2-Indolylmethylenebenzofuranones as first effective inhibitors of ABCC2

Article abstract of DOI:10.1016/j.ejmech.2016.06.039

ABC-transporters play a vital role in drugs bioavailability. They prevent intracellular accumulation of toxic compounds, rendering them a major defense mechanism against harmful substances. In this large family, ABCC2 is an apical efflux pump representing about 10% of all membrane proteins in liver and small intestine, and up to 25% in colon. In these tissues, ABCC2 plays a major role in the pharmacokinetics and pharmacodynamics of endo- and xenobiotics. To gain insight in the function of this crucial protein, we have investigated and developed the first effective inhibitors of this pump. Firstly, we set up a cellular flow cytometry assay for monitoring the drug efflux carried out by ABCC2, and used it for the screening of chemical libraries derived from several chemical classes. We found that 2-indolylmethylenebenzofuranone derivatives as promising candidates. Optimization of the hits provided new compounds that inhibit ABCC2 in the micromolar range, making them the first potent ABCC2 inhibitors reported so far. Such compounds would constitute valuable tools to further investigate the role of ABCC2 in the pharmacokinetics and pharmacodynamics of drugs.

Full text of DOI:10.1016/j.ejmech.2016.06.039

European Journal of Medicinal Chemistry 122 (2016) 408e418
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Research paper
2-Indolylmethylenebenzofuranones as first effective inhibitors of
ABCC2
,
b
,
, d,
Elisabeta Baiceanu a 1, Kim-Anh Nguyen c 1, Lucia Gonzalez-Lobato a, Rachad Nasr a,
a
b
e
f
ꢀ ꢁ
Helene Baubichon-Cortay , Felicia Loghin , Marc Le Borgne , Larry Chow ,
Ahcene Boumendjel d, Marine Peuchmaur c 2, Pierre Falson a
c
,
,
d,
, *, 2
ꢁ
a
ꢀ
ꢀ
Drug Resistance Mechanisms and Membrane Proteins Laboratory, BMSSI UMR 5086 CNRS/Universite Lyon 1, Institut de Biologie et Chimie des Proteines,
Lyon, France
b Toxicology Department, Faculty of Pharmacy, Univ. Medicine and Pharmacy€Iuliu Hatieganu€, Cluj-Napoca, Romania
c
ꢀ
ꢀ
Univ. Grenoble Alpes, Departement de Pharmacochimie Moleculaire DPM UMR 5063, 38041 Grenoble, France
d CNRS, DPM UMR 5063, 38041 Grenoble, France
e
ꢀ
ꢀ
ꢀ
ꢀ
Universite de Lyon, Universite Lyon 1, Faculte de Pharmacie e ISPB, EA 4446 Bioactive Molecules and Medicinal Chemistry, SFR Sante Lyon-Est CNRS
UMS3453 e INSERM US7, 8 Avenue Rockefeller, F-69373 Lyon Cedex 8, France
f Department of Applied Biology and Chemical Technology, and State Key Laboratory of Chirosciences, The Hong Kong Polytechnic University, Hung Hom,
Hong Kong Special Administrative Region
a r t i c l e i n f o
a b s t r a c t
Article history:
ABC-transporters play a vital role in drugs bioavailability. They prevent intracellular accumulation of
toxic compounds, rendering them a major defense mechanism against harmful substances. In this large
family, ABCC2 is an apical efflux pump representing about 10% of all membrane proteins in liver and
small intestine, and up to 25% in colon. In these tissues, ABCC2 plays a major role in the pharmacokinetics
and pharmacodynamics of endo- and xenobiotics. To gain insight in the function of this crucial protein,
we have investigated and developed the first effective inhibitors of this pump. Firstly, we set up a cellular
flow cytometry assay for monitoring the drug efflux carried out by ABCC2, and used it for the screening of
Received 22 March 2016
Received in revised form
18 June 2016
Accepted 20 June 2016
Available online 27 June 2016
Keywords:
chemical
libraries
derived
from
several
chemical
classes.
We
found
that
2-
ABC transporters
ABCC2 inhibitors
Indolylmethylenebenzofuranone
Aurones
indolylmethylenebenzofuranone derivatives as promising candidates. Optimization of the hits pro-
vided new compounds that inhibit ABCC2 in the micromolar range, making them the first potent ABCC2
inhibitors reported so far. Such compounds would constitute valuable tools to further investigate the role
of ABCC2 in the pharmacokinetics and pharmacodynamics of drugs.
Drug interactions
© 2016 Elsevier Masson SAS. All rights reserved.
1. Introduction
major pharmacological barriers [3] and exercise an important in-
fluence on the absorption, distribution, and/or elimination of drugs
ABC (ATP-binding cassette) transporters are responsible for the
ATP-dependent movement of wide variety of xenobiotics,
including drugs, lipids and metabolic products across the plasma
and intracellular membranes [1,2]. They are highly expressed in
and other xenobiotics [4]. A well-defined role in the transport of
clinically relevant drugs was especially described for the ABCB1 (P-
gp), ABCC1-5 (MRP 1e5) and ABCG2 (BCRP) [5]. Their contribution
to multidrug resistance (MDR) in tumor cells is well documented
[6], making them privileged targets to tackle MDR. These efflux
pumps are also involved in drug-drug interactions. Numerous ex-
amples of co-administration of an ABCB1 inhibitor with an ABCB1
substrate showed that there is a risk of increased blood level of the
latter and serious side effects. Such drug-drug interactions are
known for verapamil (dronedarone, quinidine, ranolazine), loper-
amide (tipranavir, ritonavir), saquinavir (tipranavir, ritonavir) and
interactions with elacridar (GF120918) [7].
a
Abbreviations: ABCC2, second member of the C subfamily of the ABC pumps;
DDI, drug-drug interactions; P-gp, P-glycoprotein; MRP, multidrug resistance pro-
tein; BCRP, breast cancer resistance protein; MDR, multidrug resistance; CCR2, CC
chemokine receptor 2; CCL2, CC chemokine ligand 2; CCR5, CC chemokine receptor
5.
* Corresponding author.
E-mail address: pierre.falson@ibcp.fr (P. Falson).
Both junior investigators equally contributed to the work.
Both senior investigators equally contributed to the work.
1
Initially identified as the first canalicular multispecific organic
2
http://dx.doi.org/10.1016/j.ejmech.2016.06.039
0223-5234/© 2016 Elsevier Masson SAS. All rights reserved.

E. Baiceanu et al. / European Journal of Medicinal Chemistry 122 (2016) 408e418
409
anion transporter, cMOAT, ABCC2 is the second member of the ABC
2.2. Biological evaluation
C-subfamily. It plays an important role in the export of organic
anions, unconjugated bile acids and xenobiotics into the bile.
Localized on the apical side of polarized cells, ABCC2 is considered
the ideal efflux pump in the terminal phase of detoxification [8].
Mutations of the ABCC2 gene are associated with DubineJohnson
syndrome, a condition in which the lack of hepatobiliary transport
of non-bile salt organic anions results in conjugated hyper-
bilirubinaemia [8,9]. Recent studies showed that ABCC2 is one of
the ABC pumps with the highest expression level in organs that are
important for the metabolism of endo- and xenobiotics, such as the
liver, kidneys and intestine. The high abundance of ABCC2 has been
quantified using quantitative proteomics and led to an estimation
of about 25% of membrane proteins in colon and10% in small in-
testine and liver [10].
The high ABCC2 expression suggests that it plays a major role in
drug pharmacokinetics and pharmacodynamics that is essential to
be investigated. This could be achieved by using efficient inhibitors
as preconized by FDA (Food and Drug Administration) and EMA
(European Medicines Agency) [11,12]. Up to date, only few ABCC2
modulators, poorly efficient and non-specific are reported [8]. The
most efficient ABCC2 inhibitor is cyclosporine A with an IC50 of
2.2.1. ABCC2-mediated calcein efflux inhibition
We set up a high-throughput flow-cytometry method to eval-
uate over 500 compounds of various chemical families covering
diverse classes of compounds, some of which were previously
shown to be inhibitors of ABC proteins (ABCB1, ABCG2, and ABCC1).
The activity of ABCC2 in cellulo was monitored by using the ABCC2-
transfected MDCKII cells (Madin-Darby Canine Kidney II cells),
relative to MDCKII control cells [17] and selected calcein as sub-
strate of the pump as previously reported. The evaluation was
conducted in the presence of 1 mM GF 120918 and 25 mM MK-571 as
inhibitors of ABCB1 and ABCC1 respectively, since these ABC pumps
are also expressed in MDCKII cells and transport calcein [18,19].
Used at 25 nM, calcein accumulation in ABCC2-MDCKII cells was
not modified in the presence of GF120918 and MK-571, showing
that the difference in calcein accumulation with parental MDCKII
cells is only due to the ABCC2 efflux activity [17].
2.2.2. Libraries screen
We tested over 500 compounds belonging to more than 20
families including flavonoid [20], flavones, flavonols, flavanones,
flavanols, chromones [21], aurones, chalcones [22], xanthones,
flavonoid dimers, quinolones, alkaloids, acridones, thiourea de-
rivatives, diphenyloxazolines, biphenyl imines, steroids and pep-
tide analogues [23], indenoindoles, benzofurans, phthalimides,
carbazoles, indazoles, carboxylic acids carbamazones. General
classes of compounds included in this study are given in Fig. 2.
These classes of compounds were chosen mostly because most
of the chemical scaffolds were previously shown to interact with
20 mM, ut it is non selective as it also inhibits ABCB1 and ABCC1
[13]. Therefore, it is mandatory to explore new ways to discover
better inhibitors.
