1-(1H-indol-3-yl)ethanamine derivatives as potent Staphylococcus aureus NorA efflux pump inhibitors

Article abstract of DOI:10.1002/cmdc.201400042

The synthesis of 37 1-(1H-indol-3-yl)ethanamine derivatives, including 12 new compounds, was achieved through a series of simple and efficient chemical modifications. These indole derivatives displayed modest or no intrinsic anti-staphylococcal activity. By contrast, several of the compounds restored, in a concentration-dependent manner, the antibacterial activity of ciprofloxacin against Staphylococcus aureus strains that were resistant to fluoroquinolones due to overexpression of the NorA efflux pump. Structure-activity relationships studies revealed that the indolic aldonitrones halogenated at position 5 of the indole core were the most efficient inhibitors of the S. aureus NorA efflux pump. Among the compounds, (Z)-N-benzylidene-2-(tert-butoxycarbonylamino)-1-(5- iodo-1H-indol-3-yl)ethanamine oxide led to a fourfold decrease of the ciprofloxacin minimum inhibitory concentration against the SA-1199B strain when used at a concentration of 0.5 mg L^-1. To the best of our knowledge, this activity is the highest reported to date for an indolic NorA inhibitor. In addition, a new antibacterial compound, tert-butyl (2-(3-hydroxyureido)-2-(1H- indol-3-yl)ethyl)carbamate, which is not toxic for human cells, was also found. A pump turn off: 1-(1H-Indol-3-yl)ethanamine derivatives such as the one shown here, were synthesized through simple chemical modifications and were shown to be efficient NorA efflux pump inhibitors. They are able to restore ciprofloxacin activity against fluoroquinolone-resistant Staphylococcus aureus strains.

Full text of DOI:10.1002/cmdc.201400042

CHEMMEDCHEM
FULL PAPERS
DOI: 10.1002/cmdc.201400042
1-(1H-Indol-3-yl)ethanamine Derivatives as Potent
Staphylococcus aureus NorA Efflux Pump Inhibitors
Arnaud Hequet,[a] Olga N. Burchak,[b] Matthieu Jeanty,[b, e] Xavier Guinchard,[b, f]
Emmanuelle Le Pihive,[c] Laure Maigre,[c] Pascale Bouhours,[a] Dominique Schneider,[c]
Max Maurin,[c, d] Jean-Marc Paris,[a] Jean-Noꢀl Denis,*[b] and Claude Jolivalt*[a]
The synthesis of 37 1-(1H-indol-3-yl)ethanamine derivatives, in-
cluding 12 new compounds, was achieved through a series of
simple and efficient chemical modifications. These indole deriv-
atives displayed modest or no intrinsic anti-staphylococcal ac-
tivity. By contrast, several of the compounds restored, in a con-
centration-dependent manner, the antibacterial activity of ci-
profloxacin against Staphylococcus aureus strains that were re-
sistant to fluoroquinolones due to overexpression of the NorA
efflux pump. Structure–activity relationships studies revealed
that the indolic aldonitrones halogenated at position 5 of the
indole core were the most efficient inhibitors of the S. aureus
NorA efflux pump. Among the compounds, (Z)-N-benzylidene-
2-(tert-butoxycarbonylamino)-1-(5-iodo-1H-indol-3-yl)ethan-
amine oxide led to a fourfold decrease of the ciprofloxacin
minimum inhibitory concentration against the SA-1199B strain
when used at a concentration of 0.5 mgLÀ1. To the best of our
knowledge, this activity is the highest reported to date for an
indolic NorA inhibitor. In addition, a new antibacterial com-
pound, tert-butyl (2-(3-hydroxyureido)-2-(1H-indol-3-yl)ethyl)-
carbamate, which is not toxic for human cells, was also found.
Introduction
The emergence and spread of pathogenic bacteria with multi-
drug resistance (MDR) to established classes of antibiotics is an
ever-growing concern worldwide.[1] MDR bacteria are not only
responsible for infections in hospital settings but also in the
community, owing to their large spread in humans, animals,
food processing, and various environments. This corresponds
to serious public health concerns and significant economic
losses, especially because of the prolonged hospitalization of
patients and increases in treatment costs.[2] Among resistant
bacteria, methicillin-resistant Staphylococcus aureus (MRSA) is
a pathogen of particular concern because it is frequently in-
volved in nosocomial infections.[3] For example, MRSA strains
accounted for 44% of the healthcare-associated infections in
2008 in the European Union, with high associated mortality
rates.[2b] MRSA is one of the examples of the consequences of
improper use of antibiotics.[3c,4] Indeed, the link between the
large-scale use of fluoroquinolones, an important class of
broad-spectrum antibiotics, and the emergence of MRSA and
other MDR bacteria has been often reported in the litera-
ture.[4,5]
Acquired resistances to fluoroquinolones may occur through
four chromosomally and/or plasmid-mediated mechanisms:
1) alterations of type II topoisomerases (DNA gyrase and topo-
isomerase IV), the targets of fluoroquinolones, 2) extrusion of
the antibiotic by overexpression of efflux pumps, and, less fre-
quently, 3) DNA gyrase protection, and 4) decreased permeabil-
ity of the bacterial membrane to these drugs.[6]
[a] Dr. A. Hequet, P. Bouhours, Dr. J.-M. Paris, Prof. Dr. C. Jolivalt
Laboratoire Charles Friedel (LCF), CNRS UMR 7223, Chimie ParisTech
11 rue Pierre et Marie Curie, 75005 Paris (France)
E-mail: claude.jolivalt@upmc.fr
[b] Dr. O. N. Burchak, Dr. M. Jeanty, Dr. X. Guinchard, Dr. J.-N. Denis
Dꢀpartement de Chimie Molꢀculaire (SeRCO)
ICMG FR-2607, CNRS UMR 5250
Universitꢀ Joseph Fourier, BP-53, 38041 Grenoble cedex 9 (France)
E-mail: jean-noel.denis@ujf-grenoble.fr
Overexpression of efflux pumps is frequently associated with
MDR. Among the five known transporter families, classified ac-
cording to their bioenergetic and structural characteristics,[7]
two of them are involved in fluoroquinolone resistance in
S. aureus: the major facilitator superfamily (MFS) and the multi-
drug and toxic compound extrusion (MATE) family.[6,7b] Several
representatives from the MFS family (such as the NorA, NorB,
NorC, MdeA, SdrM, and QacA/B proteins) and one from the
MATE family (MepA) have been described as fluoroquinolone
exporters.[8] Most of these fluoroquinolone transporters confer
an MDR phenotype to S. aureus, which restricts therapeutic op-
tions and favors dissemination of this pathogen.
[c] Dr. E. Le Pihive, Dr. L. Maigre, Prof. D. Schneider, Prof. M. Maurin
Laboratoire Adaptation et Pathogꢀnie des Micro-organismes
CNRS UMR 5163, Universitꢀ Joseph Fourier Grenoble 1
Institut Jean Roget, Campus Santꢀ
Domaine de la Merci, BP-170, 38042 Grenoble cedex 9 (France)
[d] Prof. M. Maurin
CHU de Grenoble, Universitꢀ Joseph Fourier
BP-217, 38043 Grenoble cedex 9 (France)
[e] Dr. M. Jeanty
NovAliX, Centre de recherche Pharma
Campus de Maigremont, BP-615, 27106 Val-de-Reuil Cedex (France)
[f] Dr. X. Guinchard
Institut de Chimie des Substances Naturelles (ICSN)
Centre de Recherche CNRS
Avenue de la Terrasse, 91198 Gif-sur-Yvette (France)
Several approaches have been used to overcome antibiotic
resistance: 1) a more rational use of antibiotics, 2) decrease in
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 13
&
1
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
the spread of MDR bacteria through better hygiene and
a greater awareness of this public health problem by the gen-
eral population, healthcare workers, food professionals, and
farmers, and 3) the development of novel antibiotics.[9] Chemi-
cal modifications of known antibiotics have allowed the devel-
opment of new compounds (for example, tigecycline).[10] Only
two new classes of antibiotics, with original structures and
modes of action, have been developed in the last three de-
cades: the oxazolidinones (linezolid) and the cyclic lipopepti-
des (daptomycin). However, although their use is restricted to
specific clinical situations, resistance mechanisms have already
been reported for some of them.[9a,10] An alternative strategy
to the discovery of new antibiotics consists of combining exist-
ing antibiotics with drugs potentiating their activity, such as in-
hibitors of resistance mechanisms (for example, inhibitors of
antibiotic-inactivating or target-modifying enzymes and inhibi-
tors of efflux pumps).[9b] The usefulness of such combinations
has already been demonstrated by the previous development
of the association of a b-lactam compound and a b-lactamase
inhibitor (clavulanic acid, tazobactam, and sulbactam), al-
though resistance mechanisms to b-lactamase inhibitors have
now been reported.[9b,11] Clinical trials have evaluated the use-
fulness of an aerosolized formulation of a combination of ci-
profloxacin and MC-601,205, an efflux pump inhibitor, in cystic
fibrosis patients with pulmonary infections. Although these
clinical trials have been stopped, ongoing efforts are being
made[9b,11,12] in the field of efflux pump inhibitors (EPIs), which
appear to be a promising way to restore the activity of antibi-
otics against MDR bacteria and/or to decrease the selection of
bacteria harboring an MDR phenotype.[13]
structural information on the NorA efflux pump makes it diffi-
cult to design new inhibitors from structure–activity relation-
ship studies. The search for new EPIs has been mainly per-
formed by chemical library screening or classical medicinal ap-
proaches, as illustrated in the case of the evolution from natu-
ral flavones to 2-(4-propoxyphenyl)quinoline derivatives[16b] or
from 5’-methoxyhydnocarpin-D to the more active flavono-
lignans.[25] An alternative approach allowed German and co-
workers to identify, in 2008,[16a] an ofloxacin-derived EPI, the af-
finity of which for the efflux pump was related to its quinolone
scaffold.
We recently described the evaluation of the antibacterial ac-
tivity of series of indolic compounds against 29 bacterial
strains.[26] From the four indolic sub-libraries tested, the aldoni-
trones displayed minimal inhibitory concentrations (MICs)
greater than 128 mgLÀ1 for all of the tested Gram-negative
and Gram-positive bacterial strains, including S. aureus. These
compounds exhibited no antibacterial activity of their own,
which meant that they were suitable candidates for efflux
pump inhibition testing. We report herein the higher antibac-
terial activities of combinations of ciprofloxacin with these pre-
viously synthesized indolic derivatives,[26] relative to that of ci-
profloxacin alone, against S. aureus strains resistant to fluoro-
quinolones due to overexpression of the NorA efflux pump. To
enhance the solubility of the aldonitrones and improve their
activity, an extended series of related compounds was de-
signed, synthesized, and tested for biological activity in associ-
ation with ciprofloxacin. In addition to MIC determination, we
evaluated the accumulation of ethidium bromide (a NorA sub-
strate) in S. aureus strains to confirm NorA efflux pump inhibi-
tion. The toxicity of the most active compounds was measured
against three eukaryotic cell lines to evaluate their potential
use as therapeutic agents.
