Novel 8-nitroquinolin-2(1H)-ones as NTR-bioactivated antikinetoplastid molecules: Synthesis, electrochemical and SAR study

Article abstract of DOI:10.1016/j.ejmech.2018.06.001

To study the antiparasitic 8-nitroquinolin-2(1H)-one pharmacophore, a series of 31 derivatives was synthesized in 1–5 steps and evaluated in vitro against both Leishmania infantum and Trypanosoma brucei brucei. In parallel, the reduction potential of all molecules was measured by cyclic voltammetry. Structure-activity relationships first indicated that antileishmanial activity depends on an intramolecular hydrogen bond (described by X-ray diffraction) between the lactam function and the nitro group, which is responsible for an important shift of the redox potential (+0.3 V in comparison with 8-nitroquinoline). With the assistance of computational chemistry, a set of derivatives presenting a large range of redox potentials (from ?1.1 to ?0.45 V) was designed and provided a list of suitable molecules to be synthesized and tested. This approach highlighted that, in this series, only substrates with a redox potential above ?0.6 V display activity toward L. infantum. Nevertheless, such relation between redox potentials and in vitro antiparasitic activities was not observed in T. b. brucei. Compound 22 is a new hit compound in the series, displaying both antileishmanial and antitrypanosomal activity along with a low cytotoxicity on the human HepG2 cell line. Compound 22 is selectively bioactivated by the type 1 nitroreductases (NTR1) of L. donovani and T. brucei brucei. Moreover, despite being mutagenic in the Ames test, as most of nitroaromatic derivatives, compound 22 was not genotoxic in the comet assay. Preliminary in vitro pharmacokinetic parameters were finally determined and pointed out a good in vitro microsomal stability (half-life > 40 min) and a 92% binding to human albumin.

Full text of DOI:10.1016/j.ejmech.2018.06.001

Accepted Manuscript
Novel 8-nitroquinolin-2(1H)-ones as NTR-bioactivated antikinetoplastid molecules:
Synthesis, electrochemical and SAR study
Julien Pedron, Clotilde Boudot, Sébastien Hutter, Sandra Bourgeade-Delmas,
Jean-Luc Stigliani, Alix Sournia-Saquet, Alain Moreau, Elisa Boutet-Robinet, Lucie
Paloque, Emmanuelle Mothes, Michèle Laget, Laure Vendier, Geneviève Pratviel,
Susan Wyllie, Alan Fairlamb, Nadine Azas, Bertrand Courtioux, Alexis Valentin, Pierre
Verhaeghe
PII:
S0223-5234(18)30491-4
10.1016/j.ejmech.2018.06.001
EJMECH 10471
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 28 March 2018
Revised Date: 31 May 2018
Accepted Date: 1 June 2018
Please cite this article as: J. Pedron, C. Boudot, Sé. Hutter, S. Bourgeade-Delmas, J.-L. Stigliani,
A. Sournia-Saquet, A. Moreau, E. Boutet-Robinet, L. Paloque, E. Mothes, Michè. Laget, L. Vendier,
Geneviè. Pratviel, S. Wyllie, A. Fairlamb, N. Azas, B. Courtioux, A. Valentin, P. Verhaeghe, Novel 8-
nitroquinolin-2(1H)-ones as NTR-bioactivated antikinetoplastid molecules: Synthesis, electrochemical
and SAR study, European Journal of Medicinal Chemistry (2018), doi: 10.1016/j.ejmech.2018.06.001.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Novel 8-Nitroquinolin-2(1H)-ones as NTR-bioactivated Antikinetoplastid
Molecules: Synthesis, Electrochemical and SAR Study
1
2
3
4
1
Julien Pedron , Clotilde Boudot , Sébastien Hutter , Sandra Bourgeade-Delmas , Jean-Luc Stigliani ,
1
1
5
1
1
Alix Sournia-Saquet , Alain Moreau , Elisa Boutet-Robinet , Lucie Paloque , Emmanuelle Mothes ,
6
1
1
7
7
3
Michèle Laget , Laure Vendier , Geneviève Pratviel , Susan Wyllie , Alan Fairlamb , Nadine Azas ,
2
4
1
Bertrand Courtioux , Alexis Valentin and Pierre Verhaeghe *
1
LCC-CNRS Université de Toulouse, CNRS, UPS, Toulouse, France.
2
Université de Limoges, UMR INSERM 1094, Neuroépidémiologie Tropicale, Faculté de Pharmacie,
2
rue du Dr Marcland, 87025 Limoges, France
3
IHU Méditerranée Infection, équipe VITROME « Vecteurs, Infections Tropicales et
Méditerranéennes », 19-21 boulevard Jean Moulin, 13385 Marseille cedex 05, France.
4
UMR 152 PharmaDev, Université de Toulouse, IRD, UPS, Toulouse, France.
5
Toxalim (Research Centre in Food Toxicology), Université de Toulouse, INRA, ENVT, INP-Purpan,
UPS, Toulouse, France
6
UMR MD1, U1261, AMU, INSERM, SSA, IRBA, MCT, Marseille, France.
7
University of Dundee, School of Life Sciences, Division of Biological Chemistry and Drug
Discovery, Dow Street, Dundee DD1 5EH, Scotland, United Kingdom
*
Corresponding author: Pierre Verhaeghe (pierre.verhaeghe@univ-tlse3.fr)
Tel: +33 561 333 236
https://orcid.org/0000-0003-0238-2447

ACCEPTED MANUSCRIPT
ABSTRACT
To study the antiparasitic 8-nitroquinolin-2(1H)-one pharmacophore, a series of 31 derivatives was
synthesized in 1-5 steps and evaluated in vitro against both Leishmania infantum and Trypanosoma
brucei brucei. In parallel, the reduction potential of all molecules was measured by cyclic
voltammetry. Structure-activity relationships first indicated that antileishmanial activity depends on an
intramolecular hydrogen bond (described by X-ray diffraction) between the lactam function and the
nitro group, which is responsible for an important shift of the redox potential (+0.3 V in comparison
with 8-nitroquinoline). With the assistance of computational chemistry, a set of derivatives presenting
a large range of redox potentials (from -1.1 to -0.45 V) was designed and provided a list of suitable
molecules to be synthesized and tested. This approach highlighted that, in this series, only substrates
with a redox potential above -0.6 V display activity toward L. infantum. Nevertheless, such relation
between redox potentials and in vitro antiparasitic activities was not observed in T. b. brucei.
Compound 22 is a new hit compound in the series, displaying both antileishmanial and
antitrypanosomal activity along with a low cytotoxicity on the human HepG2 cell line. Compound 22
is selectively bioactivated by the type 1 nitroreductases (NTR1) of L. donovani and T. brucei brucei.
Moreover, despite being mutagenic in the Ames test, as most of nitroaromatic derivatives, compound
2
2 was not genotoxic in the comet assay. Preliminary in vitro pharmacokinetic parameters were finally
determined and pointed out a good in vitro microsomal stability (half-life > 40 min) and a 92%
binding to human albumin.
KEYWORDS
Anti-kinetoplastids, Leishmania, Trypanosoma, 8-Nitroquinolin-2(1H)-one, Nitroreductases,
Electrochemistry.
HIGHLIGHTS
8
-nitroquinolin-2(1H)-one derivative 22 is bioactivated by L. donovani NTR1
They display suitable redox potentials thanks to an intramolecular H-bond