Hence, we firstly set up a cellular assay using cell lines over-
expressing ABCC2 and flow-cytometry in order to screen chemo-
libraries. After the evaluation of a library of more than 500
compounds belonging to diverse chemical libraries, we found 2-
indolylmethylenebenzofuran-3(2H)-ones among the most prom-
ising scaffolds (Fig. 1). In the present work, we report the cellular
test and the optimization of the hits. The ultimate goal is to set up
the structural bases needed for the development of active and se-
lective inhibitors of ABCC2, needed to explore the pathological and
clinical role of this protein.
ABC proteins. Used at 5
mM, the compounds were compared to the
reference inhibitor, cyclosporine A used at 25
mM, which corre-
sponds to its IC50. From this large screen, summarized in Fig. 3, it
turned out that compounds from the indolylmethylenebenzofur-
anone family (Fig. 1) displayed the highest inhibition of calcein
ABCC2-mediated efflux (orange bars). The structurally close
aurones and azaurones did not display any efficacy suggesting a
critical role of the indole moiety. Therefore, further investigations
were solely focused on indolylbenzofuranone class.
2. Results
2.1. Chemistry
2.2.3. Inhibition efficacy of indolylmethylenebenzofuranone
derivatives
The synthesis of 2-indolylmethylenebenzofuran-3(2H)-ones
was performed using a previously reported method [14] [15]
involving an aldol condensation between indolecarboxaldehydes
2a-s and 4,6-dihydroxybenzofuran-3(2H)-one 1 under basic con-
ditions (Scheme 1) [16].
The 4,6-dihydroxybenzofuran-3(2H)-one 1 was first synthe-
sized with 84% yield in two steps, starting from phloroglucinol and
chloroacetonitrile in HCl/Et2O (Scheme 1). The indolecarbox-
aldehydes 2aes were obtained by the mean of an alkylation in the
presence of sodium hydride 60% and alkyl halide (44e93%). The
synthesis and characterization of compounds 3a, b, del and n were
already described by our group in a precedent study [16].
The primary screen showing that indolylmethylenebenzofur-
anones were the most promising compounds prompted us to
synthesize a series of analogs and to study their influence on the
accumulation of calcein in the MDCKII-ABCC2 cells. The structures
and measured IC50 of the compounds set are shown in Table 1.
As shown in Table 1, compounds tested at 5
efficient inhibitors of ABCC2-mediated calcein efflux than cyclo-
sporine A used at 25 M. As shown through the IC50 values, the
compounds efficacies ranged from 4 M for the most active com-
pounds (3i, 3j, 3m and 3p) to 20 M for the less active ones. The SAR
analysis revealed that the substitution at the indolic nitrogen may
mM were much more
m
m
m
Fig. 1. Structures of aurones and 2-indolylmethylenebenzofuran-3(2H)-ones studied as inhibitors of ABCC2.

410
E. Baiceanu et al. / European Journal of Medicinal Chemistry 122 (2016) 408e418
Scheme 1. General pathway for the synthesis of 2-indolylmethylenebenzofuran-3(2H)-ones. Reagents and conditions: (i) ClCH2CN, HCl, ZnCl2, Et2O, 0 ꢀC; (ii) H2O, 100 ꢀC, 5 h, 84%
(for 2 steps); (iii) RX (X ¼ Br or I), NaH, DMF, rt, 18e72 h, 44e93%; (iv) KOH 50% in H2O, EtOH, reflux, 2e15 h (see detailed structures in Table 1).
tolerate a panel of substituents with equivalent activity (3d, 3e, 3f)
but some adequacy of size and/or hydrophobicity should be
respected (3f vs 3g and 3h). However, the substitution by alkyl
chains is more favorable than with aryl groups. Also, we observe
that the linkage of the indole to the benzofuranone through its C-3
is slightly more advantageous than C-2. A marked impact of the
substitution pattern of the phenyl ring of indole was observed: a
substitution at C-6 is more advantageous over C-8 (3n vs 3l). It is
noteworthy that the influence of the substitution at the phenyl ring
of indole is dependent on the substituents at the N-1 (3p vs 3q-s).
The latter is confirmed by comparing the most active compound of
the series 3m to 3l.
precise role remains unclear [38]. This is partly because only few
validated cells expressing a functional form of the pump [39] are
available, and to the lack of potent and specific ABCC2 inhibitors
[8,40]. In order to achieve the latter goal, we set up a cell-based
model using the MDCKII cells to evaluate ABCC2-mediated trans-
port and adapted it for high-throughput flow-cytometry screening,
then screened a chemolibrary of more than 500 compounds that
regroups various classes of candidates and some of which proved to
be modulators for other ABC transporters, such as ABCB1, ABCC1
and ABCG2 [21,41,42].
Our initial screen of different chemolibraries allowed us to
identify few indolylmethylenebenzofuranones as potential in-
hibitors of ABCC2. These compounds are considered as structural
In order to ensure the effectiveness of our best inhibitors, we
checked them in another cell system, using Flp-InTM-293 cells
transfected ABCC2 gene, which confirmed the above results (Fig. 4).
analogs
of
the
naturally
occurring
aurones
(Z)-2-
benzylidenebenzofuran-3-(2H)-ones. Aurones are flavone isomers
found in vegetables and especially in fruits and flowers, where they
contribute to the bright yellow color. Probably due to their scarce
occurrence in nature, the medicinal chemistry of aurones is quite
recent but in constant progress [43e45].
2.2.4. Selectivity for ABCC2 versus ABCB1 and ABCC1
In the case of ABC transporters, it is not trivial to address the
specificity issues, since this family of proteins is characterized by
substrates and inhibitors overlapping. This is particularly exacer-
bated in the case of ABCC2 for which the best reported inhibitor so
far, cyclosporine A, is not selective and fully inhibits much more
efficiently ABCB1- and ABCC1-mediated calcein efflux, at lower
Regarding selectivity for ABCC2 inhibitors, three inhibitors and
particularly 3p, displayed a higher inhibition selectivity towards
ABCC2 than ABCC1 and ABCB1, showing that ABCC2 can be spe-
cifically targeted although the substrate and inhibitor overlaps with
the two other pumps. However, this does not exclude a reactivity of
these compounds towards other transporters not tested in the
present study. The selectivity is not yet complete but this result is
encouraging for the development of a next generation of com-
pounds. The low toxicity of the identified inhibitors is also
encouraging and stimulates further investigations in order to
enhance efficacy and selectivity. The robust cellular assay designed
within this work and the easy access to indolylmethylenebenzo-
furanones are precious elements to pursue the investigation of
ABCC2 inhibitors.
As chemical tools, these inhibitors can help to gain insight into
the role of this highly expressed transporter in liver, kidneys and
intestine. They can be used to investigate drug pharmacokinetics
and pharmacodynamics for which the role of ABCC2 transport is
documented. Used in this context, to avoid clinical complications,
the purpose of the inhibitors is not to block long-term the protein,
but rather in controlled rounds.
concentrations (2.5 mM and 25 mM, respectively) (Fig. 5).
We tested our compounds series on both transporters. As shown
in Fig. 5, among the pool of the most efficient ABCC2 inhibitors, 3m,
3i, 3p and 3j, displayed a higher selectivity towards calcein ABCC2-
mediated efflux, with 3p being the more selective. Three other
compounds, 3c, 3b and 3r, although less efficient towards ABCC2
than 3p, displayed a higher selectivity towards ABCC2 versus
ABCB1 and ABCC1. At this point, it is however not possible to draw a
clear and rational structure-activity relationship to address the
selectivity issues. Such RSA could be performed in a further study
by analyzing large series and diversely substituted compounds,
especially those bearing concomitant substituents at the indolic
nitrogen and at the indole phenyl ring.
2.2.5. Innocuity of ABCC2 inhibitors on cell survival
The cellular cytotoxicity of the most potent ABCC2 inhibitors
was evaluated using the MTT assay. As shown in Fig. 6, compounds,
3m, 3i, 3p and 3j were not at all -or poorly-cytotoxic for MDCKII WT
and MDCKII ABCC2 cells, displaying IC50 > 50 mM. Particularly, the
most promising compound, 3p, displayed no visible cytotoxicity
While ABCC2 is highly expressed in organs playing key roles in
drug metabolism, surprisingly little is known about this protein and
its impact on drug pharmacokinetics, pharmacodynamics. Being
highly expressed in liver and knowing that more than 700 drugs are
associated with drug-induced liver injury [46], our ABCC2 in-
hibitors can serve as tools to build more convenient and practical
approaches to predict drug-induced liver injury and drug-drug
interactions due to the implication of this pump.
over the range of tested concentrations.
3. Discussion and conclusion
Despite the high abundance of ABCC2 in apical membranes, its

E. Baiceanu et al. / European Journal of Medicinal Chemistry 122 (2016) 408e418
Table 1 (continued )
411
Table 1
ABCC2-mediated calcein transport inhibition efficacy (IC50) of indolylmethylene-
benzofuranone for the ABCC2 inhibitors in MDCKII cells.
