz, 2 C). HRMS (ESIꢀTOF) m/z: calcd for
+
C
17
H18FN
5
O [M+H] , 328.1574; found, 328.1578.
(S)ꢀ7ꢀ(3ꢀaminopyrrolidinꢀ1ꢀyl)ꢀ5ꢀcyclopropylꢀ8ꢀfluoroꢀ[1,2,4]triazolo[1,5ꢀa]quinoxalinꢀ
4(5H)ꢀone (3b). The trifluoroacetate salt of 3b was prepared as described for 3a, starting from
compound 17b (100 mg, 0.233 mmol), to afford a white powder (69 mg, 0.156 mmol; 67%
1
yield). H NMR (600 MHz) δ 8.52 (s, 1H), 8.15 (br s, 3H), 7.84 (d, J = 13.1 Hz, 1H), 7.10 (d,
J = 8.1 Hz, 1H), 3.98 (br s, 1H), 3.78 (ddd, J = 2.5, 6.4, 10.9 Hz, 1H), 3.67ꢀ3.73 (m, 1H), 3.58
ACS Paragon Plus Environment

Page 27 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(
td, J = 3.1, 10.9 Hz, 1H), 3.47ꢀ3.53 (m, 1H), 3.01ꢀ3.07 (m, 1H), 2.28ꢀ2.42 (m, 1H), 2.00ꢀ2.18
13
(m, 1H), 1.28ꢀ1.38 (m, 2H), 0.79ꢀ0.98 (m, 2H). C NMR (151 MHz)+TFA δ ppm 158.41 (q,
J = 30.8 Hz, 1 C), 153.42 (br s, 1 C), 153.38 (s, 1 C), 147.79 (d, J = 240.8 Hz, 1 C), 143.01 (s,
1
1
C), 136.17 (d, J = 10.6 Hz, 1 C), 128.80 (s, 1 C), 115.72 ꢀ 119.43 (m, 1 C), 114.10 (d, J =
0.6 Hz, 1 C), 104.10 (d, J = 27.8 Hz, 1 C), 103.60 (br d, J = 5.0 Hz, 1 C), 53.78 (br d, J = 6.6
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
Hz, 1 C), 49.96 (s, 1 C), 47.92 (br d, J = 3.7 Hz, 1 C), 29.30 (s, 1 C), 26.60 (s, 1 C), 10.38 (br
+
d, J = 5.8 Hz, 2 C). HRMS (ESIꢀTOF) m/z: calcd for C H FN O [M+H] , 329.1526; found,
16 17
6
329.1524.
(S)ꢀ7ꢀ(3ꢀaminopyrrolidinꢀ1ꢀyl)ꢀ5ꢀcyclopropylꢀ8ꢀfluoroimidazo[1,5ꢀa]quinoxalinꢀ4(5H)ꢀ
one 2,2,2ꢀtrifluoroacetate (3c). The trifluoroacetate salt of 3c was prepared as described for
3
a, starting from compound 17c (50 mg, 0.117 mmol), to afford a white powder (40 mg,
1
0
.091 mmol, 78%). H NMR (600 MHz) δ 8.86 (s, 1H), 8.03ꢀ8.29 (m, 4H), 7.79 (s, 1H), 7.04
(d, J = 8.5 Hz, 1H), 3.90ꢀ4.04 (m, 1H), 3.38ꢀ3.81 (m, 4H), 2.82ꢀ3.00 (m, 1H), 2.27ꢀ2.41 (m,
13
1
H), 2.00ꢀ2.14 (m, 1H), 1.12ꢀ1.44 (m, 2H), 0.67ꢀ0.97 (m, 2H). C NMR (151 MHz)+TFA δ
ppm 158.54 (q, J = 32.6 Hz, 1 C), 155.56 (s, 1 C), 147.31 (d, J = 238.4 Hz, 1 C), 135.13 (d, J
10.3 Hz, 1 C), 132.70 (s, 2 C), 131.12 (s, 1 C), 128.58 (s, 1 C), 122.28 (s, 1 C), 113.46 ꢀ
20.93 (m, 1 C), 112.31 (d, J = 10.3 Hz, 1 C), 105.00 (d, J = 27.8 Hz, 1 C), 104.12 (d, J = 5.0
Hz, 1 C), 53.76 (d, J = 6.4 Hz, 1 C), 49.91 (d, J = 1.3 Hz, 1 C), 47.88 (d, J = 4.2 Hz, 1 C),
=
1
2
9.31 (s, 1 C), 25.55 (s, 1 C), 10.37 (d, J = 7.4 Hz, 1 C). HRMS (ESIꢀTOF) m/z: calcd for
+
C
17
H18FN
5
O [M+H] , 328.1574; found, 328.1563.