ACCEPTED MANUSCRIPT
Introducing a bromine atom at position 3 leads to a new antikinetoplastid hit
INTRODUCTION
Kinetoplastids are protozoan parasites responsible for deadly mammalian infections; Leishmania spp
and Trypanosoma spp being the two main genera encountered in human pathology. These parasites are
transmitted to their mammalian hosts by the bite of specific insect vectors. Among the Leishmania
genus, L. donovani and L. infantum are the two most important species responsible for visceral
leishmaniasis (VL), the most severe and often lethal clinical form of the disease. Briefly, VL occurs in
humans after the bite of a sandfly (Phlebotominae) which injects the metacyclic flagellated
promastigote stage of the parasite into the skin. Parasites are then phagocytized by macrophages and
other types of mononuclear phagocytic cells where they transform into the amastigote stage and
multiply, bypassing the immune system of the host. Infected cells disseminate in many tissues and
1
organs such as lymph nodes, liver, spleen and bone marrow, leading to death. In the Trypanosoma
genus, in addition to T. cruzi species, causing Chagas disease in South and Central America, the T.
brucei species (T. brucei gambiense or T. brucei rhodesiense) are responsible for Human African
Trypanosomiasis (HAT), also called “sleeping sickness”. During HAT, the flagellated trypomastigote
stage of the parasites penetrates into the blood stream via the bite of a “tsetse fly” and disseminates
into the whole organism. In a first step of the disease, headaches, anemia, joint pain and various organ
damage occur. Then, in a second step, the parasites invade the central nervous system, causing various
neurological changes such as sleeping disorders (responsible for the name of the disease), abnormal
2
tone and mobility, ataxia, psychiatric disorders, seizures, coma and finally, death.
It is important to note that all these kinetoplastid infections are lethal if untreated and that very few
efficient, safe and affordable drugs are available on the market for the low-income infected patients
living in developing countries, which makes these parasitic infections belong to the group of
3
“
Neglected Tropical Diseases”. Regarding recent epidemiologic data, the WHO indicates that more
than 1 billion people are at risk of contracting VL and that 300.000 new cases occur every year,
4
responsible for about 20.000 deaths. Concerning HAT, 61 million people would be at risk of
5
contracting the disease, no clear data being available about HAT mortality. Approved drugs against

ACCEPTED MANUSCRIPT
VL are amphotericin B, miltefosine, antimony V derivatives and pentamidine. The liposomal form of
amphotericin B, although being quite efficient, is also very expensive and requires parenteral
6
administration, which makes it unsuitable for treating most of infected patients. Pentamidine and
pentavalent antimony derivatives are toxic agents which use is limited by parasitic resistances and
7
poor efficacy. Miltefosine (Figure 1) remains the only orally available drug, but it is a teratogenic
6
molecule, preventing its use in women of child-bearing age. The context of anti-HAT drugs is quite
8
similar with only one orally available drug. In addition to pentamidine, suramin is used in the
treatment of the first stage of the disease, with nifurtimox in combination with eflornithine
(
NECT,Figure 1) for treatment of the second (cerebral) stage of HAT. NECT has largely replaced the
2
use of melarsoprol, a highly toxic arsenic-containing drug.
Figure 1: Structures of the antikinetoplastid drugs miltefosine and eflornithine, drug-candidate
fexinidazole and antileishmanial hit 8-nitroquinolin-2(1H)-one
Unfortunately, there are very few new chemical entities under clinical trial in 2018 against both VL
8
and HAT. In this worrying context, there is a critical need to identify new active molecules and, as
observed in the field of tuberculosis, various nitroheterocycles such as delamanid and fexinidazole are
9, 10
re-emerging as key anti-infective molecules.
Fexinidazole (Figure 1) is a 5-nitroimidazole
derivative that was first evaluated against visceral leishmaniasis in a phase II clinical trial. Although
not sufficiently efficient for use as a monotherapy for VL, fexinidazole has now demonstrated efficacy
11-12
against early and late stages of HAT in a pivotal Phase III clinical trial.
Fexinidazole is part of

ACCEPTED MANUSCRIPT
these nitroaromatic molecules which are selectively activated by parasitic nitroreductases (NTRs) to
1
3,14
generate electrophilic cytotoxic metabolites.
Unfortunately, no parasitic NTR X-ray structure is
available and most classical rational medicinal chemistry approaches cannot be used for the design of
new synthetic substrates of these key enzymes. In this context, our group initiated a research around a
1
5
class of 8-nitroquinoline derivatives with antiparasitic potential and identified an antileishmanial
1
6
pharmacophore in 2012, corresponding to the 8-nitroquinolin-2(1H)-one scaffold (Figure 1). Some
more recent modulations of the position 4 of this pharmacophore failed to achieve more active
1
7, 18
derivatives.
Here, we present a comprehensive study of the structure-activity relationships of this
pharmacophore, especially focusing on the modulation of the redox potentials. Thirty one
pharmacophore derivatives, substituted at positions 3, 4, 5, 6 and 7, were synthesized and their in vitro
cytotoxicity and activities against L. infantum (axenic amastigotes), and T. brucei brucei
(
trypomastigotes) were first determined. The original pharmacophore possessed poor antitrypanosomal
activity, whereas a new derivative (compound 22) was identified, active against all tested
kinetoplastids (L. donovani promastigotes, L donovani intramacrophage amastigotes, L. infantum
axenic amastigotes and T. brucei brucei trypomastigotes). The antileishmanial mechanism of action of
this new hit-compound will be discussed along with its mutagenicity and genotoxicity, a key aspect
concerning nitroaromatic compounds. Finally, two preliminary in vitro pharmacokinetic properties
were measured and appear favorable for further development of this anti-kinetoplastid pharmacophore.
RESULTS AND DISCUSSION
Chemistry
Thirty one derivatives of the initial 8-nitroquinolin-2-(1H)-one pharmacophore, including 14 original
ones, were prepared. As presented in Scheme 1, compounds 1-5 were obtained by classical nitration,
cyclo-condensation or N-acetylation reactions. Nitration at position 8 of 2-chloro-4-methylquinoline
afforded 6 whose chloroimine moiety reacted with perchloric acid to generate the corresponding
1
9
lactam 7, according to a simple and efficient previously reported procedure. Bromination of 7 with

ACCEPTED MANUSCRIPT
N-bromosuccinimide led to the bromomethyl derivative 8 which was reacted with sodium hydroxide
to generate hydroxymethyl derivative 9. From bromomethyl 8, by reacting with an amine, it was also
possible to access to 10 and its deprotected counterpart 11. From 2,4-dibromoquinoline, by a
sequential nitration/lactamisation procedure, 12 and 13 were obtained in good yields.
Scheme 1: Synthetic routes for the preparation of compounds 1-13. (a) H SO , HNO , rt, 1 h; (b) glyoxylic acid,
2
4
3
EtOH, 80 °C, 6 h; (c) CH COOH, Ac O, 120 °C, 3 h; (d) H SO , HNO , rt, 4 h; (e) CH COOH, Ac O, 120 °C,
3
2
2
4
3
3
2
2
4 h; (f) H SO , HNO , rt, 3 h; (g) CH CN, HClO , 100 °C, 24 h; (h) NBS, AIBN, CCl , 80 °C, 24 h; (i) NaOH,
2 4 3 3 4 4
H O/THF, 50 °C, 24 h; (j) NH (CH CH O) CH CH NHBoc, THF, 50 °C, 24 h; (k) HCl (5N in isopropanol), rt,
2
2
2
2
2
2
2
4
h; (l) H SO , HNO , rt, 2 h; (m) CH CN, HClO , MW, 100 °C, 1 h.
2 4 3 3 4
As presented in Scheme 2, the 2-chloroquinoline was the main substrate used in this work. It was first
reacted with perchloric acid to generate lactam 14 which was either nitrated at position 6, to afford
compound 15 or N-methylated with methyl iodide, to afford 16 in good yields. The nitration of 2-
chloroquinoline mainly leaded to the 2 position isomers 17 and 18 which were transformed into the
corresponding lactams 19 and 20. Compound 20 was either O-methylated with methyl iodide, forming
2
1, or selectively halogenated at position 3 by refluxing in HBr or HCl in the presence of sodium
bromate or sodium chlorate, to generate derivatives 22 and 23. Note that this specific halogenation
procedure (in situ Br or Cl formation and reaction in refluxing aqueous medium), reported by
2
2
2
0
O’Brien and co-workers, is the only one allowing selective halogenation at position 3 among many