EPIs may act by two different types of mechanisms: 1) direct
binding to the transporter, which leads to competitive or non-
competitive inhibition of the efflux of the pump’s substrate, or
2) indirect action, through the collapse of the energy-driven
efflux pump mechanisms of the bacterial cell membrane or
through interference with the regulation system of the protein Results and Discussion
pump.[11,12,14]
Chemistry
To date, many structurally diverse compounds showing
NorA and/or MepA EPI activity have been identified, including
natural[15] and synthetic ones.[12,16] Reserpine is a reference and
one of the oldest EPI compounds, but it is only effective at
concentrations that are toxic for humans.[17] More recently,
omeprazole, a series of pyrrolo[1,2-a]quinaxoline derivatives
sharing structural analogies with omeprazole,[18] piperine,[16c]
aryl and phenyl piperidine derivatives,[16a] taraquinar,[19] the syn-
thetic acridone derivative GG918,[20] and 1,4-benzothiazine de-
rivatives[21] have been shown to display inhibitory activities
against S. aureus strains overexpressing the NorA efflux pump.
Among heterocyclic compounds, the indolic scaffold is found
in many compounds with biological activities,[22] including in
NorA efflux pump inhibitors, such as the indole alkaloid reser-
pine, 5-nitro-2-phenyl-1H-indole (INF55),[23] and the more re-
cently functionalized INF55.[24]
The syntheses of the 19 indolic aldonitrones of series 4 (4aa–
ci) and 4 of the indolic hydroxyamines of series 3 (3ca and
3cd) and 5 (5ca and 5cd) were described in our previous pub-
lication.[26] The pathway to this type of heterocyclic derivatives
is presented in Scheme 1. Thus, a coupling reaction between
the three nitrones 1a–c and the nine commercially available
indoles 2a–i in acidic conditions efficiently provided indolic
N-hydroxyamines 3. The latter compounds were then oxidized
to form aldonitrones 4 with manganese dioxide in toluene at
1008C. A hydroxyaminolysis of indolic nitrones 4 was per-
formed to yield N-hydroxyamines 5 (Scheme 1).
In an attempt to increase the solubility of the indolic com-
pounds while keeping their antibacterial activity in combina-
tion with ciprofloxacin, four additional molecules were de-
signed and synthesized by chemical modifications of com-
pound 3ca, an unhalogenated compound (Scheme 2). Simulta-
neous reduction and debenzylation of the N-benzyl-N-hydroxy-
amine function under mild conditions resulted in primary
amine 6 with an excellent yield.[27] An indolic derivative of for-
The structural diversity of the known NorA EPIs could result
from the significant structural flexibility of this efflux pump,
which allows the accommodation of diverse substrates in its
active site. However, to the best of our knowledge, the lack of
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 13
&
2
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
Scheme 1. Synthesis of 3-substituted indoles. Reagents and conditions:
a) HCl, MeOH, 08C, 2 h; b) MnO2, toluene, 1008C, 5 min; c) NH2OH·HCl,
MeOH, RT, 1 h. Bn: benzyl; Boc: tert-butoxycarbonyl.
Scheme 3. Chemical modifications of the indolic derivative 3cd. Reagents
and conditions: a) TiCl3, HCl(aq), MeOH, RT, 30 min; b) PhOCONHOH, DMSO,
508C, 24h; c) MnO2, toluene, 1008C, 5 min; d) NH2OH·HCl, MeOH, RT, 1 h;
e) BtCHO, THF, 08C, 10 min then RT, 30 min (12a); f) RCOCl, Et3N, CH2Cl2,
08C, 5 min (12b,c) or RT, 2h (12e,f); g) TFAA, NEt3, CH2Cl2, 15 min (12d).
BtCHO: N-formylbenzotriazole; THF: tetrahydrofuran; TFAA: trifluoroacetic
acid.
the N-hydroxyamine into compound 11 was not possible be-
cause of debromination of the 5-bromoindole moiety under
these conditions. Synthesis of primary amine 11 required
a three-step sequence: oxidation to the aldonitrone with man-
ganese dioxide, hydroxyaminolysis to the N-hydroxyamine, and
ultimate reduction with titanium chloride in acidic medium.[30]
The treatment of secondary amine 9 with phenyl N-hydroxy-
carbamate[29] led to the corresponding urea derivative 10 with
a poor yield. Formylation of the primary amine 11 with N-for-
mylbenzotriazole and acylation with the chlorides of several
acids or trifluoroacetic anhydride resulted in the six heterocy-
clic amides 12a–f with good yields.
Scheme 2. Chemical modifications of the indolic derivative 3ca. Reagents
and conditions: a) 40% Pd(OH)2, H2, AcOH, MeOH, RT, 40h; b) TFEF, MTBE,
DMF, 15 min , 08C then RT, 1h; c) AcCl, Et3N, CH2Cl2, 08C, 5 min; d) PhOCON-
HOH, DMSO, 508C, 2h. TFEF: 2,2,2-trifluoroethylformate; MTBE: methyl tert-
butyl ether; DMF: N,N-dimethylformamide; DMSO: dimethyl sulfoxide.
Finally, another part of N-hydroxyamine 3 cd was modified,
namely the CH2NHBoc side chain was elongated (Scheme 4).
For this purpose, 3-aminopropanol was converted into N-Boc-
aminoaldehyde 13 by using N-Boc protection and oxidation
steps. Condensation of aldehyde 13 with N-benzylhydroxy-
amine resulted in aldonitrone 14. The reaction of the latter
compound with 5-bromoindole led to indolic N-hydroxyamine
15 with a satisfactory yield.
The synthesis of 37 1-(1H-indol-3-yl)ethanamine derivatives
was thus achieved through a series of simple and efficient
chemical modifications. Amongst these compounds, 12 are
new. It should be mentioned that all of the synthesized mole-
cules have a chiral center and they have been tested as a race-
mic mixture. As a result of this work, several types of indolic
compounds, such as primary and secondary amines, N-hy-
mamide, 7a, was obtained by treatment of amine 6 with tri-
fluoroethyl formate.[28] Acylation of amine 6 with acetyl chlo-
ride resulted in acetamide 7b. Finally, treatment of amine 6
with phenyl N-hydroxycarbamate led to the corresponding
N-hydroxycarbamide 8 with a modest yield.[29]
Another strategy to enhance the solubility of indoles was to
modify the nitrone group while keeping the 5-bromoindole
substitution, because these compounds were shown to be
more biologically active than nonsubstituted ones.[26] There-
fore, several chemical transformations of N-hydroxyamine 3cd
were performed and resulted in nine indolic derivatives, of
which eight are new (Scheme 3). Reduction of the N-hydroxy-
amine function of indole 3cd by TiCl3 in the presence of HCl
led to the secondary amine 9. In contrast to the case with in-
dolic compound 3ca, the use of Pd(OH)2 for the reduction of
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 13
&
3
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
None of the aldonitrones showed intrinsic antibacterial activ-
ity against SA-1199B at the highest concentrations tested. By
contrast, all of the compounds potentiated the antibacterial
activity of ciprofloxacin (CIP) added at sub-inhibitory concen-
tration. With the MIC value of ciprofloxacin against SA-1199B
being 8 mgLÀ1, three sub-inhibitory concentrations (MIC/2,
MIC/4, and MIC/8) were tested. The less active compounds
were those that were unhalogenated on the indole core and/
or the aldonitrones with a short aliphatic chain (R: Me or iBu),
rather than the CH2NHBoc side chain. Nitrones 4aa and 4ba,
which cumulate these two flaws (that is, unsubstituted at the
C5 position and with the methyl or isobutyl alkyl chain of the
aldonitrone) were the least active, with MIC values of 32 and
16 mgLÀ1, respectively, in the presence of 4 mgLÀ1 of CIP. The
most active compounds were those that were substituted at
Scheme 4. Synthesis pathway of an analogue of the indolic derivative 3cd.
Reagents and conditions: a) Boc2O, CH2Cl2, RT, 24h; b) IBX, EtOAc, reflux, 3h;
c) BnNHOH, MgSO4, CH2Cl2, RT, 3h; d) HCl, MeOH, 08C, 6h. IBX: 2-iodoxyben-
zoic acid.
droxyamines, aldonitrones, formamides, amides, and
N-hydroxycarbamides, were elaborated.
Table 1. Minimal inhibitory concentration of indolic derivatives alone or in the pres-
ence of a sub-inhibitory ciprofloxacin concentration against S. aureus 1199B and
K2378 strains.[a]
Indoles
MIC [mgLÀ1
]
Biological activity
1199B
K2378
0.25
[cipro.]:[b]
0
1
2
4
0
–
0.5
–
In a previous work,[26] the indolic compounds from
series 3 to 5 were tested for their antimicrobial activ-
ity against a panel of Gram-negative and Gram-posi-
tive bacterial strains representative of common
human pathogens. The MIC values, that is, the mini-
mal concentration allowing complete inhibition of
bacterial growth as assessed by visual control and by
absorbance measurement at 620 nm, showed that
compounds from these sub-libraries were completely
inactive when tested at 128 mgLÀ1 against Gram-
positive and Gram-negative bacteria. The lack of in-
trinsic antibacterial activity of these compounds al-
lowed us to test their activity as NorA efflux pump
inhibitors.
3ca
3cd
4aa
4ab
4ac
4ad
4ae
4ba
4bb
4bc
4bd
4be
4ca
4cb
4cc
4cd
4ce
4cf
4cg
4ch
4ci
5ca
5cd
6
7a
7b
8
9
10
11
>32
16
32
8
16
8
2
32
4
4–2
4–2
4
16
8
4
2
1
8
2
0.25
2
0.5
16
8
4
4
16
2
64
64
64
0.5
1
4
16
8
8
1
2
4
–
–
–
–
–
–
–
–
–
–
–
–
–
–
16
8
8
–
–
8
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
4
64
16
8
16
4
>128
>16
>16
>16
>16
>32
>16
>16
>16
>16
>32
>16
>16
>16
>16
>16
>16
>16
>16
>128
>32
>128
>128
>128
>128
8
>128
>16
>16
>16
>16
>32
>16
>16
>16
>16
>32
>16
8
4
2
>16
16
16
>16
64
16
–
–
–
–
–
–
–
–
–
–
–
–
>16
>16
>16
–
–
>16
–
–
–
–
–
–
–
–
–
–
–
–
–
–
8
16
>32
>16
8
4
4
32
8
2
2
0.5
4
2
2
The MIC results for aldonitrones 4aa–ci against
SA-1199B, a NorA-overexpressing strain, are summar-
ized in Table 1. First, the antibacterial activity of each
indolic compound alone was tested at a concentra-
tion range starting from the maximal concentration
to ensure complete solubilization of the compound
in Mueller–Hinton (MH) medium. Indeed, at the
128 mgLÀ1 concentration, all of the tested aldoni-
trones, except 4aa, were found to be partly insoluble
in MH medium, as assessed by visual control. The un-
halogenated aldonitrones, that is, 4aa, 4ba, and
4ca, were the more soluble compounds, with solu-
bility obtained at concentrations of >128, >32, and
>32 mgLÀ1, respectively. For the other compounds
in this series, the maximal concentration with com-
plete solubilization was 16 mgLÀ1. Therefore, the in-
dolic compounds were tested at a concentration
range of 16–0.015 mgLÀ1, as recommended by the
Clinical and Laboratory Standards Institute (CLSI) for
water-insoluble compounds used in broth dilution
susceptibility tests.