(S)ꢀ7ꢀ(3ꢀaminopyrrolidinꢀ1ꢀyl)ꢀ5ꢀcyclopropylꢀ8ꢀfluoroꢀ3Hꢀimidazo[4,5ꢀc]quinolinꢀ
4(5H)ꢀone dihydrochloride (3d). Ammonium formate (42.96 mg, 0.661 mmol) and
palladium on carbon (45 mg, 0.042 mmol) were added to a solution of 17 (30 mg, 0.066
mmol) in methanol (3 mL). The resulting mixture was refluxed for 6 h and then concentrated.
The residue was suspended in methanol and then filtered through a 0.45 µm RC membrane.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 59
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
The filtrate was purified via HPLC to afford the trifluoroacetate salt of 3d as a white powder
1
(2 mg, 0.006 mmol; 1% yield). H NMR (600 MHz) δ 8.11 (s, 1H), 7.68 (d, 1H), 7.20 (d, 1H),
3
2
.76ꢀ3.81 (m, 1H), 3.71ꢀ3.76 (m, 2H), 3.54ꢀ3.63 (m, 1H), 3.38ꢀ3.45 (m, 1H), 3.02 (tt, 1H),
13
.27ꢀ2.37 (m, 1H), 1.88ꢀ1.99 (m, 1H), 1.41ꢀ1.47 (m, 2H), 0.89ꢀ0.98 (m, 2H). C NMR (151
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
MHz) δ ppm 158.06 (br s, 1 C), 148.72 (d, J = 239.7 Hz, 1 C), 140.72 (br s, 1 C), 137.37 (s, 1
C), 137.30 (s, 1 C), 136.25 (s, 1 C), 110.00 (s, 1 C), 108.08 (d, J = 24.4 Hz, 1 C), 105.77 (br
d, J = 6.4 Hz, 1 C), 102.71 (d, J = 5.0 Hz, 1 C), 56.96 (br d, J = 5.8 Hz, 1 C), 50.59 (d, J = 1.6
Hz, 1 C), 47.60 (s, 1 C), 32.75 (s, 1 C), 26.07 (s, 1 C), 10.17 (d, J = 1.9 Hz, 2 C). HRMS
+
(ESIꢀTOF) m/z: calcd for C17
H
18
N
5
OF [2M+H] , 327.1574; found, 328.1581.
5ꢀcyclopropylꢀ8ꢀfluoroꢀ7ꢀ(pyridinꢀ4ꢀyl)imidazo[1,2ꢀa]quinoxalinꢀ4(5H)ꢀone (3e). Pyridinꢀ
4ꢀylboronic acid (169.59 mg, 1,24 mmol), potassium carbonate (343.23 mg, 2.48 mmol) and
tetrakis(triphenylphosphine)palladium(0) (43.05 mg, 0.037 mmol) were added, under an
argon atmosphere, to a solution of 16a (400 mg, 1.24 mmol) in a mixture of methanol (4 ml)
and toluene (8 ml). The resulting mixture was heated under microꢀwave irradiation for 1 h at
120°C and was then concentrated. The residue was suspended in methanol and then filtered
through a 0.45 µm RC membrane. The filtrate was purified by HPLC to afford the
1
trifluoroacetate salt of 3e as a white powder (65 mg, 0.142 mmol; 12% yield). H NMR (500
MHz) δ 8.71ꢀ8.77 (m, 2H), 8.54 (d, J = 1. 2 Hz, 1H), 8.34 (d, J = 11.0 Hz, 1H), 8.04 (d, J =
7.1 Hz, 1H), 7.70 (d, J = 4.7 Hz, 2H), 7.61 (d, J = 1.2 Hz, 1H), 3.04ꢀ3.14 (m, 1H), 1.24ꢀ1.42
1
3
(
m, 2H), 0.76ꢀ0.96 (m, 2H); C NMR (126 MHz) δ ppm 154.74 (d, J = 244.7 Hz, 1 C),
152.99 (s, 1 C), 150.15 (s, 2 C), 141.74 (s, 1 C), 136.53 (s, 1 C), 133.09 (s, 1 C), 128.12 (s, 1
C), 123.47 (d, J = 3.7 Hz, 2 C), 123.31 (d, J = 2.7 Hz, 1 C), 123.21 (s, 1 C), 118.56 (s, 1 C),
116.20 (s, 1 C), 104.68 (br d, J = 28.4 Hz, 1 C), 25.96 (s, 1 C), 9.99 (s, 2 C). HRMS (ESIꢀ
+
TOF) m/z: calcd for C18
H
13FN
4
O [M+H] , 321.1152; found, 321.1155.