ACCEPTED MANUSCRIPT
others that were tested. The 3-brominated derivative 22 reacted with methyl iodide to form O-
methylated derivative 24 or was reduced into the amino-derivative 25 by using SnCl2.
Scheme 2: Preparation of compounds 14-25 from 2-chloroquinoline. (a) CH CN, HClO , 100 °C, 72 h; (b)
3
4
H SO , HNO , rt, 1 h; (c) DMF, K CO , CH I, 80 °C, 48 h; (d) H SO , HNO , rt, 2 h; (e) CH CN, HClO , 100
2
4
3
2
3
3
2
4
3
3
4
°
C, 72 h ; (f) CH CN, HClO , 100 °C, 72 h; (g) Ar., DMF, NaH, CH I, rt, 24 h; (h) NaBrO , HBr, 100 °C, 5 h;
3
4
3
3
(
i) NaClO , HCl, 100 °C, 45 min; (j) Ar., DMF, NaH, CH I, rt, 24 h; (k) EtOH, SnCl , 80°C, 3 h.
3
3
2
Finally, as presented in Scheme 3, we also used another synthetic pathway allowing the preparation of
-nitroquinolinone derivatives bearing substituents on the benzene moiety. This strategy was reported
8
21
by Zaragoza and co-workers. First, commercial 3,3’-diethoxyethylpropionate was saponified into the
corresponding carboxylic acid which was then reacted with SOCl to form the acyl chloride. This acyl
2
chloride was then reacted with various ortho-nitroanilines, to give the corresponding N-acylated
products which were not isolated. In a final step, cyclization was operated in 98% H SO , giving the
2
4
expected products 26-31. The reaction yields are moderate when using 2-nitroanilines substituted at
position 3 or 5 by electron donating groups but turn to be very low when using 2-nitroanilines
substituted at position 4 by the same groups.
Scheme 3: Preparation of compounds 26-31 from nitroanilines subtrates. (a) CH Cl , Pyridine, rt, 24 h; (b)
2
2
H SO , rt, 3-4 h.
2
4

ACCEPTED MANUSCRIPT
Compound evaluation
In vitro activity against L. infantum and T. brucei brucei: importance of the redox potential
All synthesized molecules were screened in vitro toward both L. infantum axenic amastigotes and T.
brucei brucei trypomastigotes and compared with commercial reference drugs (amphotericin B and
miltefosine against L. infantum, suramin and eflornithine against T. brucei) along with drug-candidate
fexinidazole (Table 1). To assess antiparasitic selectivity, the cytotoxicity of all compounds was also
measured on the HepG2 human cell line, using doxorubicin as control. In parallel the LogP of all
compounds were calculated and the reduction potentials were measured by cyclic voltammetry in
DMSO. The redox potentials measured correspond to a one electron reduction/oxidation regarding the
redox couple nitro group / anion radical. Only one new hit compound (22) was evaluated in greater
depth, as the best compound in the series regarding both activity and selectivity.
Globally, apart from molecule 24, all synthesized compounds were soluble in aqueous medium and
could be evaluated in vitro. Out of the 31 tested compounds, none appeared cytotoxic on the HepG2
cell line, nine were active toward L. infantum axenic amastigotes (7 µM ≤ IC ≤ 25 µM) and thirteen
5
0
were active against T. brucei brucei trypomastigotes (2 µM ≤ IC50 ≤ 25 µM), molecule 24 being
excluded from the analysis because of its lack of solubility. As a first approach, we wanted to see if
there was a relationship between activity and lipophilicity. Computed LogP, ranging from 1.15 to 2.5
did not show any correlation, active molecules presenting either low (compound 2) or high (compound
2
2) LogP values.
We then explored the structure-activity relationships (SARs) of the pharmacophore by focusing on the
influence of the redox potentials toward the activity against L. infantum. Indeed, on the basis that this
nitroaromatic pharmacophore could be the substrate of parasitic nitroreductases, it first appeared
important to understand which chemical criteria would make this pharmacophore suitable for being
efficiently reduced into cytotoxic metabolites. The basic redox potential value of 8-nitroquinoline 1 (-
0
.84 V) and nitrobenzene (-0.85 V) are quite low and do not seem suitable for the catalytic capabilities

ACCEPTED MANUSCRIPT
of the parasitic nitroreductases (Figure 2). On the contrary, when the lactam function is combined with
the presence of a nitro group at position 8 (initial hit), a significant increase in the value of the redox
potential, up to -0.54 V, is observed and the molecule turns active. When keeping the lactam function
while moving the nitro group at position 6 or 5 (compounds 15 and 19), redox potentials become
lower and the IC values of the molecules increased, resulting in a loss of activity against L. infantum.
5
0
The same effect is noted with inactive O-methylated analog (compound 21), which redox potential
drastically fell to -0.93 V (Figure 2). Finally, adding an amide function in the ortho position of the
nitrogroup (compounds 3 and 5) did not provide an optimal value of redox potential (-0.68 V and -
0
.86 V, respectively) resulting in poor antileishmanial activity of the compounds. Thus the 8-
nitroquinolin-2-(1H)-one scaffold definitely appears to be the antileishmanial pharmacophore.
Besides, as presented in the X-ray diffraction structures (Figure 3), this pharmacophore presents a
noticeable intramolecular hydrogen bond between the lactam function and the nitro group at position
8
, responsible for a +0.3 V increase in redox potentials, allowing antileishmanial activity.
Nevertheless, molecules 26 and 30, which present the same redox potential, do not have the same
antileishmanial activity (26 being moderately active and 30 being inactive). By examining their LogP
values (respectively 2.05 and 1.38) it could be hypothesized that antileishmanial activity depends on a
compromise between redox potential values and lipophilicity.

ACCEPTED MANUSCRIPT
Table 1: Reduction potentials, calculated LogP, anti-kinetoplastid activity and cytotoxicity of the
synthesized nitroaromatic compounds 1-31.
L. infantum axenic T. brucei brucei
HepG2
E°
V/NHE)
b
Molecule
Structure
a
LogP
1.54
1.91
amastigote
IC (µM)
trypomastigote Cytotoxicity
(
IC (µM)
CC (µM)
50
50
50
Initial hit
- 0.54
15.5 +/-0.5
23.4 +/-5.7
164 +/-28
Nitro-
benzene (NB)
d
d
d
-
0.85
> 100
> 50
> 100
d
d
1
2
- 0.84
- 0.46
2.07
1.2
> 100
17.9 +/-1.8
7.3 +/-0.8
> 100
11.0 +/-2.1
125 +/-19
d
d
d
3
4
5
- 0.68
- 1.1
1.15
1.89
1.31
> 100
> 50
> 100
d
d
d
> 100
> 50
> 100
c
d
d
- 0.86
> 100
10.8 +/-1.4
20.9 +/-8.9
> 100
7
9
- 0.53
2.05
-
10.2 +/-1.0
23.4 +/-4.8
110 +/-9
d
-
-
38.3 +/-6.1
2.5 +/-1.8
> 100
d
d
1
1
0
1
-
-
> 100
> 100
d
d
-
> 100
7.6 +/-1.8
5.5 +/-1.3
> 100
1
1
3
4
- 0.51
1.86
-
20.3 +/-5.0
60 +/-2
b
d
d
d
- 1.86
> 100
> 50
> 100

ACCEPTED MANUSCRIPT
d
d
d
1
1
5
6
- 0.82
-
1.54
-
> 100
> 100
2.3 +/-0.6
> 100
> 100
d
d
d
> 50
d
d
d
1
9
- 0.67
1.54
> 100
> 50
> 100
> 100
2
2
2
2
1
2
3
4
- 0.93
- 0.45
- 0.47
- 0.86
2.51
2.36
2.2
-
88.8 +/-13.0
7.1 +/-1.5
28.5 +/-8.1
1.9 +/-0.4
1.2 +/-0.2
0.4 +/-0.1
92 +/-13
16.0 +/-0.9
85 +/-4
e
e
> 6.2
> 6.2
d
d
d
2
2
5
6
-
-
> 50
> 50
> 100
d
- 0.58
2.05
25.2 +/-3.7
23.0 +/-5.0
29.1 +/-4.7
2.1 +/-0.8
> 100
d
2
2
7
8
- 0.56
- 0.71
2.05
2.05
> 100
d
d
d
> 100
> 50
> 100
d
d
d
2
9
- 0.65
1.38
> 100
> 50
> 100
d
d
d
3
3
0
1
- 0.58
- 0.75
1.38
1.38
> 100
> 50
> 100
d
d
d
> 100
> 50
> 100
f
Doxorubicin
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
0.2 +/-0.02
5.5 +/-0.25
85 +/-8.8
g
Amphotericin B
0.06 +/-0.001
0.8 +/-0.2
3.4 +/-0.8
g
Miltefosine
g,h
c
Fexinidazole
0.4 +/-0.2
0.02 +/-0.009
15.8 +/-2.1
> 200c
h
Suramin
-
-
> 100
> 100
h
c
Eflornithine
a
b
c
d
e
f
6
Cyclic voltammetry conditions: DMSO/TBAPF , SCE/GC, 1 electron reversible reduction, values are given in V versus NHE
Weighted logP were computed with Marvin (ChemAxon)
®
Irreversible reduction
The IC50 or CC50 value was not reached at the highest tested concentration
The product could not be tested at higher concentrations due to a poor solubility in aqueous medium
Doxorubicin was used as a cytotoxic reference drug
g
Amphotericin B, Miltefosine and Fexinidazole were used as antileishmanial reference drugs