16
–
–
8
4
2
8
16
64
–
–
8
32
>128
>128
>128
32
>128
>128
>128
8
–
–
–
–
–
–
>128
>128
8
–
–
–
8
16
32
64
64
8
32
32
8
8
8
>16
64
–
–
–
–
>16
–
>32
–
–
>32
16
32
32
4
12a
12b
12c
12d
12e
12 f
15
>128
>128
>16
ꢀ32
>32
8
–
2
8
–
16
2
8
4
16
1
2
–
–
–
>16
>32
8
4
reserpine
–
8
[a] At least two independent experiments were performed for each MIC measurement.
Due to the experimental procedure, MIC values are reported to within a factor of 2.
[b] Ciprofloxacin concentrations are given in mgLÀ1
.
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 13
&
4
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
the C5 position of the indole core with a halogen atom,
namely 4bd, 4be, and 4cb–e, with maximum MIC values of
2 mgLÀ1 in the presence of 4 mgLÀ1 of CIP. Halogenated com-
pounds with a CH2NHBoc side chain showed MIC values of 4
and 0.5 mgLÀ1 for 4ci and 4ce, respectively. The indolic deriv-
atives of series 4ba–e and 4ca–e varied only with respect to
the substituents at the indole C5 position, which allowed
direct conclusions to be drawn regarding the importance of
the 5-halogen group. The presence of a nitro group in that po-
sition had already been shown to lead to more active 2-
phenyl-1H-indole derivatives, relative to the unsubstituted
ones.[24a] In the present work, we confirm the importance of
the presence of an electron-withdrawing group in the C5 posi-
tion of the indole core.
ium bromide efflux inhibition. For reserpine, a well-known
NorA inhibitor, the MIC value of which is >32 mgLÀ1 against
SA-1199B (Table 1), a concentration of 20 mgLÀ1 is considered
relevant and is widely used in EtBr accumulation tests reported
in the literature.[31] However, because of their low solubility, the
aldonitrone compounds 4cc–e and 4ch could not be accurate-
ly prepared at a concentration higher than 16 mgLÀ1. There-
fore, it was decided to assay EtBr fluorescence in the presence
of 5 mgLÀ1 of the tested compounds, that is, at a concentration
lower than their maximum solubility and MIC value.
The effects of the indolic compounds on the EtBr efflux ca-
pability of SA-1199B, with reserpine as a reference, are shown
on Figure 1 and in Table 2. All four of the tested aldonitrones
allowed a significant EtBr fluorescence increase relative to the
The MIC values of the indolic compounds were dependent
on the ciprofloxacin concentration used: for a given com-
pound, the MIC values were higher when the CIP concentra-
tion was decreased from 4 mgLÀ1 to 1 mgLÀ1. A similar rank-
ing between the tested compounds was observed at CIP con-
centrations lower than 4 mgLÀ1. However, in the presence of
a CIP concentration that was an eighth of the MIC value (that
is 1 mgLÀ1), only three compounds, 4cc–e, showed an MIC
value lower than 16 mgLÀ1. The lowest MIC value in this case,
2 mgLÀ1, was observed for 4ce. The 5-iodine derivative was
thus able to significantly potentiate CIP antibacterial activity at
a rather low concentration.
It was checked that the indolic derivatives did not induce
any potentiation of the activity of CIP at the MIC/4 concentra-
tion against SA-1199 (data not shown), the parental strain of
SA-1199B that is susceptible to CIP (MIC=0.25 mgLÀ1), to ex-
clude any synergistic effect in association with ciprofloxacin,
whereas positive results were obtained against the NorA-over-
expressing strain SA-1199B, which is consistent with the hy-
pothesis of NorA efflux pump inhibition as the mode of action.
However, the SA-1199B strain is resistant to fluoroquinolones
because of overexpression of NorA and also because of muta-
tions in the DNA topoisomerase gene grlA.[23] Thus, to address
the question of specific NorA efflux pump inhibition, four com-
pounds (4cc–e and 4ch) that were active against SA-1199B
were further tested against K2378, an S. aureus strain that over-
expresses NorA from a multicopy plasmid but has no muta-
tions on DNA gyrase encoding genes.[21] We first checked that
the MIC value of these compounds alone was >16 mgLÀ1
against K2378. The MIC values (Table 1) were then measured in
the presence of MIC/4 and MIC/8 CIP concentrations, that is,
0.5 mgLÀ1 and 0.25 mgLÀ1, respectively. All four compounds
displayed an MIC value of 2–4 mgLÀ1 against the K2378 strain,
which corroborated their potential activity as NorA efflux
pump inhibitors.
Figure 1. Effect of compounds 4cc (~), 4cd (~), 4ce (&), and 4ch (&) on
EtBr accumulation of SA-1199B. The tested compound was added after
10 min of bacterial incubation with EtBr. The positive control was reserpine
(*) at 5 mgLÀ1 and the negative control (*) was the bacteria in the pres-
ence of 1.6% DMSO.
Table 2. Ethidium bromide fluorescence increase in the presence of in-
dolic derivatives.[a]
Indoles
Fluorescence increase [%]
1199B
K2378
5 mgLÀ1
20 mgLÀ1
5 mgLÀ1
20 mgLÀ1
reserpine
4cc
4cd
4ce
4ch
5cd
8
31Æ2
116 Æ5
146Æ6
173Æ5
71Æ2
29Æ1
–
78Æ6
166
163
180
153
–
62Æ1
–
–
45Æ1
–
76Æ8
–
68Æ10
56Æ8
71Æ9
27Æ9
–
–
–
–
–
7Æ1
–
12c
15
156Æ5
98Æ15
44Æ19
–
–
[a] Data represent the meanÆ standard deviation of two or three inde-
pendent experiments performed at least in duplicate.
To confirm the EPI properties of the most active indolic com-
pounds among series 3 to 5, we evaluated their ability to inter-
fere with ethidium bromide (EtBr) efflux from SA-1199B. This
test necessitates using EPIs at concentrations at which they
have no antibacterial activity on their own, to exclude the pos-
sibility that any observed EtBr fluorescence increase could be
related to cellular death and the subsequent collapse of the
NorA pump proton driven activity rather than to active ethid-
negative control. The efflux inhibition percentage calculated
from the EtBr fluorescence increase (see the Experimental Sec-
tion for calculation details) showed that the EtBr accumulation
in the presence of the nitrones was two- to fivefold higher
than that of reserpine at the same concentration (Table 2)
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 13
&
5
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
when SA-1199B was used. The compound that was brominated
at position 6 on the indole core, 4ch, was twofold less efficient
than the analogue substituted at position 5, 4cd, which corre-
lated well with the antibacterial effects of these compounds.
Among the 5-substituted compounds, the iodine one, 4ce,
was the most active EPI, with a fivefold increase of EtBr accu-
mulation relative to that with reserpine. However, both the
chlorine and the bromine derivatives (4cc and 4 cd, respective-
ly) also show significant effects on the EtBr accumulation at
5 mgLÀ1. The chlorine compound 4cc was almost fourfold
more potent than reserpine at 5 mgLÀ1. In the presence of
S. aureus K2378, a resistant strain built to overexpress NorA,
the ethidium bromide accumulation levels obtained for both
reserpine and compound 4ch were similar to those obtained
against S. aureus 1199B, which corroborates the hypothesis of
potential inhibition of the NorA pump as a mode of action for
these compounds. However, the EtBr accumulation remained
significant but lower for 4cc–e relative to the results obtained
with S. aureus 1199B, which would suggest that these products
are not really NorA inhibitor specific.
9–12 and 15. The latter compound was similar to 3cd, except
for the length of the nitrogen side chain, with 15 including an
additional methylene group. However, this structural modifica-
tion induced no change in the biological results of 15 relative
to those of parent compound 3cd.
The nitrone group in 4cd or N-hydroxyamine in 3cd was re-
placed by a secondary amine in 9. As already shown with com-
pound 3cd, compound 9 showed a significant antibacterial ac-
tivity when tested alone against the fluoroquinolone-suscepti-
ble S. aureus ATCC 25923 strain (data not shown) and the fluo-
roquinolone-resistant SA-1199B strain (Table 1), with MIC values
of 8 and 16 mgLÀ1 for 9 and 3cd, respectively. The high toxici-
ty of compound 9 observed against three cell lines is likely to
be related to its antimicrobial activity (Table 3). When com-
bined with ciprofloxacin at 4 mgLÀ1 against SA-1199B, the MIC
Table 3. Cell growth inhibition of normal (human lung fibroblast: MCR5)
and cancer (human mouth carcinoma: KB; human colon tumor: HCT116)
cell lines.[a]
In an attempt to increase the low solubility of the indolic
compounds in series 3 to 5, which prevented an accurate de-
termination of their MIC values, four additional unhalogenated
derivatives, 6 to 8, were synthesized. It was previously ob-
served in the present work that the replacement of the nitrone
group in both 4ca and 4cd by a hydroxyamine (3ca and 3cd,
respectively) was not detrimental to the MIC value. Conse-
quently, it was decided to replace the nitrone with an amine,
6, and to an amide in a further step by formylation or acylaton
of amine 6. As expected, compounds 6 to 8 were all fully solu-
ble at 128 mgLÀ1 and were inactive alone against both the
ATCC 25923 (data not shown) and 1199B S. aureus strains
(MIC>128 mgLÀ1), which allowed unambiguous interpretation
of the results obtained in association with ciprofloxacin. How-
ever, the primary amine 6 and the two amides 7a and 7b
were not active, with MIC>128 mgLÀ1 in the presence of CIP
at MIC/4 (Table 1).
Indole
Growth inhibition [%] at 10À5 m/10À6
m
KB
MRC5
HCT116
4cb
4cc
4cd
4ce
4cf
4cg
4ch
4ci
8
0
0
57/5
56
0
0
0
46
nd
26
8
18
47
75/14
39
0
0
21
87/5
0/1
0
30
42
88/0
0/0
100/0
33/5
47/0
41/0
16/0
84/0
4/0
92/0
0/0
100/0
5/8
13/0
10/0
0/0
60/0
16/0
9
10
100/0
25/11
35/0
0/0
12c
12d
12e
12 f
15
0/0
88/0
42/11
[a] At least three independent experiments were performed for each
measurement: data represent the average. nd: not determined.
The results obtained with compound 8 appeared to deviate
from those of the same sub-library. Indeed, although the po-
tency of derivative 8 was approximately the same as that of
the parent compound 3ca in the presence of 2 mgLÀ1 of CIP,
the MIC value decreased to 0.5 mgLÀ1 when the ciprofloxacin
concentration was increased twofold. Under the same condi-
tions, the MIC value of 3ca was 16-fold higher. However, 8 was
completely inactive against the NorA-overexpressing SA-K2378,
which strongly suggests that its mode of action is not related
to the inhibition of the NorA efflux pump. As 8 significantly en-
hanced ethidium bromide accumulation in SA-1199B, a possible
mode of action could be the inhibition of S. aureus efflux
pumps other than NorA.
values of compounds 3cd and 9 decreased eightfold to 2 and
1 mgLÀ1, respectively, which corresponded to a significant syn-
ergistic effect. Similar results were obtained against K2378 for
3cd, whereas the MIC value of compound 9 against K2378 re-
mained unchanged in the presence or absence of CIP (Table 1).