Entry
11
Compounda
Indol moiety
IC50 (mM)
3k
11.4 0.7
12
13
3l
13.9 1.6
Entry
1
Compounda
Indol moiety
IC50
(mM)
3a
11.1 0.7
3m
4.1 0.8
2
3
3b
3c
9
2.2
8.2 0.9
14
15
3n
3o
6.3 0.2
4
5
3d
10.5 1.1
7.3 0.7
3e
10.6 1.2
16
17
3p
3q
4.5 0.58
6
7
3f
12.2 1.2
17.8 1.8
3g
14.4 1.4
18
19
3r
3s
15.4
2
8
3h
17.5 1.8
15.8 1.6
9
3i
3j
4.5 0.9
20
Cyclosporine A
25.6 1.4
b
IC50 (mean) ¼ mean value of compound concentration needed to inhibit 50% of
calcein transport. Data are the mean SD of 3 independent experiments.
a
Dashed line indicates the site of linkage to the benzofuranone moiety.
10
4.8 0.6

412
E. Baiceanu et al. / European Journal of Medicinal Chemistry 122 (2016) 408e418
Fig. 2. Main classes of compounds tested in the study. The graphic representation includes the general structure, the name of the class and the number of compounds screened. We
included literature references for molecules that were previously published [24e37].
4. Experimental section
[1H:
d6)
d
(DMSO-d6) ¼ 2.50 ppm,
2.05 ppm; 13C:
(acetone-d6) ¼ 29.84 ppm]. Electrospray
d
(CDCl3) ¼ 7.24 ppm,
d(acetone-
(DMSO-d6) 39.51 ppm,
¼
d
¼
4.1. Chemistry
d(CDCl3) ¼ 77.23 ppm,
d
ionization ESI mass spectra were acquired by the Analytical
Department of Grenoble University on an Esquire 3000 Plus Bruker
Daltonis instrument with a nanospray inlet. Accurate mass mea-
surements (HRMS) were carried out on a ESI/QTOF with the Waters
Xevo G2-S QTof device. The purity of compounds was determined
by HPLC (Agilent 1100 Series HPLC): all tested compounds have a
purity ꢁ 95%.
Commercially available reagents and solvents were used
without further purification. Reactions were monitored by thin-
layer chromatography (plates coated with silica gel 60 F254 from
Merck). Products were purified with column chromatography on
Silica gel 60 (230e400 mesh from Macherey-Nagel) or by auto-
matic flash chromatography with the Grace device: Reveleris X2.
Melting points were measured on a Büchi B540 apparatus and are
uncorrected. 1H and 13C NMR spectra were recorded at room
temperature in deuterated solvents on a Brüker AC-400 instrument
4.1.1. General procedure A for the synthesis of N-alkyl-indole-3-
carboxaldehyde (3c, m, oꢂs)
(400 MHz). Chemical shifts (
d) are reported in parts per million
To a suspension of NaH 60% in oil (2.25 equiv.) in dry DMF
(0.8 mL/mmol) was added, at 0 ꢀC and under nitrogen atmosphere,
(ppm) relative to TMS as internal standard or relative to the solvent

E. Baiceanu et al. / European Journal of Medicinal Chemistry 122 (2016) 408e418
413
Fig. 3. Initial screening of ABCC2-mediated calcein transport inhibitors. MDCKII WT and ABCC2 cells were incubated with 5
m
M compounds. Inhibition values were calculated as
described in materials and methods. Negative values are due to normal accumulation differences in the ABCC2-transfected cells (less calcein accumulated after treatment with
compounds).
Fig. 4. Calcein transport inhibition (%) by 5 mM compounds of ABCC2 expressed in Flp-InTM-293 (white bars) and MDCKII (grey bars) cells.
Fig. 5. Calcein transport inhibition of the human MDR ABC transporters, ABCB1 and ABCC1, by compounds of the study. Compounds were tested at 5
transporter and the corresponding control: ABCC2 MDCKII WT and MDCKII ABCC2 (black bars), NIH3T3 WT and NIH3T3 ABCB1 (white bars), BHK21 WT and BHK21 ABCC1 (grey
bars). Cyclosporine A (cyc) was used as reference compound, at 25 M for ABCC2 and ABCC1 and 2.5 M for ABCB1 (*** e p < 0.001; ** e p < 0.01; * e p < 0.05).
mM on cells expressing each
m
m
a solution of indolecarboxaldehyde (1 equiv.) in dry DMF (2.5 mL/
mmol). After stirring for 30 min at rt, alkyl halide (1.0e3.0 equiv.)
was slowly added. After stirring overnight, the reaction was
quenched by addition of water and the product was extracted with
diethyl ether. The organic layer was dried over MgSO4, filtered off
and concentrated under vacuum. The crude product was purified
by column chromatography on silica gel.
from 3-methylindole-2-carboxaldehyde (199 mg, 1.25 mmol) and
bromobutane (377 mg, 2.75 mmol). After purification by column
chromatography on silica gel (cyclohexane/EtOAc 98:2 to 96:4), the
pure product (120 mg, 0.56 mmol, 44%) was obtained as a yellow
oil. Rf ¼ 0.17 (95:5 cyclohexane/EtOAc); 1H NMR (400 MHz, CDCl3)
d
ppm 10.14 (s, 1H), 7.68 (dd, J ¼ 8.0, 1.0 Hz, 1H), 7.39 (ddd, J ¼ 8.5,
6.8, 1.0 Hz, 1H), 7.33 (dd, J ¼ 8.5, 1.1 Hz, 1H), 7.13 (ddd, J ¼ 8.0, 6.8,
1.1 Hz, 1H), 4.50 (t, J ¼ 7.4 Hz), 2.62 (s, 3H), 1.67e1.77 (m, 2H),
1.29e1.40 (m, 2H), 0.93 (t, J ¼ 7.4 Hz, 3H); 13C NMR (100 MHz,
4.1.1.1. N-Butyl-3-methylindole-2-carboxaldehyde (2c). The crude
product was prepared according to general procedure A starting
CDCl3) d ppm 181.3 (CH), 139.3 (C), 130.9 (C), 127.3 (CH), 127.1 (C),

414
E. Baiceanu et al. / European Journal of Medicinal Chemistry 122 (2016) 408e418
Fig. 6. Safety of compounds towards MDCKII WT and MDCKII ABCC2 cells (black symbols), added to WT (circles) or ABCC2 (squares) cells. Negative controls were carried out in
parallel by replacing compounds by DMSO (open symbols) at the corresponding compound concentrations.
126.8 (C), 121.5 (CH), 120.1 (CH), 110.6 (CH), 44.6 (CH2), 32.9 (CH3),
20.3 (CH2), 14.0 (CH2), 8.7 (CH3); LRMS (ESIþ) m/z (%) 216 (100)
[MþH]þ; HRMS (ESIþ) m/z calc.for C14H18NO 216.1388 [MþH]þ,
found 216.1388.
1H), 7.38e7.42 (m, 1H), 7.28 (dd, J ¼ 1.6, 1.6 Hz, 1H), 7.15 (dd, J ¼ 7.8,
7.8 Hz, 1H), 7.11 (d, J ¼ 8.9 Hz, 1H), 6.99e7.03 (m, 1H), 6.89 (dd,
J ¼ 8.9, 2.5 Hz, 1H), 5.22 (s, 2H), 3.84 (s, 3H); 13C NMR (100 MHz,
CDCl3) d ppm 184.7 (CH), 156.9 (CH), 138.7 (C), 137.9 (CH), 132.2 (C),
131.6 (CH), 130.8 (CH), 130.2 (CH), 126.4 (C), 125.7 (CH), 123.3 (C),
118.5 (C), 114.9 (C), 111.3 (CH), 103.6 (CH), 55.9 (CH2), 50.5 (CH3);
LRMS (ESIþ) m/z (%) 346 (100) [M(81Br)þH]þ, 344 (100) [M(79Br)þ
H]þ; HRMS (ESIþ) m/z calc. for C17H15BrNO2 344.0286 [MþH]þ,
found 344.0278.
4.1.1.2. N-(Cyclohexylmethyl)-5-methoxyindole-3-carboxaldehyde
(2m). The crude product was prepared according to general pro-
cedure
A
starting from 5-methoxyindole-3-carboxaldehyde
(240 mg, 1.37 mmol) and bromomethylcyclohexane (487 mg,
2.75 mmol). After purification by column chromatography on silica
gel (cyclohexane/EtOAc 7:3), the pure product (344 mg, 1.27 mmol,
93%) was obtained as a white solid. Rf ¼ 0.39 (7:3 cyclohexane/
4.1.1.4. N-Methyl-5-bromoindole-3-carboxaldehyde (2p). The crude
product was prepared according to general procedure A starting
from 5-bromoindole-3-carboxaldehyde (100 mg, 0.45 mmol) and
iodomethane (127 mg, 0.89 mmol). After purification by column
chromatography on silica gel (cyclohexane/EtOAc 6:4 to 5:5), the
pure product (86 mg, 0.36 mmol, 81%) was obtained as a yellowish
solid. Rf ¼ 0.31 (5:5 cyclohexane/EtOAc); m.p. 131e134 ꢀC; 1H NMR
EtOAc); m.p. 69e71 ꢀC; 1H NMR (400 MHz, CDCl3)
d ppm 9.92 (s,
1H), 7.77 (d, J ¼ 2.5 Hz, 1H), 7.58 (s, 1H), 7.22 (d, J ¼ 8.9 Hz, 1H), 6.93
(dd, J ¼ 8.9, 2.5 Hz, 1H), 3.92 (d, J ¼ 7.2 Hz, 2H), 3.87 (s, 3H),
1.78e1.90 (m, 1H), 1.56e1.76 (m, 5H), 1.10e1.24 (m, 3H), 0.92e1.04
(m, 2H); 13C NMR (100 MHz, CDCl3)
d ppm 184.6 (CH), 156.8 (C),
139.2 (CH), 132.6 (C), 126.3 (C), 117.8 (C), 114.6 (CH), 111.4 (CH), 103.5
(CH), 56.0 (CH3), 54.2 (CH2), 38.6 (CH), 31.1 (2xCH2), 26.3 (CH2), 25.8
(CH2); LRMS (ESIþ) m/z (%) 294 (20) [MþNa]þ, 272 (100) [MþH]þ;
HRMS (ESIþ) m/z calc. for C17H22NO2 272.1651 [MþH]þ, found
272.1654.