ACCEPTED MANUSCRIPT
h
Fexinidazole, Suramin and Eflornithine were used as anti-Trypanosoma brucei reference drugs
Figure 2: The 8-nitroquinolin-2(1H)-one pharmacophore presented a characteristic high redox
potential value.
Figure 3: X-ray structures of initial hit and new hit 22 presenting an intramolecular hydrogen bond
between the lactam function and the nitro group (Initial hit: N-H distance = 0.89 Å, H-O distance =

ACCEPTED MANUSCRIPT
2
1
.01 Å, N-H-O angle = 129°; Hit 22 : N-H distance = 0.84 Å, H-O distance = 2.01 Å, N-H-O angle =
31°).
Contrary to what was observed in Leishmania, the antitrypanosomal activity did not depend on the
redox potential values in the studied series, active molecules presenting either low (compound 15) or
high (compound 22) redox potential values. It could be hypothesized that the enzymatic capabilities of
the T. brucei NTR may be compatible with a broader spectrum of redox potentials than the ones of L.
donovani NTR.
Optimization of redox potentials: computational and in vitro studies.
From this point, we investigated the antileishmanial pharmacomodulation of 8-nitroquinolin-2(1H)-
one by introducing various substituents (or heteroatoms) at all positions of the scaffold and compared
the effect of electron-donating and electron withdrawing groups, to gain access to new optimized
derivatives. Our goal was to synthesize a series of derivatives presenting a large spectrum of redox
potentials, not only to gain a better understanding of the electrochemical process, but also to define as
precisely as possible the lowest redox potential value compatible with activity in the series. For that
purpose instead of synthesizing molecules randomly, we used ab initio calculations to estimate the
redox potential values of many hypothetical derivatives. Previous studies about the computation of
reduction potentials of nitroaromatic derivatives showed that the results are very sensitive to the basis
set applied. Diffuse s and p-type functions, combined with polarization functions were shown to be
2
2,23
important for electron affinity calculations of nitrobenzenes . Because accurate results were
obtained with the Pople's 6-311++G basis set on several molecules for which computed potentials
were compared to experimental ones, the 6-311++G(2d,2p) basis set was selected as a good
compromise between accuracy and computational cost for reduction potential calculations. The hybrid
M06-2x density functional was chosen for its accuracy to compute thermochemical parameters. The
standard redox potentials were computed with the Faraday’s law (see equation (1) in the experimental
section) from the free energy of the one-electron reduction. Figure 4 shows the correlation between the
experimental and theoretical redox potentials. It can be seen that the slope of the correlation is close to
1
.00 and the correlation coefficient (r) is 0.97. Therefore, this model was very efficient for predicting

ACCEPTED MANUSCRIPT
the standard redox potentials and was used to select the most appropriate compounds to be synthesized
and evaluated.
Figure 4: Correlation between theoretical and experimental standard redox potential of nitroquinolines
The studied series was completed with 11 derivatives of the pharmacophore presenting redox
potentials ranging from -0.75 V to -0.45 V. Figure 5 shows that introducing a heteroatom at position 4
of the scaffold (quinoxaline analog 2) or a halogen atom at position 3 (compounds 22 and 23)
increased the redox potential values by about +0.1 V, the corresponding molecules being active
against L. infantum. When modifying the pyridinone moiety by introducing a methyl group or a
bromine atom at position 4 (compounds 7 and 13, respectively), the redox potential almost remained
unchanged and the molecules were active. By substituting the benzene moiety with a methyl group at
position 5 or 6 (compounds 26 and 27), the redox potentials decreased slightly (down to -0.58 V) and
the molecules, were less efficient (IC ≈ 25 µM). Compounds bearing a methoxy group at positions 5,
50
6
or 7 (compounds 29-31) or a methyl group at position 7 (compound 28) showed lower redox
potentials, varying from -0.58 V to -0.75 V, and were totally inactive (IC > 100 µM). Thus, in this
5
0

ACCEPTED MANUSCRIPT
series, it can be noted that the optimal redox potential value to target in order to obtain antileishmanial
activity is -0.5 V and that molecules whose redox potential is positioned below -0.6 V lost all activity.
Figure 5: Modulation of the redox potential of the 8-nitroquinolin-2[1H]-one pharmacophore.
Mode of action studies, genotoxicity evaluation and preliminary pharmacokinetic properties of
2
2
Among all synthesized compounds, derivative 22 was highlighted as a new hit (Table1). It is not only
active against L. infantum (IC = 7.1 µM), but also against T. brucei brucei (IC = 1.9 µM), which
5
0
50
was not the case of the initial hit, displaying modest activity (IC50 = 23.4 µM) toward the latter
parasite. For comparison with the anti-leishmanial reference drugs, hit compound 22 appears much
less active than amphotericin B and about 10 times less active than miltefosine (with equivalent
cytotoxicity). Regarding anti-HAT drugs, hit compound 22 appears much less active than suramin, but
more active than eflornithine. It is also important to compare hit molecule 22 with drug candidate
fexinidazole, as they both belong to the nitroaromatic family: 22 is about 2 times more cytotoxic and,
depending on the parasite, between 2 to 5 times less active than this drug candidate, which remains
interesting for a hit-compound. It was then necessary to evaluate the in vitro activity of 22 against the

ACCEPTED MANUSCRIPT
amastigote stage of Leishmania in an intramacrophage model, closer to the in vivo situation. As
presented in Table 2, the test was done on L. donovani, the other species responsible for VL. Hit
molecule 22, displays a low cytotoxicity on the THP1 macrophage cell line (CC = 72 µM) and is
5
0
about 3 times less active (IC50 = 18 µM) on this intramacrophage amastigote in vitro model than
miltefosine (IC = 5.4 µM). Fexinidazole is not active on the intramacrophage amastigote stage (IC
50
5
0
2
4
>
50 µM) as it first needs to be metabolized into a sulfone derivative. Finally, as expected,
compound 14 (non nitrated) and 25 (8-amino derivative), negative controls deriving from hit 22, did
not display any antiparasitic activity.
In order to assess whether this compound series was a substrate of parasitic NTRs, the hit compound
2
2 was assayed against a wild type strain of L. donovani promastigotes and two strains overexpressing
25
26
NTR1 and NTR2 , respectively (Table 2). Compound 22 was ten times more effective against the
strain overexpressing NTR1 (IC = 0.47 µM) vs WT and NTR2 strains, indicating that it is selectively
5
0
bioactivated by NTR1 of L. donovani. Interestingly, this is the same type 1 nitroreductase that
2
5
bioactivates fexinidazole and its metabolites. Following the same approach in T. brucei brucei, it was
also demonstrated that type 1 NTR, the unique NTR enzyme expressed in this parasite, was
responsible for the bioactivation of 22 (Table 2).
Nitroaromatic drugs are often suspected to be mutagenic or genotoxic, which has considerably limited
2
7
the development of such derivatives in the past decades. The most famous in vitro test used for
evaluating mutagenicity is the Ames test, using Salmonella typhimurium strains. Nevertheless,
2
8
considering that Salmonella bacteria possess nitroreductases, the Ames test is usually positive when
evaluating nitroaromatics, with no real predicting character for humans, considering that there are no
NTRs in mammalian cells and that most of nitroaromatic derivatives only exert a genotoxic character
2
7
after being bioactivated into reduced metabolites. For that reason, it is nowadays accepted that the
comet assay or the micronucleus assay are better in vitro tools for evaluating the potential genotoxicity
2
7
of nitroaromatics. Indeed, metronidazole, one of the most famous anti-infective nitroheterocycles on
2
9
the market, is mutagenic in the Ames test, but not genotoxic in the comet assay and fexinidazole is
2
4
not genotoxic in the micronucleus assay despite being mutagenic in the Ames test. Regarding