However, because of the intrinsic activity of these compounds,
no conclusion can be drawn about their potent EPI activity.
Among the 5-Br derivatives 12, the amides with the shortest
substituents, namely H and Me groups, were fully soluble at
128 mgLÀ1. The objective of increasing the solubility of the
halogenated compounds was thus reached. Moreover, both in-
dolic formamide 12a and amide 12b showed a high MIC
value (>128 mgLÀ1), which allowed unambiguous interpreta-
tion of the results obtained in association with ciprofloxacin.
Compounds 12c–f were less soluble because of the aromatic
or trifluoroacetyl substituent on the amide. Consequently, it
can only be stated that the MIC values of 12c–e were higher
Due to the disappointing results we obtained with the unha-
logenated compounds, it was decided to further enlarge the
5-bromoindole sub-library from the common 3cd–5cd scaffold
(5-Br substitution and carbamate side chain). Indeed, com-
pounds 3cd–5cd had been found to be active as antibacterial
agents and/or as efflux inhibitors (Table 1). Substituent modifi-
cations at the side-chain nitrogen atom led to new derivatives
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 13
&
6
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
than 16–32 mgLÀ1 when used alone against SA-1199B, without
further precision. p-Bromobenzamide 12 f, the most active
compound in the 12 series, showed significant antibacterial ac-
tivity when used alone, which precluded any conclusion about
its potential EPI activity. As with compound 9, this antimicrobi-
al activity was correlated to a high cellular toxicity in the same
concentration range (Table 3). Two derivatives, 12d and 12e,
were significantly active in the presence of ciprofloxacin. Com-
pound 12e, substituted with a nicotinic acid, displayed an in-
termediate MIC value of 4 mgLÀ1 in the presence of 4 mgLÀ1
of ciprofloxacin. Trifluoroacetamide 12d was found to have an
MIC value of only 2 mgLÀ1 at the same CIP concentration. Ben-
zamide 12c was one of the most active in the 12 series, with
an MIC value as low as 1 mgLÀ1 in the presence of 4 mgLÀ1 of
CIP. In the presence of 1 mgLÀ1 of CIP, 12c exhibited an MIC
value of 8 mgLÀ1, which was one of the lowest observed in
this work under the same conditions. Compound 12c was also
active against SA-K2378 in the presence of ciprofloxacin, with
results similar to those obtained with the nitrone compound
4 cd, which showed that a charged group is not a requirement
for NorA inhibitors. Indeed, it has been hypothesized that, al-
though the three-dimensional structure of the NorA protein
can only be predicted based on models,[32] inhibitors bind to
the protein through both hydrophobic and electrostatic inter-
actions, inside a large hydrophobic binding site on the protein.
Aldonitrone derivatives 4aa–ci are charged and present large
hydrophobic areas and thus exhibit all of the known requisites
to provide favorable interactions with NorA. However, when
the nitrone group (in series 4) was substituted by various polar
ones, namely primary and secondary amines, N-hydroxyamines,
or amides, the EPI activity was retained.
ents, leading to the formation of radicals inducing cellular tox-
icity. Among the 5-bromo derivatives, 12c and 4cd, between
which the only structural difference is the replacement of the
nitrone function in 4cd by an amide in 12c, showed similar
toxicity against eukaryotic cells.
Conclusions
The present work shows the potent NorA efflux pump inhibi-
tion of a new family of indole compounds. The most efficient
were the indolic aldonitrones 4cc–ce, halogenated at posi-
tion 5 of the indole core. Both chlorine and bromine deriva-
tives showed MIC values as low as 2 mgLÀ1 in combination
with sub-inhibitory concentrations of CIP (MIC/4); these values
are similar to that observed for the best indolic structure
known to date, that is,
a 5-nitro-2-phenylindole called
INF55,[24a] which is capable of producing a fourfold increase in
S. aureus susceptibility to ciprofloxacin when co-administered
with the antibiotic at a concentration of 1.5 mgLÀ1. Moreover,
the efficiency of this lead was even increased with the 5-iodine
aldonitrone 4ce, which showed
a lower MIC value of
0.5 mgLÀ1. In addition, it was shown that the EPI activity of
these derivatives was superior to that shown for the reference
compound reserpine against both SA-1199B and K2378. These
results show a significant potentiating effect of the indolic
compounds. However, from the standpoint of the potential
use of these compounds in clinical situations, the iodine sub-
stituent is rather labile and its release has been reported to be
responsible for the toxicity of iodine-containing compounds.
Therefore, additional compounds should be synthesized with
a chlorine substituent, which is known to be less labile than
iodine.
Three compounds, 5cd, 12c, and 15, within the 5-Br series
and among those with the highest MIC values in the absence
of CIP and the lowest MIC values in the presence of CIP were
tested for their ability to allow EtBr accumulation in S. aureus
strains (Table 2). All three compounds induced an efflux inhibi-
tion of EtBr when tested with SA-1199B, with the most efficient
one being benzamide 12c, which increased EtBr fluorescence
fivefold relative to reserpine.
Replacement of the charged nitrone group by an amide one
in the 5-Br derivative 12c led to unchanged MIC values in the
presence of CIP at MIC/4, as well as a similar EtBr efflux inhibi-
tion that reached fivefold the value observed for the reference
compound reserpine, whereas the toxicity was similar to that
of the parent nitrone 4cd.
Within the amide series, tert-butyl (2-(5-bromo-1H-indol-3-
yl)-2-(nicotinamido)ethyl)carbamate 12e and tert-butyl (2-(5-
bromo-1H-indol-3-yl)-2-(2,2,2-trifluoroacetamido)ethyl)carba-
mate 12d showed no to very low toxicity at 10À5 m, a concen-
tration similar or even higher than the MIC values in the pres-
ence of CIP (MIC/4).
To assess the potential of the indolic compounds as thera-
peutic antibacterial drugs, their toxicity against several eukary-
otic lines was evaluated (Table 3). All of the compounds were
tested at the same concentrations, expressed in molar. As their
molecular weight varied from 334 gmolÀ1 (for 8) to
551 gmolÀ1 (for 12 f), the corresponding mass concentrations
varied from 3–5.5 mgLÀ1, respectively, according to the units
usually used to express MIC values. Among the nitrone deriva-
tives, the fluorine and chlorine substituents at position 6 (4cf
and 4cg, respectively) showed no toxicity at 10À5 m, namely
~4 mgLÀ1. It should be noticed that the MIC value of chlorine
compound 4cg was 8 mgLÀ1 in the presence of 2 mgLÀ1 of
CIP, that is, the compound could be used to have a therapeutic
effect against bacteria without being toxic toward mammalian
cells.
In addition, a new indole hydroxycarbamate, 8, was identi-
fied with significant antibacterial activity against SA-1199B in
combination with ciprofloxacin. However, the compound was
inactive when tested against the NorA-overexpressing strain
K2378 and was shown to be unable to inhibit EtBr efflux,
which suggests that this compound was not a NorA efflux
pump inhibitor. The mode of action of this compound and the
reason for its selective activity remains to be investigated. One
cellular target of compound 8 could be an efflux pump other
than NorA but involved in fluoroquinolone resistance.
The iodine and bromine derivatives were slightly toxic at
10À5 m as a result of the weaker carbon/halogen bond
strength, relative to that of the fluorine and chlorine substitu-
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 13
&
7
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
tert-Butyl (2-formamido-2-(1H-indol-3-yl)ethyl)carbamate (7a):
TFEF (34 mL, 0.35 mmol) was added to a solution of indolic amine 6
(88 mg, 0.32 mmol) in a mixture of dry MTBE (15 mL) and DMF
(2 mL) at 08C. The resulting mixture was stirred for 15 min at 08C
and then it was allowed to reach room temperature. After 1h,
MTBE (30 mL) was added to the mixture. The organic solution was
then washed with H2O and brine, dried over anhydrous Na2SO4, fil-
tered, and concentrated in vacuo. Pure compound 7a was ob-
tained as a white solid (93 mg, 96%): 1H NMR (300 MHz, CD3OD):
d=1.42 (s, 9H), 3.20–3.31 (m, 1H), 3.37–3.59 (m, 1H), 5.48 (dd, J1 =
8.4, J2 =5.5 Hz, 1H), 6.58 (t, J=6.5 Hz, 1H), 7.00–7.14 (m, 2H), 7.23
(s, 1H), 7.35 (d, J=8.1 Hz, 1H), 7.63 (d, J=7.9 Hz, 1H), 8.13 ppm (d,
J=0.8 Hz, 1H); 13C NMR (75 MHz, CD3OD): d=28.8, 45.6, 46.6, 79.8,
112.5, 119.7, 120.1, 120.3, 122.8, 123.0, 123.0, 127.4, 162.4,
163.4 ppm; IR (neat): n˜ =3415, 3305, 3055, 2980, 2930, 2860, 1690,
1655, 1540, 1515, 1435, 1375, 1280, 1155, 735 cmÀ1; MS (ESI+): m/z
(%): 326 (100) [M+Na]+; HRMS (ESI+): m/z calcd for C16H21N3NaO3
[M+Na]+: 326.1475; found: 326.1475; m/z calcd for C16H22N3O3
[M+H]+: 304.1656; found: 304.1656.
Experimental Section
Chemistry
General: Purchased reagents and solvents were used without pu-
rification except those indicated below. Tetrahydrofuran and
methyl tert-butyl ether were heated at reflux over sodium benzo-
phenone and then distilled. CH2Cl2, N,N-dimethylformamide,
MeOH, and Et3N were dried by heating at reflux over CaH2 and
then distilled. Acetyl chloride and benzoyl chloride were distilled
over CaCl2. Reactions were monitored by thin layer chromatogra-
phy with commercial aluminum-backed silica gel plates (Merck G
60 F254). TLC spots were viewed under ultraviolet light and by heat-
ing the plate after treatment with either a 0.5% solution of ninhy-
drin in 3% ethanolic acetic acid or a 2% solution of potassium per-
manganate in 7% aqueous Na2CO3. N-Hydroxyamines were detect-
ed with a 1% triphenyltetrazolium chloride in EtOH (red color).