ACS Paragon Plus Environment

Page 29 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
5
ꢀcyclopropylꢀ8ꢀfluoroꢀ7ꢀ(pyridinꢀ4ꢀyl)ꢀ[1,2,4]triazolo[1,5ꢀa]quinoxalinꢀ4(5H)ꢀone (3f).
The trifluoroacetate salt of 3f was prepared as described for 3e, starting from compound 16b
(
200 mg, 0.619 mmol). Purification under conditions similar to 3e afforded a white powder
1
(
109 mg, 0.250 mmol; 41% yield). H NMR (600 MHz) δ 8.87 (d, J = 6.5 Hz, 2H), 8.67 (s,
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
H), 8.18 (d, J = 6.8 Hz, 1H), 8.14 (d, J = 10.1 Hz, 1H), 7.95 (d, J = 5.1 Hz, 2H), 3.07ꢀ3.26
13
(
m, 1H), 1.22ꢀ1.42 (m, 2H), 0.83ꢀ1.06 (m, 2H). C NMR (151 MHz) +TFA δ ppm 158.69 (q,
J = 36.7 Hz, 1 C), 155.46 (d, J = 247.4 Hz, 1 C), 154.30 (s, 1 C), 153.11 (s, 1 C), 148.06 (br s,
2 C), 145.13 (br s, 1 C), 144.90 (s, 1 C), 128.88 (d, J = 2.1 Hz, 1 C), 125.16 (d, J = 3.2 Hz, 2
C), 124.73 (d, J = 11.1 Hz, 1 C), 124.48 (d, J = 14.0 Hz, 1 C), 119.65 (d, J = 2.6 Hz, 1 C),
1
16.05 (q, J = 291.4 Hz, 1 C), 104.34 (d, J = 29.1 Hz, 1 C), 26.93 (s, 1 C), 10.40 (s, 2 C).
+
HRMS (ESIꢀTOF) m/z: calcd for C H FN O [M+H] , 322.1104; found, 322.1111.
1
7
12
5
5ꢀcyclopropylꢀ8ꢀfluoroꢀ7ꢀ(pyridinꢀ4ꢀyl)imidazo[1,5ꢀa]quinoxalinꢀ4(5H)ꢀone (3g). The
trifluoroacetate salt of 3g was prepared as described for 3e, starting from compound 16c (200
mg, 0.619 mmol). Purification under conditions similar to 3e afforded a white powder (110
1
mg, 0.270 mmol, 41%). H NMR (600 MHz) δ 9.10 (d, J = 0.7 Hz, 1H), 8.84ꢀ8.97 (m, 2H),
8
.46 (d, J = 11.2 Hz, 1H), 8.01ꢀ8.11 (m, 3H), 7.90 (d, J = 0.7 Hz, 1H), 2.84ꢀ3.24 (m, 1H),
13
1
.07ꢀ1.41 (m, 2H), 0.58ꢀ1.01 (m, 2H). C NMR (151 MHz) +TFA δ ppm 158.75 (q, J = 37.1
Hz, 1 C), 155.17 (s, 1 C), 155.36 (d, J = 246.4 Hz, 1 C), 146.64 (br s, 2 C), 134.27 (s, 1 C),
31.80 (s, 1 C), 128.99 (d, J = 2.4 Hz, 1 C), 125.53 (d, J = 3.7 Hz, 2 C), 123.62 (d, J = 11.4
1
Hz, 1 C), 123.21 (s, 1 C), 122.75 ꢀ 122.88 (m, 1 C), 122.67 ꢀ 122.97 (m, 1 C), 119.48 (d, J =
2.7 Hz, 1 C), 115.89 (q, J = 290.4 Hz, 1 C), 105.27 (d, J = 29.4 Hz, 1 C), 25.90 (s, 1 C), 10.40
+
(
s, 2 C). HRMS (ESIꢀTOF) m/z: calcd for C H FN O [M+H] , 321.1152; found, 321.1163.
1
8
13
4
5ꢀcyclopropylꢀ8ꢀfluoroꢀ7ꢀ(pyridinꢀ3ꢀyl)imidazo[1,2ꢀa]quinoxalinꢀ4(5H)ꢀone
(3h).
Pyridinꢀ3ꢀylboronic acid (125.92 mg, 1.02 mmol), cesium carbonate (546.18 mg, 1.68 mmol)
and tetrakis(triphenylphosphine)palladium(0) (53.81 mg, 0.047 mmol) were added, under an
ACS Paragon Plus Environment