ACCEPTED MANUSCRIPT
compound 22, the Ames test (in metabolizing conditions) and the comet assay were done in parallel
and revealed that, although being mutagenic in the Ames test (Table 3) at 0.25 or 2.5 mM, compound
2
2 was not genotoxic in the comet assay after 2 or 72 h of exposure, at 1, 10, 20 or 40 µM (Figure 6),
concentrations chosen from the HepG2 CC10% (40 µM). This lack of genotoxicity is a key point when
thinking about further development of this pharmacophore. Finally, preliminary in vitro
pharmacokinetic evaluations were done with compound 22 (Table 2). It presented a very good
microsomal stability (T1/2 > 40 min) and had a strong, but not extreme binding to human albumin
(
92%), in relation with the relatively high value of its calculated logP (= 2.36).
Table 2: Complementary in vitro biological evaluations done on hit compound 22
In vitro studies
a
Cytotoxicity: CC THP1 (µM)
72 +/-6
18 +/-2
5
0
b
IC L. donovani intramacrophagic amastigotes (µM)
5
0
IC L. donovani promastigotes
50
5
.9 +/-0.12
wild type strain (µM)
IC L. donovani promastigotes
50
0
.47 +/-0.02
NTR1 over-expressing strain (µM)
IC L. donovani promastigotes
50
4
.6 +/-0.12
7.7 +/-1.0
NTR2 over-expressing strain (µM)
IC T. b. brucei trypomastigotes
5
0
1
wild type strain (µM)
IC T. b. brucei trypomastigotes
5
0
3
.9 +/-0.1
NTR1 over-expressing strain (µM)
Microsomal stability: T1/2 (min)
Binding % to human albumin
> 40
92
a
Reference compounds: Amphotericin B CC50 = 3.6 +/-0.7 µM; Miltefosine CC50 > 40 µM; Fexinidazole CC50 > 62.5 µM
Reference compounds: Amphotericin B IC50 = 0.4 µM; Miltefosine IC50 = 5.4 µM; Fexinidazole IC50 > 50 µM
b

ACCEPTED MANUSCRIPT
Table 3: Ames test results for hit compound 22
0.25 mM
2.5 mM
TA97a + S9 mix
TA98 + S9 mix
TA100 + S9 mix
Positive
Positive
Positive
Negative
Positive
Positive
Positive
Positive
TA102 + S9 mix
benzo[a]pyrene was used as a positive control
Figure 6: results of the comet assay for hit compound 22.
Positive control = MMS = 1 mM methylmethanesulfonate

ACCEPTED MANUSCRIPT
CONCLUSION
Starting from the previously identified 8-nitroquinolin-2(1H)-one antileishmanial pharmacophore, an
electrochemistry-guided SAR study revealed that an intramolecular hydrogen bond between the lactam
ring and the nitro group is responsible for a +0.3 V increase in the redox potential value (compared
with 8-nitroquinoline) resulting in improved antileishmanial activity. A computational model allowed
the prediction of the redox potential values of any derivative belonging to this series, and we
synthesized a set of derivatives presenting a large range of redox potential values. Then, it was
highlighted that, to be active, molecules belonging to this series must display a redox potential value
of about -0.5 V and that molecules with a redox potential below -0.6 V were inactive. By introducing
electron-withdrawing groups on the pyridone moiety of the scaffold, a new original hit compound was
found: 22. Presenting a low cytotoxicity on the human HepG2 cell line (CC = 92 µM), it displays
5
0
activity toward L. infantum axenic amastigotes (IC = 7.1 µM), L. donovani promastigotes (IC = 5.9
5
0
50
µM) and L. donovani intramacrophagic amastigotes (IC = 18 µM), but also toward T. brucei brucei
5
0
(
IC = 1.9 µM). Moreover, we demonstrated that 22 was not genotoxic in the comet assay and that it
50
was selectively bioactivated by the mitochondrial NTR1 of L. donovani and T. brucei brucei. This
whole study, completed with two encouraging preliminary in vitro pharmacokinetic parameters (good
microsomal stability and 92% binding to human albumin) now opens the way to the rational
conception, guided with E° values, of new efficient, selective and safe NTR1 bioactivated 8-
nitroquinolin-2(1H)-ones against VL and HAT.
ASSOCIATED CONTENT
Supporting Information

ACCEPTED MANUSCRIPT
Additional analysis of the hit compound 22 as well as biological methods, Ames test and Comet assay
protocols, electrochemistry and computational studies methods, X-Ray data, microsomal stability and
plasma protein binding assay results.
AUTHOR INFORMATION
Corresponding author
*
E-mail: pierre.verhaeghe@lcc-toulouse.fr. Phone: +33 561333236.
ORCID
Pierre Verhaeghe: 0000-0003-0238-2447
Notes
The authors declare no competing financial interest
ACKNOWLEDGMENTS
A. Fairlamb and S. Wyllie are supported by funding from the Wellcome Trust (WT105021). J. Pedron
thanks the Université Paul Sabatier de Toulouse and the Région Occitanie / Pyrénées-Méditerranée for
PhD funding. C. Piveteau from Institut Pasteur de Lille, C. Bijani from Laboratoire de Chimie de
Coordination de Toulouse and the mass spectrometry team of the Institut de Chimie de Toulouse are
also acknowledged for their support, respectively for pharmacokinetics, NMR and HRMS
experiments.
ABBREVIATIONS
NTR, Nitroreductase; SAR, Structure-Activity Relationship; rt, room temperature; NBS, N-
bromosuccinimide; AIBN, Azobisisobutyronitrile; THF, Tetrahydrofuran; MW, Microwaves; DMF,
N,N-Dimethylformamide, Ar., Argon; NHE, Normal Hydrogen Electrode; IC , half maximal
5
0
inhibitory concentration; CC , half maximal cytotoxic concentration; TBAPF Tetrabutylammonium
50
6,

ACCEPTED MANUSCRIPT
hexafluorophosphate; SCE, Saturated Calomel Electrode; GC, Glassy Carbon; E°, standard redox
potential; E°exp, standard experimental redox potential; E°calc, standard calculated redox potential;
HRMS, High-resolution mass spectrometry; NMR, Nuclear Magnetic Resonance; VL, Visceral
leishmaniasis; HAT, Human African Trypanosomiasis.
EXPERIMENTAL SECTION
Chemistry
All reagents and solvents were obtained from commercial sources (Fluorochem®, Sigma-Aldrich® or
Alfa Aesar®) and used as received. The progress of the reactions was monitored by pre-coated thin
®
layer chromatography (TLC) sheets ALUGRAM SIL G/UV from Macherey-Nagel and were
2
54
1
13
visualized by ultraviolet light at 254 nm. The NMR H and C spectra were recorded on a Bruker
UltraShield 300 MHz, a Bruker IconNMR 400 MHz, or a Bruker Avance NEO 600 MHz instrument, at
the Laboratoire de Chimie de Coordination and data are presented as follows: chemical shifts in parts
per million (δ) using tetramethylsilane (TMS) as reference, coupling constant in Hertz (Hz),
multiplicity by abbreviations : (s) singlet, (d) doublet, (t) triplet, (q) quartet, (dd) doublets of doublets,
(
m) multiplet and (br s) broad singlet. Melting points are uncorrected and were measured on a Stuart
Melting Point SMP3 instrument. HRMS measurements were done on a GCT Premier Spectrometer (DCI,
CH , HRMS) or Xevo G2 QTOF (Waters, ESI+, HRMS) instrument at the Université Paul Sabatier,
4
®
Toulouse. Microwave reactions were realized in a CEM Discover microwave reactor. All products were
1
purified by chromatography columns and then recrystallized. Purity of the products was determined by H
NMR.
3
0
31
General procedure for the preparation of 8-nitroquinoline 1 , 7-amino-8-nitroquinoline 4 , 2-chloro-
3
2
17
16
4
-methyl-8-nitroquinoline 6 , 2,4-dibromo-8-nitroquinoline 12 , 6-nitroquinolin-2(1H)-one 15 , 2-
16
16
chloro-5-nitroquinoline 17 and 2-chloro-8-nitroquinoline 18 :
H SO (98%) was added onto 1 equiv. of the quinoline derivatives, cooled with an ice bath. 5 equiv. of
2
4
6
5% HNO were then added dropwise at 0°C and the reaction mixture was stirred at rt for 1-4 h. The
3