Product purification by gravity column chromatography was per-
formed by using Macherey–Nagel silica gel 60 (70–230 mesh). Infra-
red spectra were recorded on a Nicolet Impact-400 Fourier trans-
form infrared spectrometer or on a Nicolet Magna-550 Fourier
transform infrared spectrometer equipped with an attenuated total
reflection (ATR) device and the data are reported in reciprocal cen-
tert-Butyl (2-acetamido-2-(1H-indol-3-yl)ethyl)carbamate (7b):
Acetyl chloride (23 mL, 0.6 mmol) was added to a solution of indolic
amine 6 (138 mg, 0.5 mmol) and Et3N (101 mg, 1.0 mmol) in dry
CH2Cl2 (5 mL) at 08C. The resulting mixture was stirred for 5 min at
08C and then the CH2Cl2 was evaporated. Diethyl ether (50 mL)
was added to the mixture, and the resulting organic solution was
washed with an aqueous solution of HCl (1m), a saturated aqueous
solution of NaHCO3, and brine, dried over anhydrous Na2SO4, fil-
tered, and concentrated in vacuo. Column chromatography with
EtOAc/pentane (10:90–99:1) yielded pure compound 7b as
1
timeters (cmÀ1). H NMR and 13C NMR spectroscopy was performed
1
on a Bruker Advance300 spectrometer. Chemical shifts for H spec-
tra are values downfield from the internal solvent signal (CD3OD:
d=3.31 ppm; [D6]DMSO: d=2.5 ppm) or from tetramethylsilane in
CDCl3 (d=0.00 ppm) and are reported as follows: chemical shifts
(d) in parts per million (ppm), multiplicity, coupling constants (J) in
Hertz (Hz), and integration. Low-resolution mass spectra were re-
corded on a ThermoFinnigan PolarisQ ion-trap spectrometer by
using direct chemical ionization (DCI; ammonia/isobutane, 63:37)
or on a Bruker Esquire 3000 plus (ESI). High-resolution mass spectra
were recorded on a Bruker maXis mass spectrometer by the “Fꢂd-
ꢂration de Recherche” ICOA/CBM (FR2708) platform, Orleans,
France. Elemental analyses were performed at the “Service d’Ana-
lyse Elꢂmentaire” of the Dꢂpartement de Chimie Molꢂculaire, Uni-
versitꢂ Joseph Fourier (Grenoble, France).
1
a white solid (125 mg, 79%): H NMR (300 MHz, CD3OD): d=1.42 (s,
9H), 1.97 (s, 3H), 3.48 (dd, J1 =13.6, J2 =8.5 Hz, 1H), 3.59 (dd, J1 =
13.6, J2 =5.5 Hz, 1H), 5.40 (dd, J1 =8.5, J2 =5.5 Hz, 1H), 6.99–7.13
(m, 2H), 7.22 (s, 1H), 7.34 (dt, J1 =8.0, J2 =1.0 Hz, 1H), 7.62 ppm
(dt, J1 =7.5, J2 =1.2 Hz, 1H); 13C NMR (75 MHz, CD3OD): d=22.9,
28.8, 45.6, 48.0, 80.2, 112.4, 114.9, 119.8, 120.1, 122.7, 123.1, 127.6,
138.2, 158.8, 172.9 ppm; IR (neat): n˜ =3310, 3215, 2970, 2930, 1690,
1630, 1525, 1435, 1285, 1225, 1170, 1150, 740 cmÀ1; MS (ESI+): m/z
(%): 340 (100) [M+Na]+; HRMS (ESI+): m/z calcd for C17H23N3NaO3
[M+Na]+: 340.1632; found: 340.1631; m/z calcd for C17H24N3O3
[M+H]+: 318.1812; found: 318.1811.
The preparation of indolic compounds 3ca, 3cd, 4aa–4ci, 5ca,
and 5cd was previously described.[26]
tert-Butyl (2-amino-2-(1H-indol-3-yl)ethyl)carbamate (6):[30] The
Pearlman catalyst (Pd(OH)2, 152 mg, 40% w/w) was added to
tert-Butyl (2-(3-hydroxyureido)-2-(1H-indol-3-yl)ethyl)carbamate
(8): A flask was charged with the amino compound 6 (100 mg,
0.36 mmol), phenyl N-hydroxycarbamate (56 mg, 0.36 mmol), and
DMSO (1 mL). The mixture was stirred for 2 h at 508C. The crude
solution was diluted in EtOAc (25 mL) and washed with brine and
water. The organic layer was dried over anhydrous MgSO4 and con-
centrated in vacuo. After flash chromatography (EtOAc/pentane,
7:3 then neat EtOAc), compound 8 was obtained as a white solid
a
stirred solution of indolic N-hydroxyamine 3ca (381 mg,
1.0 mmol) in a mixture of MeOH (20 mL) and acetic acid (0.7 mL).
The argon atmosphere was replaced by hydrogen. The resulting
mixture was stirred at room temperature for 40 h. It was then fil-
tered through Celite. The resulting filtrate was treated with an
aqueous solution of NaOH (6n) until pH 12 was reached. The
MeOH was then evaporated in vacuo. The resulting aqueous mix-
ture was extracted with EtOAc (3ꢃ20 mL). The combined organic
layers were washed with H2O and brine, dried over anhydrous
Na2SO4, filtered, and concentrated in vacuo. Pure compound 6 was
1
(60 mg, 50%): mp 1888C; H NMR (300 MHz, [D6]DMSO): d=1.37 (s,
9H), 3.32–3.48 (m, 2H), 5.12–5.20 (m, 1H), 6.68 (d, J=8.7 Hz, 1H),
6.80 (t, J=5.4 Hz, 1H), 6.97 (t, J=7.4 Hz, 1H), 7.07 (t, J=7.4 Hz,
1H), 7.23 (d, J=2.1 Hz, 1H), 7.34 (d, J=7.4 Hz, 1H), 7.61 (d, J=
7.4 Hz, 1H), 8.33 (brs, 1H), 8.52 (brs, 1H), 10.89 ppm (brs, 1H);
13C NMR (75 MHz, [D6]DMSO): d=29.2, 45.5, 46.9, 78.6, 112.3, 116.0,
119.3, 119.8, 121.9, 123.3, 127.1, 137.2, 156.7, 161.8 ppm; IR (neat):
n˜ =3415, 3390, 3355, 3195, 2975, 1685, 1620, 1560, 1520,
1170 cmÀ1; MS (ESI+): m/z (%): 357 (100) [M+Na]+; HRMS (ESI+):
m/z calcd for C16H23N4O4 [M+H]+: 335.1714; found: 335.1713.
1
obtained as a white solid (260 mg, 95%): mp 145–1468C; H NMR
(300 MHz, CDCl3): d=1.44 (s, 9H), 1.61 (brs, 2H), 3.33–3.43 (m, 1H),
3.53–3.61 (m, 1H), 4.40 (dd, J1 =7.2, J2 =5.6 Hz, 1H), 4.90 (brs, 1H),
7.09–7.23 (m, 3H), 7.36 (dd, J1 =8.0, J2 =0.8 Hz, 1H), 7.70 (d, J=
7.7 Hz, 1H), 8.33 ppm (brs, 1H); 13C NMR (75 MHz, CDCl3): d=28.4,
47.5, 48.7, 79.2, 111.3, 117.8, 119.3, 119.6, 121.0, 122.3, 126.0, 136.6,
156.7 ppm; IR (neat): n˜ =3404, 3260, 2970, 2920, 1690, 1575, 1440,
1365, 1295, 1250, 1155, 980, 740 cmÀ1; MS (ESI+): m/z (%): 298 (19)
[M+Na]+, 276 (10) [M+H]+, 203 (100); elemental analysis: calcd
for C15H21N3O2: C 65.43, H 7.69, N 15.26; found: C 65.22, H 7.69, N
15.19.
tert-Butyl (2-(benzylamino)-2-(5-bromo-1H-indol-3-yl)ethyl)car-
bamate (9): A 15% aqueous solution of titanium trichloride
(9.1 mL, 8.6 mmol) was added to a stirred solution of indolic N-hy-
droxylamine 3cd (1.83 g, 3.96 mmol) in MeOH (40 mL). The result-
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 13
&
8
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
ing mixture was stirred at room temperature for 30 min. A large
excess of an aqueous solution of NaOH 20% saturated with NaCl
was added. MeOH was then removed in vacuo and the crude mix-
ture was extracted with EtOAc (3ꢃ20 mL). The combined organic
layers were washed with H2O and brine, dried over anhydrous
Na2SO4, filtered, and concentrated in vacuo. After flash chromatog-
raphy (EtOAc/pentane, 4:6 then 7:3), pure compound 9 was ob-
80.3, 113.4, 114.1, 114.5, 122.4, 124.8, 125.6, 129.2, 136.8, 163.4,
168.5 ppm; IR (neat): n˜ =3270, 3170, 3080, 2975, 2930, 1690, 1655,
1530, 1450, 1380, 1290, 1255, 1150, 790, 755 cmÀ1; MS (ESI+): m/z
(%): 404 (100) and 406 (97) [M+Na]+; HRMS (ESI+): m/z calcd for
C16H21BrN3O3 [M+H]+: 382.0761 and 384.0742; found: 382.0760
and 384.0741.
1
tert-Butyl
(2-acetamido-2-(5-bromo-1H-indol-3-yl)ethyl)carba-
tained as a white solid (1.20 g, 58%): H NMR (300 MHz, [D6]DMSO):
mate (12b): Compound 12b was obtained from indolic amine 11
(177 mg, 0.5 mmol) and acetyl chloride (23 mL, 0.6 mmol) by the
same procedure as compound 7b. The product was a white solid
d=1.34 (s, 9H), 2.41 (brs, 1H), 3.23 (t, J=6.0 Hz, 2H), 3.59 (dd, AB
system, J=13.5, dAÀdB =42.0 Hz, 2H), 3.99 (t, J=6.0 Hz, 1H), 6.81
(brs, 1H), 7.16 (dd, J=8.7, J=1.8 Hz, 1H), 7.20–7.33 (m, 7H), 7.83
(d, J=1.8 Hz, 1H), 11.10 ppm (s, 1H); 13C NMR (75 MHz, [D6]DMSO):
d=28.1, 46.0, 50.4, 54.4, 77.4, 110.8, 113.3, 115.1, 121.4, 123.2,
124.5, 126.3, 127.7, 127.9, 128.3, 135.2, 141.0, 155.6 ppm; IR (neat):
n˜ =3421, 3292, 3027, 1683, 1506, 1453, 1391, 1365, 1248, 1163,
1100, 884 cmÀ1; MS (DCI; ammonia/isobutane, 63:37): m/z (%): 444
(100) and 446 (97) [M+H]+; HRMS (ESI+): m/z calcd for
C22H27BrN3O2 [M+H]+: 444.1281 and 446.1263; found: 444.1281
and 446.1264; elemental analysis: calcd for C22H26BrN3O2: C 59.46,
H 5.86, N 9.46; found: C 59.25, H 5.92, N 9.33.