(400 MHz, CDCl3)
1H), 7.40 (dd, J ¼ 8.7, 1.9 Hz, 1H), 7.18 (d, J ¼ 8.7 Hz, 1H), 3.83 (s, 3H);
13C NMR (100 MHz, CDCl3)
ppm 184.3 (CH), 139.8 (CH), 136.7 (C),
d
ppm 9.91 (s, 1H), 8.42 (d, J ¼ 1.8 Hz, 1H), 7.62 (s,
d
127.2 (C), 126.9 (CH), 124.9 (CH), 117.7 (C), 116.9 (C), 111.5 (CH), 34.1
(CH3); LRMS (ESIþ) m/z (%) 240 (100) [M(81Br)þH]þ, 238 [M(79Br)þ
H]þ (95); HRMS (ESIþ) m/z calc. for C10H9BrNO 237.9868 [MþH]þ,
found 237.9862.
4.1.1.3. N-(3-Bromobenzyl)-5-methoxyindole-3-carboxaldehyde (2o).
The crude product was prepared according to general procedure A
starting from 5-methoxyindole-3-carboxaldehyde (100 mg,
0.57 mmol) and 3-bromobenzyl bromide (285 mg,1.14 mmol). After
purification by column chromatography on silica gel (cyclohexane/
EtOAc 7:3), the pure product (133 mg, 0.39 mmol, 68%) was ob-
tained as a yellow oil. Rf ¼ 0.32 (7:3 cyclohexane/EtOAc); 1H NMR
4.1.1.5. N-Ethyl-5-bromoindole-3-carboxaldehyde (2q). The crude
product was prepared according to general procedure A starting
from 5-bromoindole-3-carboxaldehyde (100 mg, 0.45 mmol) and
bromoethane (97 mg, 0.89 mmol). After purification by column
chromatography on silica gel (cyclohexane/EtOAc 7:3), the pure
product (96 mg, 0.38 mmol, 85%) was obtained as a yellowish
(400 MHz, CDCl3)
d
ppm 9.91 (s, 1H), 7.78 (d, J ¼ 2.5 Hz, 1H), 7.61 (s,

E. Baiceanu et al. / European Journal of Medicinal Chemistry 122 (2016) 408e418
415
powder. Rf ¼ 0.13 (7:3 cyclohexane/EtOAc); m.p. 113e115 ꢀC; 1H
8:2 to 5:5) and recrystallization in acetonitrile, the pure product
(61 mg, 0.17 mmol, 47%) was obtained as red powder. Rf ¼ 0.18
(96:4 CH2Cl2/MeOH); m.p. 197e199 ꢀC; 1H NMR (400 MHz,
NMR (400 MHz, CDCl3)
d
ppm 9.84 (s,1H, CH]O), 8.37 (d, J ¼ 1.9 Hz,
1H, H4), 7.64 (s, 1H, H2), 7.32 (dd, J ¼ 8.7, 2.0 Hz, 1H, H6), 7.15 (d,
J ¼ 8.7 Hz, 1H, H7), 4.13 (q, J ¼ 7.3 Hz, 2H, CH2), 1.49 (t, J ¼ 7.3 Hz, 3H,
Acetone-d6)
d
ppm 9.88 (bs, 1H), 9.26 (bs, 1H), 7.60 (dd, J ¼ 8.0,
CH3); 13C NMR (100 MHz, CDCl3)
d
ppm 184.3 (CH]O), 138.2 (C2),
0.7 Hz, 1H), 7.44 (dd, J ¼ 8.3, 1.0 Hz, 1H), 7.23 (ddd, J ¼ 8.0, 7.1, 1.0 Hz,
1H), 6.83 (s, 1H), 7.07 (ddd, J ¼ 8.3, 7.1, 0.7 Hz,1H), 6.30 (d, J ¼ 1.5 Hz,
1H), 6.17 (d, J ¼ 1.5 Hz,1H), 4.31 (t, J ¼ 7.3 Hz), 2.42 (s, 3H), 1.69e1.78
(m, 2H), 1.26e1.38 (m, 2H), 0.90 (t, J ¼ 7.4 Hz, 3H); 13C NMR
135.7 (C8), 126.9 (C5), 126.9 (C6), 124.7 (C4), 117.5 (C3), 116.5 (C9),
111.6 (C7), 42.2 (CH2), 15.0 (CH3); LRMS (ESIþ) m/z (%) 254 (95)
[M(81Br)þH]þ, 252 (100) [M(79Br)þH]þ; HRMS (ESIþ) m/z calc. for
C
11H11BrNO 252.0024 [MþH]þ, found 252.0018.
(100 MHz, Acetone-d6) d ppm 180.7 (C), 168.6 (C), 168.4 (C), 159.0
(C), 148.4 (C), 139.0 (C), 129.5 (C), 129.4 (C), 124.1 (CH), 120.1 (CH),
120.1 (CH), 115.5 (C), 110.6 (CH), 104.4 (C), 100.1 (CH), 98.7 (CH), 92.1
(CH), 44.4 (CH2), 33.3 (CH3), 20.8 (CH2), 14.1 (CH2), 10.7 (CH3); LRMS
(ESIþ) m/z (%) 364 (100) [MþH]þ; HRMS (ESIþ) m/z calc. for
4.1.1.6. N-Butyl-5-bromoindole-3-carboxaldehyde (2r). The crude
product was prepared according to general procedure A starting
from 5-bromoindole-3-carboxaldehyde (307 mg, 1.37 mmol) and
bromobutane (377 mg, 2.75 mmol). After purification by column
chromatography on silica gel (cyclohexane/EtOAc 7:3), the pure
product (358 mg, 1.28 mmol, 93%) was obtained as a yellow oil.
Rf ¼ 0.21 (7:3 cyclohexane/EtOAc); 1H NMR (400 MHz, CDCl3)
C
22H22NO4 364.1549 [MþH]þ, found 364.1537.
4.1.2.2. (Z)-2-[N-(Cyclohexylmethyl)-5-methoxyindol-3-
ylmethylene]-4,6-dihydroxybenzofuran-3(2H)-one (3m). The crude
product was prepared according to general procedure B starting
from 1 (60 mg, 0.36 mmol) and aldehyde 2m (125 mg, 0.46 mmol).
After purification by column chromatography on silica gel (cyclo-
hexane/EtOAc 8:2 to 4:6) and recrystallization in acetonitrile, the
pure product (36 mg, 0.09 mmol, 24%) was obtained as red powder.
d
ppm 9.84 (s, 1H), 8.37 (d, J ¼ 1.9 Hz, 1H), 7.61 (s, 1H), 7.31 (dd,
J ¼ 8.7, 1.9 Hz, 1H), 7.15 (d, J ¼ 8.7 Hz, 1H), 4.07 (t, J ¼ 7.2 Hz, 2H),
1.74e1.84 (m, 2H), 1.21e1.38 (m, 2H), 0.90 (t, J ¼ 7.4 Hz, 3H); 13C
NMR (100 MHz, CDCl3)
d ppm 184.3 (CH), 139.0 (CH), 135.9 (C),
126.9 (C), 126.9 (CH), 124.7 (CH), 117.4 (C), 116.4 (C), 111.7 (CH), 47.3
(CH2), 31.8 (CH2), 20.1 (CH2), 13.7 (CH3); LRMS (ESIþ) m/z (%) 304
(10), 302 (10), 282 (100) [M(81Br)þH]þ, 280 (95) [M(79Br)þH]þ;
HRMS (ESIþ) m/z calc. for C13H15BrNO 280.0337 [MþH]þ, found
280.0338.
m.p. 188e191 ꢀC; 1H NMR (400 MHz, Acetone-d6)
d ppm 9.70 (bs,
1H), 8.88 (bs, 1H), 8.07 (s, 1H), 7.59 (d, J ¼ 2.4 Hz, 1H), 7.45 (d,
J ¼ 8.9 Hz, 1H), 7.09 (s, 1H), 6.90 (dd, J ¼ 8.9, 2.4 Hz, 1H), 6.34 (d,
J ¼ 1.7 Hz, 1H), 6.12 (d, J ¼ 1.7 Hz, 1H), 4.14 (d, J ¼ 7.3 Hz, 2H), 3.92 (s,
3H), 1.87e2.02 (m, 1H), 1.58e1.76 (m, 5H), 1.16e1.26 (m, 3H),
4.1.1.7. N-(Cyclohexylmethyl)-5-bromoindole-3-carboxaldehyde (2s).
The crude product was prepared according to general procedure A
starting from 5-bromoindole-3-carboxaldehyde (200 mg,
0.90 mmol) and bromomethylcyclohexane (316 mg, 1.80 mmol).