ACCEPTED MANUSCRIPT
reaction mixture was successively poured into ice, neutralized with NaOH and extracted three times
with dichloromethane. The organic layer was washed with water, dried over anhydrous Na SO and
2
4
evaporated in vacuo. The crude residue was purified by chromatography on silica gel using adapted
eluent and recrystallized if necessary to give compounds 1, 4, 6, 12, 15, 17 and 18.
8
-nitroquinoline 1 (C H N O ) was purified by chromatography on silica gel using dichloromethane as
9 6 2 2
an eluent, separated from its 5-nitro isomer and isolated to yield a pale yellow solid (35%, 2.7 mmol,
1
4
70 mg), mp 90 °C (Lit: 90 °C), H NMR (400 MHz, CDCl ) δ: 7.53-7.66 (m, 2 H), 8.04 (d, J = 8.0
3
1
3
Hz, 2 H), 8.27 (dd, J = 1.6 and 8.4 Hz, 1 H, H7), 9.70 (dd, J = 1.7 and 4.2 Hz, 1 H, H2). C NMR
(
(
100 MHz, CDCl ) δ: 122.8 (CH), 123.8 (CH), 125.3 (CH), 129.0 (C), 132.0 (CH), 136.1 (CH), 139.5
3
C), 148.2 (C), 152.6 (CH).
7
-amino-8-nitroquinoline 4 (C H N O2) was purified by chromatography on silica gel using
9 7 3
dichloromethane/ethyl acetate (80/20) as an eluent and isolated to yield an orange solid (46%, 0.16
1
mmol, 30 mg), mp 193 °C (Lit: 194 °C), H NMR (400 MHz, CDCl ) δ: 5.61 (br s, 2 H, NH ) ; 7.00 (d,
3
2
J=9.0 Hz, 1 H, H6), 7.28 (dd, 1 H, J=4.4 and 8.1 Hz, H3), 7.69 (d, J= 9.0 Hz, 1H, H5), 8.00 (dd, 1 H,
13
J=1.8 and 8.1 Hz, H4), 8.92 (dd, 1 H, J=1.8 and 4.4 Hz, H2). C NMR (100 MHz, Acetone d6): 118.90
CH), 119.8 (CH), 120.9 (C), 131.8 (CH), 135.7 (CH), 139.7 (C), 142.8 (C), 143.1 (C), 151.2 (CH).
-chloro-4-methyl-8-nitroquinoline 6 (C H ClN O ) was purified by chromatography on silica gel
(
2
10
7
2
2
using cyclohexane/acetone (70/30) as an eluent, separated from its 6-nitro isomer, isolated and
recrystallized in isopropanol to yield a white solid in (53%, 14.8 mmol, 3.3 g), mp 143 °C (Lit: 142-
1
1
7
43 °C), H NMR (400 MHz, CDCl ) δ: 2.74 (d, J=1.0 Hz, 3 H, CH ), 7.35 (d, J=1.0 Hz, 1 H, H3),
3
3
.62-7.66 (m, 1 H, H6), 8.00 (dd, J= 7.6 and 1.3 Hz, 1 H, H5), 8.18 (dd, J= 8.4 and 1.3 Hz, 1 H,
13
H7). C NMR (150 MHz, CDCl ) δ: 18.9 (CH ), 124.2 (CH), 124.5 (CH), 125.4 (CH), 127.8 (CH),
3
3
1
27.9 (C), 138.9 (C), 147.8 (C), 147.9 (C), 153.2 (C).
2
,4-dibromo-8-nitroquinoline 12 (C H Br N O ) was purified by chromatography on silica gel using
9
4
2
2
2
cyclohexane/acetone (9/1) as an eluent and isolated to yield a white solid (56%, 6 mmol, 2 g), mp 129
1
°
C (Lit: 129 °C), H NMR (400 MHz, CDCl ) δ: 7.72-7.76 (m, 1 H, H6), 8.00 (s, 1 H, H3), 8.08 (dd,
3

ACCEPTED MANUSCRIPT
1
3
J= 7.6 et 1.4 Hz, 1 H, H5), 8.39 (dd, J= 8.5 et 1.4 Hz 1 H, H7). C NMR (100 MHz, DMSO) δ: 125.3
CH), 126.9 (CH), 127.5 (C), 130.9 (CH), 131.0 (CH), 135.0 (C), 139.7 (C), 143.6 (C), 147.5 (C).
-nitroquinolin-2(1H)-one 15 (C H N O ) was recrystallized in isopropanol to yield a yellow solid
(
6
9
6
2
3
1
(
91%, 6.3 mmol, 1.2 g), mp 279 °C (Lit: 279 °C), H NMR (300 MHz, DMSO d6) δ: 6.68 (d, J= 9.8
Hz, 1 H, H3), 7.43 (d, J= 9.0 Hz, 1 H, H8), 8.12 (d, J= 9.8 Hz, 1 H, H4), 8.33 (dd, J= 2.6 and 9.0 Hz, 1
1
3
H, H7), 8.70 (d, J= 2.6 Hz, 1 H, H5), 12,3 (br s, 1 H, NH). C NMR (75 MHz, DMSO d6) δ: 116.3
CH), 118.7 (C), 124.0 (CH), 124.5 (CH), 125.3 (CH), 140.3 (CH), 141.7 (C), 143.5 (C), 162.1 (C).
-chloro-5-nitroquinoline 17 (C H ClN O ) was purified by chromatography on silica gel using
(
2
9
5
2
2
cyclohexane/acetone (8/2) as an eluent, isolated and recrystallized in isopropanol to yield a yellow
1
solid (14% , 4.3 mmol, 0.9 g), mp 134 °C (Lit: 133-134 °C), H NMR (300 MHz, CDCl ) δ: 7.63 (d,
3
J=9,1 Hz, 1 H, H4), 7.82-7.87 (m, 1 H, H7), 8.34 (dd, J= 8,6 and 0,7 Hz, 1 H, H8), 8.40 (dd, J= 7.6
1
3
and 1.2 Hz, 1 H, H6), 8.99 (d, J=9.2 Hz, 1 H, H3). C NMR (100 MHz, CDCl ) δ: 120.1 (C), 125.0
3
(
CH), 125.6 (CH), 129.0 (CH), 135.1 (CH), 135.7 (CH), 145.6 (C), 148.2 (C), 152.6 (C).
-chloro-8-nitroquinoline 18 (C H ClN O ) was purified by chromatography on silica gel using
2
9
5
2
2
cyclohexane/acetone (8/2) as an eluent, isolated and recrystallized in isopropanol to yield a yellow
1
solid (50%, 15.3 mmol, 3.2 g), mp 152 °C (Lit: 152 °C), H NMR (400 MHz, CDCl ) δ: 7.56 (d, J=8.7
3
Hz, 1 H, H4), 7.65-7.69 (m, 1 H, H6), 8.07 (dd, J= 8.3 and 1.4 Hz, 1 H, H5), 8.11 (dd, J= 7.6 and 1.4
1
3
Hz, 1 H, H7), 8.23 (d, J=8.7 Hz, 1 H, H3). C NMR (150 MHz, CDCl ) δ: 124.6 (CH), 124.9 (CH),
3
1
25.8 (CH), 127.6 (C), 131.8 (CH), 138.6 (CH), 139.0 (C), 147.1 (C), 153.6 (C).
Preparation of 8-nitroquinoxalin-2(1H)-one 2
Ethanol (25 mL) was added onto 1 g of 3-nitro-1,2-phenylenediamine (6.53 mmol, 1 equiv.). 0.6 g of
glyoxylic acid monohydrate (6.53 mmol, 1 equiv.) was then added at rt and the reaction mixture was
stirred at 80°C for 6 h. The reaction mixture was poured into water and extracted three times with
dichloromethane. The organic layer was washed with water, dried over anhydrous MgSO and
4