1
(150 mg, 76%): H NMR (300 MHz, CD3OD): d=1.41 (s, 9H), 1.97 (s,
3H), 3.43–3.58 (m, 2H), 5.34 (dd, J1 =7.8, J2 =6.3 Hz, 1H), 7.19 (dd,
J1 =8.6, J2 =1.7 Hz, 1H), 7.26 (s, 1H), 7.27 (d, J=8.6 Hz, 1H),
7.77 ppm (d, J=1.7 Hz, 1H); 13C NMR (75 MHz, CD3OD): d=22.8,
28.8, 45.4, 47.7, 80.2, 113.3, 114.1, 114.9, 122.4, 124.7, 125.5, 129.4,
136.8, 158.6, 172.9 ppm; IR (neat): n˜ =3425, 3325, 2980, 2930, 1680,
1640, 1530, 1460, 1365, 1275, 1165, 885, 800 cmÀ1; MS (ESI+): m/z
(%): 418 (100) and 420 (97) [M+Na]+; HRMS (ESI+): m/z calcd for
C17H23BrN3O3 [M+H]+: 396.0917 and 398.08999; found: 396.0916
and 398.0897.
tert-Butyl (2-(1-benzyl-3-hydroxyureido)-2-(5-bromo-1H-indol-3-
yl)ethyl)carbamate (10): Compound 10 was synthesized by the
same procedure as compound 8 with reaction time and tempera-
ture of 24 h at 508C. From indolic amine 9 (100 mg, 0.22 mmol)
and phenyl N-hydroxycarbamate (41 mg, 0.27 mmol), pure com-
pound 10 was obtained as a white solid (22 mg, 20%): 1H NMR
(300 MHz, CDCl3): d=1.44 (s, 9H), 3.45–3.68 (m, 2H), 4.21 (dd, AB
system, J=16.5, dAÀdB =128.3 Hz, 2H), 4.98 (brs, 1H), 5.71 (t, J=
7.2 Hz, 1H), 6.81 (brs, 1H), 6.93 (d, J=5.7 Hz, 2H), 7.10–7.30 (m,
6H), 7.40 (brs, 1H), 7.66 (d, J=2.4 Hz, 1H), 8.66 ppm (brs, 1H);
13C NMR (75 MHz, CDCl3): d=28.5, 42.8, 46.4, 51.2, 80.4, 111.6,
113.0, 113.8, 121.8, 124.3, 126.1, 127.0, 127.8, 128.7, 129.0, 135.0,
137.3, 156.7, 162.5 ppm; MS (ESI+): m/z (%): 525 (100) and 527 (97)
[M+Na]+.
tert-Butyl
(2-benzamido-2-(5-bromo-1H-indol-3-yl)ethyl)carba-
mate (12c): Compound 12c was obtained from indolic amine 11
(60 mg, 0.17 mmol) and benzoyl chloride (20 mL, 0.17 mmol) by the
same procedure as compound 7b. The product was a white solid
1
(45 mg, 58%): mp 1878C; H NMR (300 MHz, [D6]DMSO): d=1.33 (s,
9H), 3.44–3.52 (m, 2H), 5.42–5.49 (m, 1H), 7.00 (t, J=5.8 Hz, 1H),
7.17 (d, J=6.4 Hz, 1H), 7.32 (d, J=8.8 Hz, 1H), 7.38 (s, 1H), 7.44 (t,
J=7.2 Hz, 2H), 7.51 (t, J=7.2 Hz, 1H), 7.82–7.88 (m, 3H), 8.55 (d,
J=8.8 Hz, 1H), 11.14 ppm (brs, 1H); 13C NMR (75 MHz, [D6]DMSO):
d=28.1, 44.1, 46.2, 77.6, 111.1, 113.3, 114.5, 121.0, 123.4, 124.2,
127.3, 128.0, 130.9, 134.61, 134.65, 134.71, 155.8, 165.8 ppm; IR
(neat): n˜ =3375, 3310, 1661, 1629, 1519, 1276, 1163 cmÀ1; MS
(ESI+): m/z (%): 480 (100) and 482 (97) [M+Na]+; HRMS (ESI+): m/z
calcd for C22H25BrN3O3 [M+H]+: 458.1074 and 460.1056; found:
458.1069 and 460.1050.
tert-Butyl
(2-amino-2-(5-bromo-1H-indol-3-yl)ethyl)carbamate
(11): Compound 11 was synthesized in a three-step sequence from
indolic N-hydroxyamine 3cd according to the procedure previously
described.[27] Pure compound 11 was obtained as a white solid
(overall yield: 23%): mp 1518C; 1H NMR (300 MHz, CD3OD): d=
1.41 (s, 9H), 3.24–3.43 (m, 2H), 4.27 (dd, J1 =7.2, J2 =6.0 Hz, 1H),
7.18 (dd, J1 =8.4, J2 =1.8 Hz, 1H), 7.25 (s, 1H), 7.26 (d, J=8.4 Hz,
1H), 7.81 ppm (d, J=1.8 Hz, 1H); 13C NMR (75 MHz, CD3OD): d=
28.8, 49.1, 49.5, 80.2, 113.1, 114.0, 118.0, 122.4, 124.2, 125.3, 129.4,
136.8, 156.7 ppm; IR (neat): n˜ =3380, 3355, 3290, 3120, 3015, 2980,
2895, 1680, 1510, 1450, 1365, 1275, 1245, 1160, 915, 790 cmÀ1; MS
(ESI+): m/z (%): 354 and 356 (12) [M+H]+, 281 (100) and 283 (97);
HRMS (ESI+): m/z calcd for C15H21BrN3O2 [M+H]+: 354.0812 and
356.0793; found: 354.0812 and 356.0793.
tert-Butyl
(2-(5-bromo-1H-indol-3-yl)-2-(2,2,2-trifluoroacetami-
do)ethyl)carbamate (12d): Compound 12d was obtained from in-
dolic amine 11 (80 mg, 0.23 mmol) and trifluoroacetic anhydride
(35 mL, 0.25 mmol) by the same procedure as compound 7b as
a
white solid (70 mg, 69%): mp 1768C; 1H NMR (300 MHz,
[D6]DMSO): d=1.36 (s, 9H), 3.43–3.48 (m, 2H), 5.25–5.32 (m, 1H),
7.03 (t, J=5.7 Hz, 1H), 7.20 (dd, J1 =8.7, J2 =2.0 Hz, 1H), 7.34 (d,
J=8.7 Hz, 1H), 7.38 (d, J=2.0 Hz, 1H), 7.72 (s, 1H), 9.63 (d, J=
8.7 Hz, 1H), 11.24 ppm (brs, 1H); 13C NMR (75 MHz, [D6]DMSO): d=
28.0, 43.4, 46.6, 77.8, 111.4, 112.3, 113.5, 115.9 (JCÀF =288.8 Hz),
120.6, 123.7, 127.7, 155.9 (JCÀF =288.8 Hz), 127.7, 134.7, 155.6 (JCÀF
=
15.8 Hz), 156.0 ppm; IR (neat): n˜ =3440, 3325, 1675, 1535, 1455,
1275, 1160 cmÀ1; MS (ESI+): m/z: 472 and 474 [M+Na]+; HRMS
(ESI+): m/z calcd for C17H20BrFN3O3 [M+H]+: 450.0635 and
452.0616; found: 450.0629 and 452.0610.
tert-Butyl
(2-formamido-2-(5-bromo-1H-indol-3-yl)ethyl)carba-
mate (12a): A solution of BtCHO (85 mg, 0.58 mmol) in THF (2 mL)
was added to a solution of indolic amine 11 (205 mg, 0.58 mmol)
in THF (8 mL) at 08C. The resulting mixture was stirred for 10 min
at 08C and it was then allowed to warm to room temperature.
After 30 min, the THF was removed in vacuo, then CH2Cl2 (50 mL)
and an aqueous solution of NaOH (2n, 20 mL) were added. After
extraction, the organic solution was washed with brine, dried over
anhydrous Na2SO4, filtered, and concentrated in vacuo. Column
chromatography with EtOAc/pentane (10:90 then 99:1) yielded
pure compound 12a as a white solid (140 mg, 63%): 1H NMR
(300 MHz, CD3OD): d=1.41 (s, 9H), 3.41–3.60 (m, 2H), 5.43 (dd, J1 =
8.0, J2 =6.1 Hz, 1H), 7.18–7.29 (m, 3H), 7.79 (d, J=1.8 Hz, 1H),
8.14 ppm (s, 1H); 13C NMR (75 MHz, CD3OD): d=28.8, 45.5, 46.3,
tert-Butyl (2-(5-bromo-1H-indol-3-yl)-2-(nicotinamido)ethyl)car-
bamate (12e): Compound 12e was obtained from indolic amine
11 (80 mg, 0.23 mmol) and nicotinoyl chloride (48 mg, 0.27 mmol)
by the same procedure as compound 7b as a white solid (60 mg,
58%): mp 1958C; 1H NMR (300 MHz, [D6]DMSO): d=1.32 (s, 9H),
3.46–3.50 (m, 2H), 5.43–5.49 (m, 1H), 7.03 (t, J=6.0 Hz, 1H), 7.18
(d, J=8.4 Hz, 1H), 7.32 (d, J=8.4 Hz, 1H), 7.40 (s, 1H), 7.49 (dd,
J1 =8.0, J2 =4.8 Hz, 1H), 7.83 (s, 1H), 8.19 (d, J=8.0 Hz, 1H), 8.68
(d, J=3.6 Hz, 1H), 8.77 (d, J=8.8 Hz, 1H), 9.02 (s, 1H), 11.16 ppm
(s, 1H); 13C NMR (75 MHz, [D6]DMSO): d=29.1, 45.1, 47.3, 78.6,
112.2, 114.4, 115.1, 121.9, 124.2, 124.4, 125.3, 129.0, 131.1, 135.7,
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 13
&
9
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
136.0, 149.5, 152.6, 156.8, 165.4 ppm; IR (neat): n˜ =3325, 3210,
2970, 1680, 1635, 1545, 1270, 1165 cmÀ1; MS (ESI+): m/z: 459 and
461 [M+H]+; HRMS (ESI+): m/z calcd for C21H24BrN4O3 [M+H]+:
459.1026 and 461.1008; found: 459.1024 and 461.1006.
layers were combined, washed with brine, dried over anhydrous
MgSO4, and evaporated. The crude material was purified by flash
chromatography (EtOAc/pentane, 4:6). The solid was washed with
pentane to afford the desired hydroxyamine 15 as a white solid
(685 mg, 40%): 1H NMR (300 MHz, [D6]DMSO): d=1.36 (s, 9H),
1.81–1.92 (m, 1H), 2.22–2.33 (m, 1H), 2.88–2.97 (m, 2H), 3.51 (ABq
system, J=13.8, dAÀdB =17.4 Hz, 2H), 3.99 (dd, J=7.8, J=6.0 Hz,
1H), 6.74 (brs, 1H), 7.15–7.28 (m, 6H), 7.33 (d, J=9.0 Hz, 1H), 7.34
(s, 1H), 7.71 (s, 1H), 7.90 (d, J=1.8 Hz, 1H), 11.16 ppm (s, 1H);
13C NMR (75 MHz, [D6]DMSO): d=28.2, 33.0, 38.0, 60.4, 61.6, 77.2,
110.9, 113.1, 113.2, 122.5, 123.1, 125.8, 126.3, 127.6, 128.7, 128.8,
135.0, 139.5, 155.4 ppm; IR (neat): n˜ =3410, 3290, 2970, 1735, 1685,
1510, 1455 cmÀ1; MS (ESI+): m/z (%): 474 (100) and 476 (97) [M+
H]+; HRMS (ESI+): m/z calcd for C23H29BrN3O3 [M+H]+: 474.1387
and 476.1369; found: 474.1380 and 476.1367.
tert-Butyl (2-(4-bromophenyl)acetamido-2-(5-bromo-1H-indol-3-
yl)ethyl)carbamate (12 f): Compound 12 f was obtained from in-
dolic amine 11 (60 mg, 0.17 mmol) and 2-(4-bromophenyl)acetyl
chloride (204 mL, 0.20 mmol) by the same procedure as compound
7b as a white solid (50 mg, 54%): mp 1318C; 1H NMR (300 MHz,
CD3OD): d=1.34 (s, 9H), 3.29–3.43 (m, 2H), 3.41 (s, 2H), 5.17–5.23
(m, 1H), 6.82 (t, J=6.0 Hz, 1H), 7.17 (dd, J1 =8.4, J2 =1.6 Hz, 1H),
7.21 (d, J=8.4 Hz, 2H), 7.29–7.31 (m, 2H), 7.45 (d, J=8.4 Hz, 2H),
7.66 (s, 1H), 8.30 (d, J=8.8 Hz, 1H), 11.13 ppm (brs, 1H); 13C NMR
(75 MHz, CD3OD): d=29.1, 42.6, 45.1, 46.3, 78.6, 112.1, 114.3, 115.2,
120.4, 122.0, 124.5, 124.8, 128.9, 131.9, 132.1, 135.8, 136.8, 156.6,
170.0 ppm; IR (neat): n˜ =3425, 3340, 1680, 1635, 1535, 1490, 1460,
1275, 1170 cmÀ1; MS (ESI+): m/z: 572, 573, 574, 575, 576, and 577
[M+Na]+; HRMS (ESI+): m/z calcd for C23H25Br2N3O3 [M+H]+:
550.0335, 552.0316, 554.0300; found: 550.0328, 552.0310,
554.0292.