After purification by column chromatography on silica gel (cyclo-
hexane/EtOAc 9:1 to 8:2), the pure product (186 mg, 0.58 mmol,
65%) was obtained as a yellowish crystal. Rf ¼ 0.37 (7:3 cyclo-
1.02e1.16 (m, 2H); 13C NMR (100 MHz, Acetone-d6)
d ppm 181.2 (C),
167.7 (C), 167.2 (C), 158.5 (C), 156.5 (C), 146.0 (C), 134.9 (CH), 132.8
(C), 129.4 (C), 113.9 (CH), 112.4 (CH), 108.6 (C), 105.2 (C), 105.0 (CH),
101.6 (CH), 98.1 (CH), 91.9 (CH), 56.0 (CH3), 53.8 (CH2), 39.5 (CH),
31.4 (2xCH2), 27.0 (CH2), 26.4 (2xCH2); LRMS (ESIþ) m/z (%) 420
(100) [MþH]þ; HRMS (ESIþ) m/z calc. for C25H26NO5 420.1811
[MþH]þ, found 420.1803.
hexane/EtOAc); m.p. 89e92 ꢀC; 1H NMR (400 MHz, CDCl3)
d ppm
9.90 (s, 1H), 8.41 (d, J ¼ 2.0 Hz, 1H), 7.60 (s, 1H), 7.35 (dd, J ¼ 8.7,
2.0 Hz, 1H), 7.18 (d, J ¼ 8.7 Hz, 1H), 3.92 (d, J ¼ 7.3 Hz, 2H), 1.76e1.89
(m, 1H), 1.53e1.76 (m, 5H), 1.05e1.23 (m, 3H); 13C NMR (100 MHz,
4.1.2.3. (Z)-2-[N-(3-Bromobenzyl)-5-methoxyindol-3-ylmethylene]-
4,6-dihydroxybenzofuran-3(2H)-one (3o). The crude product was
prepared according to general procedure B starting from 1 (25 mg,
0.15 mmol) and aldehyde 2o (104 mg, 0.30 mmol). After purifica-
tion by column chromatography on silica gel (cyclohexane/EtOAc
8:2 to 5:5) and recrystallization in acetonitrile, the pure product
(24 mg, 0.05 mmol, 32%) was obtained as orange powder. Rf ¼ 0.16
(96:4 CH2Cl2/MeOH); m.p. > 270 ꢀC (decomposition); 1H NMR
CDCl3) d ppm 184.4 (CH), 139.6 (CH), 136.3 (C), 127.0 (CH), 126.9 (C),
124.8 (CH), 117.3 (C), 116.5 (C), 112.0 (CH), 54.0 (CH2), 38.5 (CH), 31.0
(2xCH2), 26.2 (CH2), 25.7 (2xCH2); LRMS (ESIþ) m/z (%) 344 (20),
322 (95) [M(81Br)þH]þ, 320 (100) [M(79Br)þH]þ; HRMS (ESIþ) m/z
calc. for C16H19BrNO 320.0650 [MþH]þ, found 320.0663.
(400 MHz, DMSO-d6)
d ppm 10.72 (bs, 1H), 10.72 (bs, 1H), 8.24 (s,
4.1.2. General procedure B for the synthesis of compounds (3c, m,
oꢂs)
1H), 7.56 (d, J ¼ 2.3 Hz, 1H), 7.45e7.50 (m, 2H), 7.39 (d, J ¼ 8.9 Hz,
1H), 7.29 (dd, J ¼ 8.1, 8.1 Hz, 1H), 7.17e7.23 (m, 1H), 7.02 (s, 1H), 6.84
(dd, J ¼ 8.9, 2.4 Hz, 1H), 6.21 (d, J ¼ 1.7 Hz, 1H), 6.06 (d, J ¼ 1.7 Hz,
To
a solution of 4,6-dihydroxybenzofuran-3(2H)-one 1 in
ethanol (3 mL/mmol) were added an aqueous solution of potassium
hydroxide (50%, 5 mL/mmol) and a benzaldehyde derivative
(1.0e3.5 equiv.). The solution was refluxed until TLC showed
complete disappearance of the starting material (2e5 h). After
cooling, ethanol was removed under reduced pressure, then the
residue was diluted into distilled water (50 mL/mmol) and an
aqueous solution of hydrochloric acid (10%) was added to adjust the
pH to 2e3. The mixture was then extracted with ethyl acetate or
dichloromethane. The combined organic layers were washed with
water and brine, dried over MgSO4, filtered off and concentrated
under reduced pressure to afford the corresponding crude (Z)-2-
benzylidenebenzofuran-3(2H)-one derivative.
1H), 5.54 (s, 2H), 3.83 (s, 3H); 13C NMR (100 MHz, DMSO-d6)
d ppm
178.3 (C), 166.9 (C), 166.4 (C), 157.9 (C), 154.9 (CH), 145.2 (C), 140.4
(C), 133.4 (CH), 130.9 (CH), 130.8 (CH), 130.4 (CH), 129.7 (C), 128.1
(C), 126.0 (CH), 121.8 (C), 112.9 (CH), 111.7 (CH), 107.9 (C), 103.7 (C),
102.4 (CH), 101.1 (CH), 97.5 (CH), 90.3 (CH), 55.4 (CH3), 48.9 (CH2);
LRMS (ESI-) m/z (%) 492 (80) [M(81Br)-H]ꢂ, 490 (70) [M(79Br)-H]ꢂ,
255 (85), 239 (70), 157 (100); HRMS (ESI-) m/z calc. for C25H17BrNO5
490.0290 [M-H]ꢂ, found 490.0294.
4.1.2.4. (Z)-2-(N-Methyl-5-bromoindol-3-ylmethylene)-4,6-
dihydroxybenzofuran-3(2H)-one (3p). The crude product was pre-
pared according to general procedure B starting from 1 (37 mg,
0.22 mmol) and aldehyde 2p (105 mg, 0.44 mmol). After purifica-
tion by column chromatography on silica gel (cyclohexane/EtOAc
8:2 to 5:5) and recrystallization in acetonitrile, the pure product
(35 mg, 0.09 mmol, 41%) was obtained as orange powder. Rf ¼ 0.16
(96:4 CH2Cl2/MeOH); m.p. > 300 ꢀC (decomposition); 1H NMR
4.1.2.1. (Z)-2-(N-Butyl-3-methylindol-2-ylmethylene)-4,6-
dihydroxybenzofuran-3(2H)-one (3c). The crude product was pre-
pared according to general procedure B starting from 1 (60 mg,
0.36 mmol) and aldehyde 2c (100 mg, 0.46 mmol). After purifica-
tion by column chromatography on silica gel (cyclohexane/EtOAc
(400 MHz, Acetone-d6)
d ppm 9.72 (bs, 1H), 8.92 (bs, 1H), 8.24 (d,

416
E. Baiceanu et al. / European Journal of Medicinal Chemistry 122 (2016) 408e418
J ¼ 1.8 Hz, 1H), 8.14 (s, 1H), 7.49 (d, J ¼ 8.7 Hz, 1H), 7.41 (dd, J ¼ 8.7,
1.8 Hz,1H), 7.04 (s,1H, CH), 6.34 (d, J ¼ 1.7 Hz,1H), 6.13 (d, J ¼ 1.7 Hz,
(33), 210 (50); HRMS (ESIþ) m/z calc. for C24H23BrNO4 468.0810
[MþH]þ, found 468.0809.
1H), 3.99 (s, 3H); 13C NMR (100 MHz, Acetone-d6)
d ppm 181.1 (C),
167.4 (C), 167.4 (C), 158.6 (C), 146.7 (C), 136.9 (C), 136.1 (CH), 130.2
(C), 126.2 (CH), 122.7 (CH), 114.9 (C), 113.0 (CH), 108.5 (C), 105.1 (C),
103.6 (CH), 98.3 (CH), 91.9 (CH), 33.8 (CH3); LRMS (ESI-) m/z (%) 386
(95) [M(81Br)-H]ꢂ, 384 (100) [M(79Br)-H]ꢂ; HRMS (ESI-) m/z calc.
for C18H11BrNO4 383.9871 [M-H]ꢂ, found 383.9870.
4.2. Biology
4.2.1. Reagents
Calcein-AM, elacridar (GF120918), MK-571 and cyclosporine A
were obtained from Sigma-Aldrich (France). All other reagents
were purchased from other companies (such as Euromedex France)
at the highest available purity grade (>95%).
4.1.2.5. (Z)-2-(N-Ethyl-5-bromoindol-3-ylmethylene)-4,6-
dihydroxybenzofuran-3(2H)-one (3q). The crude product was pre-
pared according to general procedure B starting from 1 (27 mg,
0.16 mmol) and aldehyde 2q (83 mg, 0.33 mmol). After purification
by column chromatography on silica gel (cyclohexane/EtOAc 8:2 to
5:5) and recrystallization in acetonitrile, the pure product (21 mg,
0.05 mmol, 33%) was obtained as orange powder. Rf ¼ 0.15 (96:4
CH2Cl2/MeOH); m.p. > 270 ꢀC (decomposition); 1H NMR (400 MHz,
4.2.2. Cell lines
The MDCKII WT and MDCKII ABCC2 cell lines were kindly pro-
vided by Prof. Dr. Piet Borst, (The Netherlands Cancer Institute
Amsterdam, Netherlands). These are epithelial kidney cells and the
MDCKII-ABCC2 cell line is obtained by transfection with a pCMV-
cMOAT retrovirus (clone 17; MDCKII-MOAT17) [39].