ACCEPTED MANUSCRIPT
evaporated in vacuo. The crude residue was purified by chromatography on silica gel using
dichloromethane/diethyl ether (80/20) as an eluent. 8-nitroquinoxalin-2(1H)-one 2 was isolated and
recrystallized in acetonitrile to yield an orange solid (46%, 2.98 mmol, 570 mg).
1
Compound 2 (C H N O ): mp 162 °C, H NMR (400 MHz, CDCl ) δ: 7.45-7.50 (m, 1 H, H6), 8.25
8
5
3
3
3
(
dd, J=1.5 and 8.2 Hz, 1 H, H5), 8.39 (d, J=2.2 Hz, 1 H, H3), 8.55 (dd, J=1.5 and 8.2 Hz, 1 H, H7),
1
3
1
1
1
1.37 (br s, 1 H, NH). C NMR (100 MHz, CDCl ) δ: 122.6 (CH), 127.8 (C), 128.0 (CH), 133.1 (C),
3
+
33.8 (C), 137.4 (CH), 152.6 (CH), 153.6 (C). HRMS (DCI CH ) calcd for C H N O [M+H]
4
8
6
3
3
92.0409, found 192.0412.
3
3
Preparation of N-(2-Nitrophenyl)-acetamide 3
Acetic acid (25 mL) was added onto 525 mg of 2-nitroaniline (3.8 mmol, 1 equiv.). The reaction
mixture was then stirred at 120°C and 720 µL of acetic anhydride (7.6 mmol, 2 equiv.) were added
dropwise. After 3 h under reflux, the reaction mixture was poured into ice, neutralized with Na CO
2
3
and extracted three times with dichloromethane. The organic layer was washed with water, dried over
anhydrous Na SO and evaporated in vacuo. N-(2-Nitrophenyl)-acetamide 3 was isolated and
2
4
recrystallized in acetonitrile to yield a yellow solid (87%, 3.3 mmol, 604 mg).
1
Compound 3 (C H N O ): mp 93 °C (Lit: 92-94 °C), H NMR (400 MHz, CDCl ) δ: 2.29 (s, 3 H,
8
8
2
3
3
CH ), 7.16-7.20 (m, 1 H, H4), 7.63-7.67 (m, 1 H, H5), 8.21 (dd, J= 8.5 and 1.3 Hz, 1 H, H6), 8,77 (dd,
3
1
3
J= 8.6 and 1.3 Hz, 1 H, H3), 10.33 (br s, 1H, NH). C NMR (100 MHz, CDCl ): 25.60 (CH ), 122.2
3
3
(
CH), 123.2 (CH), 125.7 (CH), 134.8 (C), 135.9 (CH), 136.3 (C), 169.0 (C).
Preparation of 7-acetylamino-8-nitroquinoline 5
Glacial acetic acid (20 mL) was added onto 60 mg of 7-amino-8-nitroquinoline (0.32 mmol, 1 equiv.).
The reaction mixture was then stirred at 120°C and 180 µL of acetic anhydride (1.6 mmol, 6 equiv.)
were added dropwise. After 24 h under reflux, the reaction mixture was poured into ice, neutralized
with K CO and extracted three times with dichloromethane. The organic layer was washed with
2
3

ACCEPTED MANUSCRIPT
water, dried over anhydrous MgSO and evaporated in vacuo. The crude residue was purified by
4
chromatography on silica gel using dichloromethane/ethyl acetate (90/10) as an eluent. 7-acetylamino-
8
-nitroquinoline 5 was isolated and recrystallized in isopropanol to yield a yellow solid (53%, 0.17
mmol, 39 mg).
Compound 5 (C H N O ): mp 189 °C, H NMR (300 MHz, CDCl ) δ: 2.28 (s, 3 H, CH ), 7.49 (dd,
1
1
1
9
3
2
3
3
J=4.3 et 8.3 Hz, 1 H, H3), 7.96 (d, 1H, J= 9.2 Hz, H 6), 8.19 (dd, J=1.7 et 8.3 Hz, 1 H, H4), 8.35 (br s,
1
3
1
H, NH), 8.54 (d, J=9.2 Hz, 1 H, H5), 8.99 (dd, J=1.7 et 4.3 Hz, 1 H, H2). C NMR (150 MHz,
CDCl ) δ: 25.0 (CH ), 121.7 (CH), 121.9 (CH), 125.2 (C), 127.8 (C), 131.4 (CH), 131.8 (C), 135.8
3
3
+
(
CH), 140.3 (C), 152.6 (CH), 168.7 (C). HRMS (DCI CH ) calcd for C H N O [M+H] 232.0722,
4 11 10 3 2
found 232.0719.
3
4
General procedure for the preparation of 4-methyl-8-nitroquinolin-2(1H)-one 7 , quinolin-2(1H)-one
35
16
16
1
4 , 5-nitroquinolin-2(1H)-one 19 and 8-nitroquinolin-2(1H)-one 20 .
Acetonitrile (25 mL) and 70 % perchloric acid solution (25 mL) were added onto 1 equiv. of the
quinoline derivatives. The reaction mixture was stirred at 100°C overnight. The reaction mixture was
then poured into ice, neutralized with KOH and extracted twice with dichloromethane. The organic
layer was washed with water, dried over anhydrous MgSO and evaporated in vacuo. The crude
4
residues were purified by chromatography on silica gel using adapted eluent and recrystallized if
necessary to give compounds 7, 14, 19, 20.
4
-methyl-8-nitroquinolin-2(1H)-one 7 (C H N O ) was isolated and recrystallized in acetonitrile to
10 8 2 3
1
yield a yellow solid (92%, 11 mmol, 2.2 g), mp 203 °C (Lit: 196 °C), H NMR (400 MHz, DMSO) δ:
3
.33 (s, 3 H, CH ), 6.63 (s, 1 H, H3), 7.39-7.44 (m, 1 H, H6), 8.19 (dd, J= 8.0 and 1.3 Hz, 1H, H5),
3
1
3
8
.42 (dd, J= 8.4 and 1.6 Hz, 1 H, H7), 10.98 (br s, 1 H, NH). C NMR (DMSO, 100 MHz) δ: 19.6
(
CH ), 121.6 (CH), 122.0 (CH), 122.6 (C), 127.9 (CH), 133.1 (CH), 133.6 (C) 134.6 (C), 148.9 (C),
3
1
61.0 (C).
quinolin-2(1H)-one 14 (C H NO) was purified by chromatography on silica gel using
9
7
dichloromethane/ethyl acetate (8/2) as an eluent and isolated to yield a white solid (74%, 22.7 mmol,

ACCEPTED MANUSCRIPT
1
3
7
9
.3 g), mp 198 °C (Lit: 193 °C), H NMR (400 MHz, CDCl ) δ: 6.72 (d, J= 9.5 Hz, 1 H, H3), 7.20-
3
.25 (m, 1 H, H6), 7.40-7,42 (m, 1 H, H5), 7.50-7.54 (m, 1 H, H7), 7.56-7.58 (m, 1 H, H8), 7.82 (d, J=
1
3
.5 Hz,1 H, H4), 11.97 (br s, 1 H, NH). C NMR (100 MHz, CDCl ) δ: 116.3 (CH), 119.9 (C), 121.3
3
(
CH), 122.7 (CH), 127.7 (CH), 130.7 (CH), 138.5 (C), 141.1 (CH), 164.7 (C).
-nitroquinolin-2(1H)-one 19 (C H N O ) was isolated and recrystallized in acetonitrile to yield a
5
9
6
2
3
1
yellow solid (92%, 2.2 mmol, 418 mg), mp 302 °C (Lit: 302 °C), H NMR (400 MHz, DMSO d6) δ:
.76 (d, J=10.0 Hz, 1 H, H3), 7.65-7.73 (m, 2 H, H7 H8), 7.88 (dd, J= 7.7 and 1.3 Hz, 1 H, H6), 8.25
6
1
3
(
(
d, J= 10.0 Hz, 1 H, H4), 12.3 (br s, 1 H, NH). C NMR (100 MHz, DMSO d6) δ: 111.9 (C), 118.9
CH), 121.5 (CH), 125.8 (CH), 130.6 (CH), 134.7 (CH), 140.7 (C), 146.7 (C), 161.3 (C).
8
-nitroquinolin-2(1H)-one 20 (C H N O ) was isolated and recrystallized in isopropanol to yield a
9 6 2 3
1
yellow solid (96%, 10.6 mmol, 2.02 g), mp 163 °C (Lit: 163 °C), H NMR (400 MHz, CDCl ) δ: 6.79
3
(
(
dd, J= 9.7 and 2.0 Hz, 1H, H3), 7.32-7.36 (m, 1 H, H6), 7.82 (d, J= 9.7 and 1.4 Hz, 1 H, H4), 7.91
1
3
dd, J= 7.6 Hz, 1 H, H5), 8.53 (dd, J= 8.4 and 1.4 Hz, 1 H, H7), 11.3 (br s, 1H, NH). C NMR (100
MHz, DMSO d6) δ: 121.4 (CH), 122.1 (C), 123.8 (CH), 127.9 (CH), 133.1 (C), 133.9 (C), 135.6
CH), 139.9 (CH), 161.6 (C).
(
Preparation of 4-bromomethyl-8-nitroquinolin-2(1H)-one 8
Carbon tetrachloride (15 mL) was added onto 400 mg of 4-methyl-8-nitroquinolin-2(1H)-one (1.9
mmol, 1 equiv.) and 1.7 g of N-bromosuccinimide (9.7 mmol, 5 equiv.). The reaction mixture was
heated at 80°C. Thent 31 mg of AIBN (0.24 mmol, 0.1 equiv.) were added, and stirred overnight. The
reaction mixture was poured into water, extracted twice with dichloromethane and once with ethyl
acetate. The combined organic layers were washed with water, dried over anhydrous MgSO and
4
evaporated in vacuo. The crude residue was purified by chromatography on silica gel using ethyl
acetate as an eluent. 4-bromomethyl-8-nitroquinolin-2(1H)-one 8 was isolated to yield a yellow solid
(
21%, 0.42 mmol, 120 mg).
1
Compound 8 (C H BrN O ): mp 161 °C, H NMR (400 MHz, CDCl ) δ: 4.60 (s, 2 H, CH ), 6.87 (d,
1
0
7
2
3
3
2
J=2.0 Hz, 1 H, H3), 7.40-7.44 (m, 1 H, H6), 8.18 (dd, J= 1.3 and 8.8 Hz, 1 H, H5), 8.57 (dd, J= 1.3