Biology
Bacterial strains: Four S. aureus strains were used.[8c,d,16b,33] The wild-
type clinical isolate S. aureus SA-1199, the NorA-overproducing
mutant SA-1199B, and the fluoroquinolone-resistant strain SA-
K2378, which overexpresses the NorA efflux pump and was pro-
duced by cloning norA and its promoter into plasmid pCU1 and in-
troducing the construct into SA-K1902, were generously provided
by G. W. Kaatz (John D. Dingell Department of Veterans Affairs
Medical Center, Detroit, IL, USA). S. aureus ATCC 25923 was pur-
chased from the Institut Pasteur (CRBIP, Paris, France). All strains
were grown at 378C in Mueller–Hinton (MH) broth (Bio-Rad, Mitry
Mory, France) or spread on MH agar plates for counting. Colony
forming unit (CFU) monitoring was carried out by counting the col-
onies present in 2ꢃ10 mL of serial log dilutions of bacteria inocu-
lum spotted on MH agar plates. Plates were examined for growth
after one night at 378C.
tert-Butyl (3-oxopropyl)carbamate (13): In a dry flask, 3-aminopro-
panol (3.0 g, 40 mmol) was dissolved in dry CH2Cl2 (40 mL) at 08C
under argon. Boc2O (10.5 g, 48 mmol) was then added portionwise.
The solution was stirred at room temperature for 24 h. After addi-
tion of CH2Cl2, the organic layer was washed twice with a saturated
Na2CO3 aqueous solution and water, dried over anhydrous MgSO4,
and evaporated in vacuo. The crude material was purified by filtra-
tion through a short pad of silica (EtOAc/pentane, 4:6) to afford
the desired carbamate as a colorless oil (6.16 g, 88%). A flask was
charged with the carbamate (875 mg, 5.0 mmol), EtOAc (25 mL),
and IBX (3.5 g, 12.5 mmol). The heterogeneous solution was
heated at reflux with stirring for 3 h. The mixture was then filtered
to remove the excess of IBX and the solvent was evaporated. The
desired aldehyde 13 was obtained without any further purification
Media, antibiotics, and culture conditions: MH broth (Bio Rad) was
used for all bacteria overnight cultures, susceptibility testing, and
ethidium bromide accumulation experiments. Ciprofloxacin was
obtained from Sigma–Aldrich (Saint Quentin Fallavier, France). Re-
serpine was purchased from Alfa Aesar (Schiltigheim, France).
Stock solutions (ciprofloxacin and EtBr) were prepared in sterile
water, except for reserpine, which was dissolved in DMSO.
1
as a colorless liquid (820 mg, 95%): H NMR (300 MHz, CDCl3): d=
1.43 (s, 9H), 2.70 (t, J=5.7 Hz, 2H), 3.42 (q, J=5.7 Hz, 2H), 4.97
(brs, 1H), 9.81 ppm (t, J=0.9 Hz, 1H); 13C NMR (75 MHz, CDCl3):
d=28.4, 34.2, 44.4, 79.6, 155.9, 201.4 ppm.
(Z)-N-(3-((tert-butoxycarbonyl)amino)propylidene)-1-phenylme-
thanamine oxide (14): N-Benzylhydroxyamine (585 mg, 4.76 mmol)
and aldehyde 13 (800 mg, 4.62 mmol) were dissolved in CH2Cl2
(15 mL) at room temperature. Anhydrous MgSO4 (2.5 g) was added
to this solution. The mixture was stirred for 3 h at room tempera-
ture. The solution was then filtered and evaporated in vacuo. The
resulting oil was purified by flash chromatography (neat EtOAc,
then EtOAc with 3–6% of MeOH) to afford the desired product 14
Susceptibility testing/MIC determination: For each tested strain,
a bacterial inoculum of 5ꢃ105 CFUmLÀ1 was prepared by diluting
an overnight culture in sterile MH broth. The inoculum was
checked by CFU counting. Indolic compounds (100 mL) were dis-
pensed in a 96-well microplate after twofold serial dilutions in MH
broth by using a Biomek 2000 Workstation (Beckman Coulter, Ville-
pinte, France). The bacterial inoculum (100 mL; prepared as de-
scribed above) was then added into each well. The growth of bac-
terial strains was measured by using a DTX 880 microplate reader
(Beckman Coulter, Villepinte, France) by monitoring absorption at
620 nm after 0, 2, 4, 6, and 24 h of incubation at 378C. All of the
tested compounds were dissolved in DMSO at a concentration of
10 mgmLÀ1 before dilution into MH broth for use in MIC determi-
nation. The MIC value was defined as the minimal concentration of
the indolic derivatives that completely inhibited cell growth during
24 h incubation at 378C. Indolic molecules were first tested at final
concentrations ranging from 0.125–128 mgLÀ1. For compounds
with solubilities lower than 128 mgLÀ1 in MH broth, the concentra-
tion range was adjusted with the maximal concentration decreased
to 8 mgmLÀ1. The highest final DMSO concentration used (2.56%
v/v) induced no significant bacterial growth inhibition. The MIC de-
termination with ciprofloxacin (8 mgLÀ1 for SA-1199B, 2 mgLÀ1 for
1
as a colorless oil (1.20 g, 91%): H NMR (300 MHz, CDCl3): d=1.42
(s, 9H), 2.65 (q, J=6.3 Hz, 2H), 3.31 (q, J=6.3 Hz, 2H), 4.86 (s, 2H),
5.16 (brs, 1H), 6.80 (t, J=6.3 Hz, 1H), 7.34–7.41 ppm (m, 5H);
13C NMR (75 MHz, CDCl3): d=27.9, 28.5, 37.6, 69.4, 128.3, 129.0,
129.3, 132.8, 137.2 ppm; HRMS (ESI+): m/z calcd for C15H23N2O3
[M+H]+: 279.1703; found: 279.1705.
tert-Butyl (3-(benzyl(hydroxy)amino)-3-(5-bromo-1H-indol-3-yl)-
propyl)carbamate (15): In a dry flask, under an argon atmosphere,
freshly distilled acetyl chloride (513 mL, 7.19 mmol) was added to
cold dry MeOH (15 mL). The mixture was stirred for 10 min at 08C,
then a mixture of 5-bromoindole (705 mg, 3.60 mmol) and nitrone
14 (1.0 g, 3.60 mmol) in dry MeOH (15 mL) was added. After the
addition, the mixture was stirred at 08C for 6 h. The solution was
then quenched with saturated NaHCO3 aqueous solution. The
aqueous layer was extracted three times with CH2Cl2. The organic
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 13
&10
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
SA-K2378, and 0.25 mgLÀ1 for SA-1199) was performed as a control
on each microplate. The accepted variance in the MIC values can
be estimated to a twofold difference due to the microdilution
method used. All experiments were at least duplicated in two inde-
pendent runs.
Keywords: antibiotics · efflux pumps · indoles · inhibitors ·
Staphylococcus aureus · structure–activity relationships
´
[1] a) E. K. Jagusztyn-Krynicka, A. Wyszynska, Pol. J. Microbiol. 2008, 57, 91–
Ethidium bromide accumulation assays: The accumulation of ethid-
ium bromide by SA cells with or without inhibitors was determined
by fluorimetry. SA strains were grown overnight in MH broth, dilut-
ed 20-fold in fresh, pre-warmed MH broth, and grown in a water
bath at 378C under agitation (130 rpm) for 3.5 h, which corre-
sponds to the mid-log phase of growth. Samples (10 mL) were
aseptically withdrawn from the culture, the bacteria were harvest-
ed by centrifugation (4000 g, 10 min), the supernatant was discard-
ed, and the bacterial cells were resuspended in fresh sterile pre-
warmed MH broth (20 mL) containing EtBr (final concentration:
10 mgL<M->1) and incubated again in a water bath at 378C under
agitation (130 rpm) for 30 min. Aliquots (1 mL) were taken at fixed
intervals (0, 5, 10, 15, 20, 25, and 30 min) for fluorescence measure-
ment (lex =530 nm, lem =600 nm). After 10 min, the test com-
pounds were added (final concentration: 5 or 20 mgLÀ1) and the
fluorescence was measured for an additional 20 min. The fluores-
cence intensity difference between the compound-containing
assay and the fluorescence baseline before addition was indicative
of the ability of the compound to inhibit the efflux of EtBr. The
result was expressed as a percentage of the fluorescence differ-
ence after and before the addition of the potential efflux pump in-
hibitor. The fluorescence intensity (in arbitrary units) before addi-
tion was calculated as the mean of the fluorescence at 0, 5, and
10 min. Fluorescence intensity after inhibitor addition was calculat-
ed as the mean of the fluorescence at 15, 20, 25, and 30 min. It
was checked that the fluorescence remained constant during the
experiment in the absence of tested compound. Reserpine was
used as a positive control. Experiments were repeated two or three
times and were expressed as mean values ÆSD.
Cellular toxicity tests: In vitro cytotoxicity was assayed on three cell
lines: KB (human mouth carcinoma), MCR5 (human lung fibroblast),
and HCT116 (human colon tumor). Cells were plated in 96-well
tissue-culture microplates in complete medium (200 mL) and treat-
ed 24 h later with stock solution of compounds (2 mL) dissolved in
DMSO by using a Biomek 3000 apparatus (Beckman–Coulter). Con-
trols received the same volume of DMSO (1% final volume). After
72 h of exposure, 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxy-
phenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) reagent (Promega)
was added and incubated for 3 h at 378C. The absorbance was
monitored at 490 nm and the results were expressed as the inhibi-
tion of cell proliferation calculated as the ratio [1À(OD490 treated/
OD490 control)]ꢃ100 in triplicate experiments. Results are present-
ed as the percentage of cellular growth inhibition in the presence
of 10À5 m and 10À6 m concentrations of the tested indolic deriva-
tives.