Acetone-d6)
d
ppm 9.74 (bs, 1H), 8.94 (bs, 1H), 8.25 (d, J ¼ 1.8 Hz,
To evaluate the selectivity of our compounds on other ABC
transporters, we used the NIH3T3 parental cell line and NIH3T3/
ABCB1 drug resistant cell line transfected with human MDR1/A-
G185 [49], purchased from American Type Culture Collection
(Manassas, VA) and used as described previously [23]. The BHK21
(Baby Hamster Kidney-21) cells and BHK21-ABCC1 stably trans-
fected with wild-type ABCC1 cells were used to evaluate the effect
on ABCC1 [50]. Flp-In™-293 cells were transfected by electropo-
ration using Neon® Transfection System (ThermoFisher scientific)
with pcDNA5-FRT-ABCC2 plasmid and pOG44 Flp-recombinant
expression vector. After selection by hygromycine B (Thermo-
Fisher scientific), the Flp-In system allows a stable integration and
expression of ABCC2 to deliver single copy isogenic cell lines. Flp-In-
293-ABCC2 cells were grown at 37 ꢀC in 5% CO2.
1H), 8.20 (s, 1H), 7.54 (d, J ¼ 8.7 Hz, 1H), 7.40 (dd, J ¼ 8.7, 1.9 Hz, 1H),
7.04 (s, 1H), 6.35 (d, J ¼ 1.7 Hz, 1H), 6.13 (d, J ¼ 1.7 Hz, 1H), 4.40 (q,
J ¼ 7.3 Hz, 2H), 1.52 (t, J ¼ 7.3 Hz, 3H); 13C NMR (100 MHz, Acetone-
d6) d ppm 181.1 (C), 167.9 (C), 167.4 (C), 158.6 (C), 146.7 (C), 135.8 (C),
134.6 (CH), 130.4 (C), 126.2 (CH), 122.9 (CH), 114.8 (C), 113.1 (CH),
108.7 (C), 105.1 (C), 103.6 (CH), 98.3 (CH), 92.0 (CH), 42.5 (CH2), 15.7
(CH3); LRMS (ESI-) m/z (%) 400 (100) [M(81Br)-H]ꢂ, 398 (90)
[M(79Br)-H]ꢂ, 255 (40), 157 (60); HRMS (ESI-) m/z calc. for
C
19H13BrNO4 398.0028 [M-H]ꢂ, found 398.0028.
4.1.2.6. (Z)-2-(N-Butyl-5-bromoindol-3-ylmethylene)-4,6-
dihydroxybenzofuran-3(2H)-one (3r). The crude product was pre-
pared according to general procedure B starting from 1 (60 mg,
0.36 mmol) and aldehyde 2r (129 mg, 0.46 mmol). After purifica-
tion by recrystallization in acetonitrile, the pure product (148 mg,
0.35 mmol, 96%) was obtained as red powder. Rf ¼ 0.16 (96:4
CH2Cl2/MeOH); m.p. > 260 ꢀC (decomposition); 1H NMR (400 MHz,
4.2.3. Cell culture
All cell lines were grown at 37 ꢀC in 5% CO2. MDCKII and NIH3T3
cells were cultured in Dulbecco’s modified Eagles’s medium
(DMEM high glucose) (PAA) supplemented with 10% fetal bovine
serum (FBS, PAA, GE Healthcare Life sciences, Velizy-Villacoublay,
France), 1% penicillin/streptomycin (PAA, GE Healthcare Life sci-
ences, Velizy-Villacoublay, France), without selection for the
ABCC2-transfected cell line, while for the NIH3T3-ABCB1 cell line
we used 60 ng/mL of colchicine (Sigma- Aldrich company, Saint
Quentin Fallavier, France). BHK-21 cells were cultured in DMEM-
F12, GlutaMAX Supplement culture medium (Gibco-Life Technol-
ogies, Saint Aubin, France) supplemented with 15 mM HEPES, 1%
penicillin/streptomycin (PAA) and 5% of heat-inactivated fetal
bovine serum (PAA, GE Healthcare Life sciences, Velizy-
Villacoublay, France), in the presence of 200 mM hygromycin for
transfected cells.
Acetone-d6)
d
ppm 9.76 (bs, 1H), 8.93 (bs, 1H), 8.25 (d, J ¼ 1.8 Hz,
1H), 8.19 (s, 1H), 7.55 (d, J ¼ 8.7 Hz, 1H), 7.39 (dd, J ¼ 8.7, 1.8 Hz, 1H),
7.04 (s, 1H), 6.34 (d, J ¼ 1.3 Hz, 1H), 6.14 (d, J ¼ 1.3 Hz, 1H), 4.36 (t,
J ¼ 7.2 Hz, 2H), 1.85e1.94 (m, 2H), 1.33e1.43 (m, 2H), 0.95 (t,
J ¼ 7.4 Hz, 3H); 13C NMR (100 MHz, Acetone-d6)
d ppm 181.1 (C),
167.4 5 (C), 167.4 (C), 158.6 (C), 146.6 (C), 136.1 (C), 135.1 (CH), 130.3
(C), 126.1 (CH), 122.8 (CH), 114.8 (C), 113.2 (CH), 108.6 (C), 105.1 (C),
103.6 (CH), 98.2 (CH), 91.9 (CH), 47.4 (CH2), 32.9 (CH2), 20.6 (CH2),
13.9 (CH3); LRMS (ESIþ) m/z (%) 430 (50) [M(81Br)þH]þ, 428 (60)
[M(79Br)þH]þ, 381 (100), 353 (60), 227 (57), 210 (57); HRMS (ESIþ)
m/z calc. for C21H19BrNO4 428.0497 [MþH]þ, found 428.0495.
4.1.2.7. (Z)-2-[N-(Cyclohexylmethyl)-5-bromoindol-3-ylmethylene]-
4,6-dihydroxybenzofuran-3(2H)-one (3s). The crude product was
prepared according to general procedure starting from 1 (63 mg,
0.38 mmol) and aldehyde 2s (241 mg, 0.75 mmol). After purifica-
tion by column chromatography on silica gel (cyclohexane/EtOAc
8:2 to 5:5) and recrystallization in acetonitrile, the pure product
(100 mg, 0.21 mmol, 56%) was obtained as orange powder. Rf ¼ 0.20
(96:4 CH2Cl2/MeOH); m.p. > 270 ꢀC (decomposition); 1H NMR
4.2.4. ABCC2-mediated calcein transport
4.2.4.1. High-throughput flow cytometry analysis for ABCC2 in-
hibitors screening. After a thorough understanding of the transport
of calcein in the MDCKII cells, we chose the ideal conditions to set
up a high-throughput screening protocol. For this purpose, we used
the Macsquant VYB (Miltenyi Biotec) cytometer integrated with a
microplate reader and an autosampler. Calcein-AM was used as a
(400 MHz, DMSO-d6)
d ppm 10.75 (s, 1H), 10.73 (s, 1H), 8.25 (d,
J ¼ 1.9 Hz, 1H), 8.10 (s, 1H), 7.58 (d, J ¼ 8.7 Hz, 1H), 7.35 (dd, J ¼ 8.7,
1.9 Hz, 1H), 6.97 (s, 1H), 6.23 (d, J ¼ 1.8 Hz, 1H), 6.06 (d, J ¼ 1.8 Hz,
1H), 4.14 (d, J ¼ 7.3 Hz, 2H), 1.75e1.90 (m, 1H), 1.43e1.72 (m, 5H,
substrate at 25 nM, in the presence of 1
MK-571 to inhibit endogenous ABCB1 and ABCC1. Our reference for
ABCC2 inhibition was 25 M cyclosporine. Using 96-well plates, the
mM GF120918 and 25 mM
m
5H), 0.93e1.24 (m, 5H); 13C NMR (100 MHz, DMSO-d6)
d
ppm 178.2
control cell line and the ABCC2-transfected cell line, we could
screen up to 20 compounds/h. Cells were seeded at a density of
3 ꢃ 103 cells/well for 48 h at 37 ꢀC in 5% CO2, then treated for
(C), 166.9 (C), 166.5 (C), 157.9 (C), 145.5 (C), 135.2 (C), 134.4 (CH),
128.7 (C),124.9 (CH),121.8 (CH),113.4 (C), 113.0 (CH), 107.1 (C),103.5
(C), 101.7 (CH), 97.5 (CH), 90.4 (CH), 52.1 (CH2), 38.2 (CH), 30.0
(2xCH2), 25.8 (CH2), 25.1 (2xCH2); LRMS (ESIþ) m/z (%) 470 (90)
[M(81Br)þH]þ, 468 (100) [M(79Br)þH]þ, 381 (65), 353 (70), 331
30 min with 5
mM of the inhibitor candidate, washed with a
phosphate buffered saline (PBS) solution and detached using
Accumax or Accutase enzyme solutions. Calcein was excited using

E. Baiceanu et al. / European Journal of Medicinal Chemistry 122 (2016) 408e418
417
488 nm laser; the intracellular fluorescence was monitored using
the B1 (525/50 nm) channel, and at least 5000 events were
collected. The percentage of inhibition was calculated by using the
following equation: % inhibition ¼ (ABCSþI e ABCS)/(ControlSþI e
ABCS)*100, where ABCSþI is the intracellular fluorescence of the
transfected cells in the presence of substrates þ inhibitors, ABCS is
the intracellular fluorescence of the transfected cells with only
substrate, and ControlSþI corresponds to the intracellular fluores-
cence of control cells in the presence of substrates and inhibitors. In
the initial screening that led to the identification of the class of
compounds optimized as ABCC2 inhibitors, compounds were
tested only once.
analyses of compounds. We would like to thank chemists involved
in chemical syntheses of this library, in particular Drs. Zouhair
Bouaziz, Christelle Marminon, Pascal Nebois, Laurent Ettouati,
Sylvie Radix, Thierry Lomberget, Nadia Walchshofer, Roland Barret
and Marie-Emmanuelle Million. We also thank ChemAxon for
providing us a license to their cheminformatics software Instant
JChem. EB was financially supported by the Erasmus þ PhD
mobility program accessed from the University of Medicine and
Pharmacy€Iuliu Hatieganu€, Cluj-Napoca, Romania, and a fellowship
received from the French National Research Agency, ANR-13-BSV5-
0001-01 (to PF and AB).