ACCEPTED MANUSCRIPT
1
3
and 8.3 Hz, 1 H, H7), 11.44 (br s, 1 H, NH). C NMR (100 MHz, CDCl ,) δ: 27.9 (CH ), 120.2 (C),
3
2
1
21.3 (CH), 123.7 (CH), 128.4 (CH), 132.4 (CH), 133.6 (C), 134.4 (C), 146.1 (C), 160.7 (C). HRMS
+
(
DCI CH ) calcd for C H BrN O [M+H] 282.9718, found 282.9723.
4
10
8
2
3
Preparation of 4-hydroxymethyl-8-nitroquinolin-2(1H)-one 9
Fifty millilitres of a mixture H O/THF (1/1) were added onto 150 mg of 4-bromomethyl-8-
2
nitroquinolin-2(1H)-one (0.53 mmol, 1 equiv). 106 mg of NaOH (2.6 mmol, 5 equiv.) were added and
the reaction mixture was stirred at 50°C for 24 h. The reaction mixture was then poured into water,
extracted twice with dichloromethane and twice with ethyl acetate. The combined organic layers were
washed with water, dried over anhydrous MgSO and evaporated in vacuo. The crude residue was
4
purified by chromatography on silica gel using ethyl acetate as an eluent. 4-hydroxymethyl-8-
nitroquinolin-2(1H)-one 9 was isolated to yield a yellow solid (43%, 0.23 mmol, 50 mg).
1
Compound 9 (C H N O ): mp 211 °C, H NMR (300 MHz, DMSO) δ: 4.80 (d, J=5.5 Hz, 2 H, CH ),
1
0
8
2
4
2
5
8
.67 (t, J=5.5 Hz, 1 H, OH), 6.73 (s, 1 H, H3), 7.37-7.42 (m, 1 H, H6), 8,13 (d, J= 7.9 Hz, 1 H, H5),
1
3
.42 (d, J= 8.3 Hz, 1 H, H7), 11,01 (br s, 1H, NH). C NMR (150 MHz, DMSO) δ: 39.5 (CH ), 122.0
2
(
CH), 122.1 (C), 123.0 (CH), 128.2 (CH), 133.3 (CH), 133.8 (CH), 136.5 (C), 141.3 (C), 161.6 (C).
+
HRMS (DCI CH ) calcd for C H N O [M+H] 221.0562, found 221.0558.
4
10
9
2
4
Preparation of 4-{[2-(N-boc-2-aminoethoxy)ethoxy]ethyl}aminomethyl-8-nitroquinolin-2(1H)-one 10
THF (50 mL) was added onto 500 mg of 4-bromomethyl-8-nitroquinolin-2(1H)-one (1.8 mmol, 1
equiv.) and 1.3 g of N-Boc-2,2′-(ethylenedioxy)-diethylamine (5.3 mmol, 3 equiv.). Triethylamine
(
375 µL, 2.7 mmol, 1.5 equiv.) were then added and the reaction mixture was stirred at 50°C
overnight. The reaction mixture was evaporated in vacuo. The crude product was then resolubilized in
ethyl acetate, washed three times with water, dried over anhydrous MgSO and evaporated in vacuo.
4
The crude residue was purified by chromatography on silica gel using diethyl ether/methanol/ 20%
aqueous
ammonia
solution
(8.9/1/0.1)
as
an
eluent.
4-{[2-(N-boc-2-

ACCEPTED MANUSCRIPT
aminoethoxy)ethoxy]ethyl}aminomethyl-8-nitroquinolin-2(1H)-one 10 was isolated to yield an orange
oil (84%, 1.5 mmol, 680 mg).
1
Compound 10 (C H N O ): H NMR (600 MHz, DMSO) δ: 1.36 (s, 9H, tBu), 2.5 (s, 1 H, NH amine),
2
1
30
4
7
2.75-2.77 (m, 2 H, CH 3’), 3.03-3.06 (m, 2 H, CH 10’), 3.36-3.39 (m, 2 H, CH 4’), 3.51-3.54 (m, 6 H,
2 2 2
3
xCH 6’ 7’ 9’), 4.02 (s, 2 H, CH 1’), 6.73 (br s, 1 H, NH carbamate), 6.76 (s, 1 H, H3), 7.38-7.42 (m, 1
2
2
H, H6), 8.31 (dd, J= 8.1 and 1.3 Hz, 1H, H5), 8.42 (dd, J= 8.1 and 1.3 Hz, 1 H, H7), 11.03 (br s, 1 H,
NH).
1
3
C NMR (150 MHz, DMSO) δ: 28.7 (3 CH tBu), 40.1 (CH ), 48.7 (CH ), 49.9 (CH ), 69.6 (CH ),
3
,
2
2
2
2
7
0.0 (CH ), 70.1 (CH ), 70.5 (CH ), 78.0 (C), 120.5 (CH), 121.4 (C), 121.6 (CH), 127.8 (CH), 132.9
2 2 2
(
[
CH), 133.6 (C), 134.5 (C), 150.6 (C), 156.0 (C), 161.1 (C). HRMS (ESI) calcd for C H N O
21 31 4 7
+
M+H] 450.2114, found 450.2118.
Preparation of 4-{[2-(2-aminoethoxy)ethoxy]ethyl}aminomethyl-8-nitroquinolin-2(1H)-one 11
Hydrochloric acid (15 mL of 5 N HCL in isopropanol) was added onto 600 mg of 4-{[2-(N-boc-2-
aminoethoxy)ethoxy]ethyl}aminomethyl-8-nitroquinolin-2(1H)-one (1.33 mmol, 1 equiv.) and the
reaction mixture was then stirred at rt for 4 h. The reaction mixture was successively poured into
water, neutralized with K CO and extracted several times with dichloromethane. The organic layer
2
3
was
aminoethoxy)ethoxy]ethyl}aminomethyl-8-nitroquinolin-2(1H)-one 11 was isolated to yield an oil in
43%, 0.57 mmol, 200 mg).
Compound 11 (C H N O ): H NMR (400 MHz, DMSO) δ: 2.5 (s, 1 H, NH amine), 2.63-2.66 (m, 2
dried
over
anhydrous
Na SO
and
evaporated
in
vacuo.
4-{[2-(2-
2
4
(
1
1
6
22
4
5
H, CH 10’), 2.74-2.77 (m, 2 H, CH 3’), 3.35-3.38 (m, 2 H, CH 9’), 3.49-3.55 (m, 6 H, 3xCH 4’ 6’
2
2
2
2
7
1
’), 4.01 (d, J= 1.9 Hz, 2 H, CH 1’), 6.76 (br s, 1 H, H3), 7.38-7.42 (m, 1 H, H6), 8.30 (dd, J= 8.1 and
2
1
3
.3 Hz, 1 H, H5), 8.40 (dd, J= 8.1 and 1.3 Hz, 1 H, H7), 11.0 (br s, 1 H, NH lactame). C NMR (150
MHz, DMSO) δ: 41.7 (CH ), 48.7 (CH ), 50.0 (CH ), 70.0 (CH ), 70.1 (CH ), 70.6 (CH ), 73.2 (CH ),
2
2
2
2
2
2
2
1
20.4 (CH), 121.4 (C), 121.5 (CH), 127.7 (CH), 132.8 (CH), 133.7 (C), 134.6 (C), 150.7 (C), 161.2
+
(
C). HRMS (ESI) calcd for C H N O [M+H] 351.1668, found 351.1674.
16
23
4
5