98; b) L. B. Rice, Curr. Opin. Microbiol. 2009, 12, 476–481.
[2] a) O. Danilchanka, C. Mailaender, M. Niederweis, Antimicrob. Agents Che-
mother. 2008, 52, 2503–2511; b) R. Kock, K. Becker, B. Cookson, J. E. van
Gemert-Pijnen, S. Harbarth, J. Kluytmans, M. Mielke, G. Peters, R. L. Skov,
M. J. Struelens, E. Tacconelli, A. Navarro Tornꢂ, W. Witte, A. W. Friedrich,
Euro Surveill. 2010, 15, 19688; c) A. F. Shorr, N. Haque, C. Taneja, M.
Zervos, L. Lamerato, S. Kothari, S. Zilber, S. Donabedian, M. B. Perri, J.
Spalding, G. Oster, J. Clin. Microbiol. 2010, 48, 3258–3262; d) N. Wood-
ford, D. M. Livermore, J. Infect. 2009, 59, S4–16.
[3] a) H. Boucher, L. G. Miller, R. R. Razonable, Clin. Infect. Dis. 2010, 51,
S183–197; b) M. Z. David, R. S. Daum, Clin. Microbiol. Rev. 2010, 23,
616–687; c) L. D. Liebowitz, Int. J. Antimicrob. Agents 2009, 34, S11–13.
[4] D. L. Paterson, Clin. Infect. Dis. 2004, 38, S341–345.
[5] a) S. J. Dancer, J. Antimicrob. Chemother. 2008, 61, 246–253; b) Y. C. Lin,
K. C. Hsia, Y. C. Chen, W. H. Sheng, S. C. Chang, M. H. Liao, S. Y. Li, Antimi-
crob. Agents Chemother. 2010, 54, 2078–2084; c) C. Pant, T. J. Sferra, A.
Deshpande, A. Minocha, Eur. J. Intern. Med. 2011, 22, 561–568; d) R.
Rossotti, A. Orani, Eur. J. Clin. Microbiol. Infect. Dis. 2012, 31, 1517–1522.
[6] a) A. Fꢄbrega, S. Madurga, E. Giralt, J. Vila, Microb. Biotechnol. 2009, 2,
40–61; b) J. Ruiz, J. Antimicrob. Chemother. 2003, 51, 1109–1117.
[7] a) D. C. Bay, K. L. Rommens, R. J. Turner, Biochim. Biophys. Acta Bio-
membr. 2008, 1778, 1814–1838; b) K. A. Hassan, R. A. Skurray, M. H.
Brown, J. Mol. Microbiol. Biotechnol. 2007, 12, 180–196; c) K. Poole, J.
Antimicrob. Chemother. 2005, 56, 20–51.
[8] a) Y. Ding, Y. Onodera, J. C. Lee, D. C. Hooper, J. Bacteriol. 2008, 190,
7123–7129; b) J. Huang, P. W. O’Toole, W. Shen, H. Amrine-Madsen, X.
Jiang, N. Lobo, L. M. Palmer, L. Voelker, F. Fan, M. N. Gwynn, D. McDevitt,
Antimicrob. Agents Chemother. 2004, 48, 909–917; c) G. W. Kaatz, F.
McAleese, S. M. Seo, Antimicrob. Agents Chemother. 2005, 49, 1857–
1864; d) G. W. Kaatz, S. M. Seo, C. A. Ruble, Antimicrob. Agents Chemo-
ther. 1993, 37, 1086–1094; e) F. McAleese, P. Petersen, A. Ruzin, P. M.
Dunman, E. Murphy, S. J. Projan, P. A. Bradford, Antimicrob. Agents Che-
mother. 2005, 49, 1865–1871; f) H. Nakaminami, N. Noguchi, M. Sasatsu,
Antimicrob. Agents Chemother. 2010, 54, 4107–4111.
[9] a) G. Alvan, C. Edlund, A. Heddini, Drug Resist. Updates 2011, 14, 70–76;
b) J. Fernebro, Drug Resist. Updates 2011, 14, 125–139.
[10] I. M. Gould, J. Antimicrob. Chemother. 2011, 66 (Suppl. 4), iv17–iv21.
[11] B. Marquez, Biochimie 2005, 87, 1137–1147.
[12] B. Zechini, I. Versace, Recent Pat. Anti-Infect. Drug Discovery 2009, 4, 37–
50.
[13] A. Louie, D. L. Brown, W. Liu, R. W. Kulawy, M. R. Deziel, G. L. Drusano,
Antimicrob. Agents Chemother. 2007, 51, 3988–4000.
[14] M. Askoura, W. Mottawea, T. Abujamel, I. Taher, Libyan J. Med. 2011, 6,
5870.
[15] a) J. G. Holler, S. B. Christensen, H. C. Slotved, H. B. Rasmussen, A.
Guzman, C. E. Olsen, B. Petersen, P. Molgaard, J. Antimicrob. Chemother.
2012, 67, 1138–1144; b) M. Stavri, L. J. Piddock, S. Gibbons, J. Antimi-
crob. Chemother. 2007, 59, 1247–1260.
[16] a) N. German, P. Wei, G. W. Kaatz, R. J. Kerns, Eur. J. Med. Chem. 2008, 43,
2453–2463; b) S. Sabatini, F. Gosetto, G. Manfroni, O. Tabarrini, G. W.
Kaatz, D. Patel, V. Cecchetti, J. Med. Chem. 2011, 54, 5722–5736; c) P. L.
Sangwan, J. L. Koul, S. Koul, M. V. Reddy, N. Thota, I. A. Khan, A. Kumar,
N. P. Kalia, G. N. Qazi, Bioorg. Med. Chem. 2008, 16, 9847–9857; d) N.
Thota, M. V. Reddy, A. Kumar, I. A. Khan, P. L. Sangwan, N. P. Kalia, J. L.
Koul, S. Koul, Eur. J. Med. Chem. 2010, 45, 3607–3616.
Acknowledgements
[17] F. J. Schmitz, A. C. Fluit, M. Luckefahr, B. Engler, B. Hofmann, J. Verhoef,
H. P. Heinz, U. Hadding, M. E. Jones, J. Antimicrob. Chemother. 1998, 42,
807–810.
[18] C. Vidaillac, J. Guillon, C. Arpin, I. Forfar-Bares, B. B. Ba, J. Grellet, S.
Moreau, D. H. Caignard, C. Jarry, C. Quentin, Antimicrob. Agents Chemo-
ther. 2007, 51, 831–838.
[19] I. Leitner, J. Nemeth, T. Feurstein, A. Abrahim, P. Matzneller, H. Lagler, T.
Erker, O. Langer, M. Zeitlinger, J. Antimicrob. Chemother. 2011, 66, 834–
839.
This work was supported by a grant from the Agence Nationale
de la Recherche ANR, France (ANR-08EBIO-012). SA-1199, SA-
1199B, and K2378 were kindly provided by Prof. Glenn W. Kaatz
(John D. Dingell Department of Veterans Affairs Medical Center
Detroit, IL, USA). The authors thank Dr. Thierry Cresteil and Dr.
Genevieve Aubert (UPR 2301 CNRS, Imagif, Gif-sur-Yvette) for per-
forming the cytotoxicity tests.
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 13
&
11
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
[20] S. Gibbons, M. Oluwatuyi, G. W. Kaatz, J. Antimicrob. Chemother. 2003,
51, 13–17.
[21] S. Sabatini, G. W. Kaatz, G. M. Rossolini, D. Brandini, A. Fravolini, J. Med.
Chem. 2008, 51, 4321–4330.
[22] a) J. C. A. Tanaka, C. C. da Silva, A. J. B. de Oliveira, C. V. Nakamura, F. B. P.
Dias, Braz. J. Med. Biol. Res. 2006, 39, 387–391; b) L. Mamelli, S. Petit, J.
Chevalier, C. Giglione, A. Lieutaud, T. Meinnel, I. Artaud, J.-M. Pages,
PLoS One 2009, 4, e6443.
[23] P. N. Markham, E. Westhaus, K. Klyachko, M. E. Johnson, A. A. Neyfakh,
Antimicrob. Agents Chemother. 1999, 43, 2404–2408.
[26] O. N. Burchak, E. L. Pihive, L. Maigre, X. Guinchard, P. Bouhours, C. Joli-
valt, D. Schneider, M. Maurin, C. Giglione, T. Meinnel, J. M. Paris, J. N.
Denis, Bioorg. Med. Chem. 2011, 19, 3204–3215.
[27] X. Guinchard, Y. Vallee, J.-N. Denis, Org. Lett. 2007, 9, 3761–3764.
[28] D. R. Hill, C. N. Hsiao, R. Kurukulasuriya, S. J. Wittenberger, Org. Lett.
2002, 4, 111 – 113.
[29] B. Thavonekham, Synthesis 1997, 1189–1194.
[30] X. Guinchard, Y. Vallee, J.-N. Denis, J. Org. Chem. 2007, 72, 3972–3975.
[31] G. W. Kaatz, S. M. Seo, L. O’Brien, M. Wahiduzzaman, T. J. Foster, Antimi-
crob. Agents Chemother. 2000, 44, 1404–1406.
[32] N. P. Kalia, P. Mahajan, R. Mehra, A. Nargotra, J. P. Sharma, S. Koul, I. A.
Khan, J. Antimicrob. Chemother. 2012, 67, 2401–2408.
[33] G. W. Kaatz, S. M. Seo, Antimicrob. Agents Chemother. 1995, 39, 2650–
2655.
[24] a) J. I. Ambrus, M. J. Kelso, J. B. Bremner, A. R. Ball, G. Casadei, K. Lewis,
Bioorg. Med. Chem. Lett. 2008, 18, 4294–4297; b) S. Samosorn, B. Tanwir-
at, N. Muhamad, G. Casadei, D. Tomkiewicz, K. Lewis, A. Suksamrarn, T.
Prammananan, K. C. Gornall, J. L. Beck, J. B. Bremner, Bioorg. Med. Chem.
2009, 17, 3866–3872.
[25] N. R. Guz, F. R. Stermitz, J. B. Johnson, T. D. Beeson, S. Willen, J. Hsiang,
K. Lewis, J. Med. Chem. 2001, 44, 261–268.
Received: January 16, 2014
Published online on && &&, 0000
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 13
&12
&
ÞÞ
These are not the final page numbers!

FULL PAPERS
A pump turn off: 1-(1H-Indol-3-yl)-
ethanamine derivatives such as the one
shown here, were synthesized through
simple chemical modifications and were
shown to be efficient NorA efflux pump
inhibitors. They are able to restore
ciprofloxacin activity against fluoroqui-
nolone-resistant Staphylococcus aureus
strains.
A. Hequet, O. N. Burchak, M. Jeanty,
X. Guinchard, E. Le Pihive, L. Maigre,
P. Bouhours, D. Schneider, M. Maurin,
J.-M. Paris, J.-N. Denis,* C. Jolivalt*
&& – &&
1-(1H-Indol-3-yl)ethanamine
Derivatives as Potent Staphylococcus
aureus NorA Efflux Pump Inhibitors
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 13
&13
&
ÞÞ
These are not the final page numbers!