In order to calculate the IC50 of calcein transport in MDCKII WT
and ABCC2 cells, we tested several concentrations of the inhibitor
and plotted the inhibition curves, then extracted the concentration
for which 50% of calcein transport is inhibited. An additional
confirmatory assay was carried on in a similar manner in Flp-
In™293 and Flp-In™293 cells transfected with ABCC2. The in-
References
[1] K.M. Giacomini, S.-M. Huang, D.J. Tweedie, L.Z. Benet, K.L.R. Brouwer, X. Chu,
A. Dahlin, R. Evers, V. Fischer, K.M. Hillgren, K.A. Hoffmaster, T. Ishikawa,
D. Keppler, R.B. Kim, C.A. Lee, M. Niemi, J.W. Polli, Y. Sugiyama, P.W. Swaan,
J.A. Ware, S.H. Wright, S.W. Yee, M.J. Zamek-Gliszczynski, L. Zhang, Membrane
transporters in drug development, Nat. Rev. Drug Discov. 9 (2010) 215e236.
[2] H. Glavinas, P. Krajcsi, J. Cserepes, B. Sarkadi, The role of ABC transporters in
drug resistance, metabolism and toxicity, Curr. Drug Deliv. 1 (2004) 27e42.
[3] M. Pinto, D. Digles, G.F. Ecker, Computational models for predicting the
interaction with ABC transporters, Drug Discov. Today Technol. 12 (2014)
e69e77.
hibitors were tested in flow cytometry at 5
loaded with 25 nM calcein-AM, and used 25
reference.
m
m
M, the cells were
M cyclosporine as
4.2.4.2. Selectivity of inhibitor candidates towards ABCB1 and ABCC1.
NIH-3T3 cells transfected with ABCB1 were seeded at a density of
3.5 ꢃ 103 cells/well into 96-wells culture plates and incubated
48 h at 37 ꢀC in 5% CO2, whereas BHK21 cells transfected with
ABCC1 were seeded at 104 cells/well for 48 h. The transport
mediated by the ABC transporters was evaluated by exposing the
cells to calcein-AM (25 nM) in the presence or absence of the tested
[4] E.M. Leslie, R.G. Deeley, S.P. Cole, Multidrug resistance proteins: role of P-
glycoprotein, MRP1, MRP2, and BCRP (ABCG2) in tissue defense, Toxicol. Appl.
Pharmacol. 204 (2005) 216e237.
[5] A.H. Schinkel, J.W. Jonker, Mammalian drug efflux transporters of the ATP
binding cassette (ABC) family: an overview, Adv. Drug Deliv. Rev. 64 (2012)
138e153.
[6] R.J. Kathawala, P. Gupta, C.R. Ashby Jr., Z.S. Chen, The modulation of ABC
transporter-mediated multidrug resistance in cancer: a review of the past
decade, Drug Resist. Updates 18 (2015) 1e17.
[7] F. Montanari, G.F. Ecker, Prediction of drug-ABC-transporter interaction e
recent advances and future challenges, Adv. Drug Deliv. Rev. 86 (2015) 17e26.
[8] E. Baiceanu, G. Crisan, F. Loghin, P. Falson, Modulators of the human ABCC2:
hope from natural sources? Future Med. Chem. 7 (2015) 2041e2063.
[9] A.H. Schinkel, J.W. Jonker, Mammalian drug efflux transporters of the ATP
binding cassette (ABC) family: an overview, Adv. Drug Deliv. Rev. 55 (2003)
3e29.
[10] M. Drozdzik, C. Groer, J. Penski, J. Lapczuk, M. Ostrowski, Y. Lai, B. Prasad,
J.D. Unadkat, W. Siegmund, S. Oswald, Protein abundance of clinically relevant
multidrug transporters along the entire length of the human intestine, Mol.
Pharm. 11 (2014) 3547e3555.
[11] S.M. Huang, J.M. Strong, L. Zhang, K.S. Reynolds, S. Nallani, R. Temple,
S. Abraham, S.A. Habet, R.K. Baweja, G.J. Burckart, S. Chung, P. Colangelo,
D. Frucht, M.D. Green, P. Hepp, E. Karnaukhova, H.S. Ko, J.I. Lee, P.J. Marroum,
J.M. Norden, W. Qiu, A. Rahman, S. Sobel, T. Stifano, K. Thummel, X.X. Wei,
S. Yasuda, J.H. Zheng, H. Zhao, L.J. Lesko, New era in drug interaction evalu-
ation: US Food and Drug Administration update on CYP enzymes, trans-
porters, and the guidance process, J. Clin. Pharmacol. 48 (2008) 662e670.
[12] Guideline on the Investigation of Drug Interactions, 2012. http://www.ema.
europa.eu/docs/en_GB/document_library/Scientific_guideline/2012/07/
WC500129606.pdf (accessed 16.03.16).
compounds at 5
using Accumax or Accutase enzyme solutions. We used cyclo-
sporine A as a positive reference in our tests at 2.5 M for ABCB1
and 25 M for ABCC1.
mM, then washed with PBS solution and detached
m
m
4.2.5. Cytotoxicity
Cell survival was studied using the MTT (3-(4,5-
dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) colori-
metric assay. We tested the cytotoxicity of compounds on MDCKII
WT and MDCKII ABCC2 cell lines. Cells were seeded in 96-well
plates at the right density so that they don’t reach 100% conflu-
ence after 72 h of incubation at 37 ꢀC. After 24 h, we cultivated cells
in the presence or absence of the tested compounds (0e50
After 72 h, we added 22 L of MTT solution (5 mg/mL) to each well
and the plates were incubated at 37 ꢀC for 3 h. The medium was
discarded, and replaced with 100 L solution of DMSO/isopropanol
mM).
m
m
[13] H.M. Wortelboer, M. Usta, A.E. van der Velde, M.G. Boersma, B. Spenkelink,
J.J. van Zanden, I.M. Rietjens, P.J. van Bladeren, N.H. Cnubben, Interplay be-
tween MRP inhibition and metabolism of MRP inhibitors: the case of curcu-
min, Chem. Res. Toxicol. 16 (2003) 1642e1651.
[14] C. Beney, A.M. Mariotte, A. Boumendjel, An efficient synthesis of 4,6 dime-
thoxyaurones, Heterocycles 55 (2001) 967e972.
(1:1). After 10e20 min of gentle shaking, absorbance was measured
in a microplate reader at 570 nm, and this was corrected with the
absorbance measured at 690 nm. Data shown represents the
mean SD of at least three independent experiments.
[15] S. Okombi, D. Rival, S. Bonnet, A.M. Mariotte, E. Perrier, A. Boumendjel, Dis-
covery of benzylidenebenzofuran-3(2H)-one (Aurones) as inhibitors of
tyrosinase derived from human melanocytes, J. Med. Chem. 49 (2006)
329e333.
[16] A. Meguellati, A. Ahmed-Belkacem, W. Yi, R. Haudecoeur, M. Crouillere,
R. Brillet, J.M. Pawlotsky, A. Boumendjel, M. Peuchmaur, B-ring modified
aurones as promising allosteric inhibitors of hepatitis C virus RNA-dependent
RNA polymerase, Eur. J. Med. Chem. 80 (2014) 579e592.
[17] J.J. van Zanden, H.M. Wortelboer, S. Bijlsma, A. Punt, M. Usta, P.J. Bladeren,
I.M. Rietjens, N.H. Cnubben, Quantitative structure activity relationship
studies on the flavonoid mediated inhibition of multidrug resistance proteins
1 and 2, Biochem. Pharmacol. 69 (2005) 699e708.
[18] R. Evers, M. Kool, A.J. Smith, L. van Deemter, M. de Haas, P. Borst, Inhibitory
effect of the reversal agents V-104, GF120918 and Pluronic L61 on MDR1 Pgp-,
MRP1- and MRP2-mediated transport, Brit. J. Cancer 83 (2000) 366e374.
[19] A.L. Dogan, O. Legrand, A.M. Faussat, J.Y. Perrot, J.P. Marie, Evaluation and
comparison of MRP1 activity with three fluorescent dyes and three modula-
tors in leukemic cell lines, Leuk. Res. 28 (2004) 619e622.
4.3. Statistical analysis
All experiments were run in duplicates or triplicates, with
positive and negative controls, and then analyzed statistically using
specific software and analysis tools. We used FlowJo V10 for the
analysis of flow-cytometry data, and Excel 2013 package for the
interpretation and graphical representation of data. Statistical dif-
ferences between group means were determined by student’s two-
tailed t-test for unpaired samples.
Acknowledgements
We are grateful to Dr. Laurent Ettouati, Mrs Christine Ranquet
and Mrs Elodie Monniot for their assistance in the stock manage-
ment of the small molecule library of EA 4446 B2MC and HPLC/MS
[20] A. Boumendjel, S. Macalou, G. Valdameri, A. Pozza, C. Gauthier, O. Arnaud,
E. Nicolle, S. Magnard, P. Falson, R. Terreux, P.A. Carrupt, L. Payen, A. Di Pietro,

418
E. Baiceanu et al. / European Journal of Medicinal Chemistry 122 (2016) 408e418
Targeting the multidrug ABCG2 transporter with flavonoidic inhibitors:
in vitro optimization and in vivo validation, Curr. Med. Chem. 18 (2011)
3387e3401.
in
a
one-pot oxidation/cyclization reaction, Tetrahedron 67 (2011)
8731e8739.
[34] W. Zeinyeh, H. Xia, P. Lawton, S. Radix, C. Marminon, P. Nebois,
N. Walchshofer, Synthesis and modulation properties of imidazo[4,5-b]pyr-
idin-7-one and indazole-4,7-dione derivatives towards the Cryptosporidium
parvum CpABC3 transporter, Eur. J. Med. Chem. 45 (2010) 2480e2488.
[35] W. Zeinyeh, G. Alameh, S. Radix, C. Grenot, C. Dumontet, N. Walchshofer,
Design, synthesis and evaluation of progesterone-adenine hybrids as bivalent
inhibitors of P-glycoprotein-mediated multidrug efflux, Bioorg. Med. Chem.
Lett. 20 (2010) 3165e3168.
[36] W. Zeinyeh, Z. Mahiout, S. Radix, T. Lomberget, A. Dumoulin, R. Barret,
C. Grenot, L. Rocheblave, E.L. Matera, C. Dumontet, N. Walchshofer, Proges-
terone-adenine hybrids as bivalent inhibitors of P-glycoprotein-mediated
multidrug efflux: design, synthesis, characterization and biological evaluation,
Steroids 77 (2012) 1177e1191.
[21] G. Valdameri, E. Genoux-Bastide, B. Peres, C. Gauthier, J. Guitton, R. Terreux,
S.M. Winnischofer, M.E. Rocha, A. Boumendjel, A. Di Pietro, Substituted
chromones as highly potent nontoxic inhibitors, specific for the breast cancer
resistance protein, J. Med. Chem. 55 (2012) 966e970.
[22] E. Winter, P. Devantier Neuenfeldt, L.D. Chiaradia-Delatorre, C. Gauthier,
R.A. Yunes, R.J. Nunes, T.B. Creczynski-Pasa, A. Di Pietro, Symmetric bis-
chalcones as a new type of breast cancer resistance protein inhibitors with
a mechanism different from that of chromones, J. Med. Chem. 57 (2014)
2930e2941.
[23] O. Arnaud, A. Koubeissi, L. Ettouati, R. Terreux, G. Alame, C. Grenot, et al.,
Potent and fully noncompetitive peptidomimetic inhibitor of multidrug
resistance P-glycoprotein, J. Med. Chem. 53 (2010) 6720e6729.
[24] M. Hadjeri, A.M. Mariotte, A. Boumendjel, Alkylation of 2-phenyl-4-
quinolones: synthetic and structural studies, Chem. Pharm. Bull. 49 (2001)
1352e1355.
[25] C. Hundsdorfer, H.J. Hemmerling, J. Hamberger, M. Le Borgne, P. Bednarski,
C. Gotz, F. Totzke, J. Jose, Novel indeno[1,2-b]indoloquinones as inhibitors of
the human protein kinase CK2 with antiproliferative activity towards a broad
panel of cancer cell lines, Biochem. Biophys. Res. Comm. 424 (2012) 71e75.
[26] F. Bois, C. Beney, A.M. Mariotte, A. Boumendjel, A one-step synthesis of 5-
hydroxyflavones, Synlett 1999 (1999) 1480e1482.
[27] O. Arnaud, A. Boumendjel, A. Geze, M. Honorat, E.L. Matera, J. Guitton,
W.D. Stein, S.E. Bates, P. Falson, C. Dumontet, A. Di Pietro, L. Payen, The
acridone derivative MBLI-87 sensitizes breast cancer resistance protein-
expressing xenografts to irinotecan, Eur. J. Cancer 47 (2011) 640e648.
[28] C. Hundsdorfer, H.J. Hemmerling, C. Gotz, F. Totzke, P. Bednarski, M. Le
Borgne, J. Jose, Indeno[1,2-b]indole derivatives as a novel class of potent
human protein kinase CK2 inhibitors, Bioorg. Med. Chem. 20 (2012)
2282e2289.
[29] G. Jabor Gozzi, Z. Bouaziz, E. Winter, N. Daflon-Yunes, D. Aichele,
A. Nacereddine, C. Marminon, G. Valdameri, W. Zeinyeh, A. Bollacke, J. Guillon,
A. Lacoudre, N. Pinaud, S.M. Cadena, J. Jose, M. Le Borgne, A. Di Pietro, Con-
verting potent indeno[1,2-b]indole inhibitors of protein kinase CK2 into se-
lective inhibitors of the breast cancer resistance protein ABCG2, J. Med. Chem.
58 (2015) 265e277.
[30] F. Alchab, L. Ettouati, Z. Bouaziz, A. Bollacke, J.G. Delcros, C.G. Gertzen,
H. Gohlke, N. Pinaud, M. Marchivie, J. Guillon, B. Fenet, J. Jose, M. Le Borgne,
Synthesis, biological evaluation and molecular modeling of substituted indeno
[1,2-b]indoles as inhibitors of human protein kinase CK2, Pharmaceuticals 8
(2015) 279e302.
[31] G.J. Gozzi, Z. Bouaziz, E. Winter, N. Daflon-Yunes, M. Honorat, N. Guragossian,
C. Marminon, G. Valdameri, A. Bollacke, J. Guillon, N. Pinaud, M. Marchivie,
S.M. Cadena, J. Jose, M. Le Borgne, A. Di Pietro, Phenolic indeno[1,2-b]indoles
as ABCG2-selective potent and non-toxic inhibitors stimulating basal ATPase
activity, Drug Des. Dev. Ther. 9 (2015) 3481e3495.
[37] Z. Bouaziz, S. Issa, J. Gentili, A. Gratz, A. Bollacke, M. Kassack, J. Jose,
ꢀ
L. Herfindal, G. Gausdal, S.O. Doskeland, C. Mullie, P. Sonnet, C. Desgrouas,
N. Taudon, G. Valdameri, A. Di Pietro, M. Baitiche, M. Le Borgne, Biologically
active carbazole derivatives: focus on oxazinocarbazoles and related com-
pounds, J. Enz. Inh. Med. Chem. 30 (2015) 180e188.
[38] O. Fardel, E. Jigorel, M. Le Vee, L. Payen, Physiological, pharmacological and
clinical features of the multidrug resistance protein 2, Biomed. Pharmacother.
59 (2005) 104e114.
[39] R. Evers, M. Kool, V.L. Deemter, H. Janssen, J. Calafat, L.C. Oomen,
C.C. Paulusma, R.P. Oude Elferink, F. Baas, A.H. Schinkel, P. Borst, Drug export
activity of the human canalicular multispecific organic anion transporter in
polarized kidney MDCK cells expressing cMOAT (MRP2) cDNA, J. Clin. Investig.
101 (1998) 1310e1319.
[40] F. Tang, K. Horie, R.T. Borchardt, Are MDCK cells transfected with the human
MDR1 gene a good model of the human intestinal Mucosa? Pharm. Res. 19
(2002) 765e772.
[41] G. Valdameri, L. Pereira Rangel, C. Spatafora, J. Guitton, C. Gauthier, O. Arnaud,
A. Ferreira-Pereira, P. Falson, S.M.B. Winnischofer, M.E.M. Rocha, C. Tringali,
A. Di Pietro, Methoxy stilbenes as potent, specific, untransported, and non-
cytotoxic inhibitors of breast cancer resistance protein, ACS Chem. Biol. 7
(2012) 322e330.
[42] E. Winter, F. Lecerf-Schmidt, G. Gozzi, B. Peres, M. Lightbody, C. Gauthier,
C. Ozvegy-Laczka, G. Szakacs, B. Sarkadi, T.B. Creczynski-Pasa, A. Boumendjel,
A. Di Pietro, Structure-activity relationships of chromone derivatives toward
the mechanism of interaction with and inhibition of breast cancer resistance
protein ABCG2, J. Med. Chem. 56 (2013) 9849e9860.
[43] V. Zwick, A.O. Chatzivasileiou, N. Deschamps, M. Roussaki, C.A. Simoes-Pires,
A. Nurisso, I. Denis, C. Blanquart, N. Martinet, P.A. Carrupt, A. Detsi,
M. Cuendet, Aurones as histone deacetylase inhibitors: identification of key
features, Bioorg. Med. Chem. Lett. 24 (2014) 5497e5501.
[44] C. Zwergel, F. Gaascht, S. Valente, M. Diederich, D. Bagrel, G. Kirsch, Aurones:
interesting natural and synthetic compounds with emerging biological po-
tential, Nat. Prod. Commun. 7 (2012) 389e394.
[32] T. Lomberget, R. Barret, A new access to efaroxan and its 5-amino derivatives,
Tetrahedron Lett. 49 (2008) 715e718.
[45] A. Boumendjel, Aurones: a subclass of flavones with promising biological
potential, Curr. Med. Chem. 10 (2003) 2621e2630.
[33] F. Baragona, T. Lomberget, C. Duchamp, N. Henriques, E. Lo Piccolo, P. Diana,
A. Montalbano, R. Barret, Synthesis of 5-substituted 2,3-dihydrobenzofurans
[46] Y. Xu, Z. Dai, F. Chen, S. Gao, J. Pei, L. Lai, Deep learning for drug-induced liver
injury, J. Chem. Inf. Model. 55 (2015) 2085e2093.