Collateral sensitivity of resistant MRP1-overexpressing cells to flavonoids and derivatives through GSH efflux

Article abstract of DOI:10.1016/j.bcp.2014.05.017

The multidrug resistance protein 1 (MRP1) is involved in multidrug resistance of cancer cells by mediating drug efflux out of cells, often in co-transport with glutathione (GSH). GSH efflux mediated by MRP1 can be stimulated by verapamil. In cells overexp

Full text of DOI:10.1016/j.bcp.2014.05.017

Accepted Manuscript
Title: Collateral sensitivity of resistant MRP1-overexpressing
cells to flavonoids and derivatives through GSH efflux
Author: Doriane Lorendeau Lauriane Dury Estelle
˜
Genoux-Bastide Florine Lecerf-Schmidt Claudia Simo
¨
es-Pires Pierre-Alain Carrupt Raphael Terreux Sandrine
`
´ `
Magnard Attilio Di Pietro Ahcene Boumendjel Helene
Baubichon-Cortay
PII:
S0006-2952(14)00294-9
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.bcp.2014.05.017
BCP 11984
To appear in:
BCP
Received date:
Revised date:
Accepted date:
14-4-2014
16-5-2014
19-5-2014
Please cite this article as: Lorendeau D, Dury L, Genoux-Bastide E, Lecerf-Schmidt
¨
˜
¨
¨
F, Simo es-Pires C, Terreux P-AC, <ce:inter-ref id=intr0005xlink:href=mailto:pierre-
¨
¨
Raphael,
alain.carrupt@unige.ch>pierre-alain.carrupt@unige.ch</ce:inter-ref>,
Magnard S, Di Pietro A, Boumendjel A, Baubichon-Cortay H, Collateral sensitivity of
resistant MRP1-overexpressing cells to flavonoids and derivatives through GSH efflux,
Biochemical Pharmacology (2014), http://dx.doi.org/10.1016/j.bcp.2014.05.017
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Collateral sensitivity of resistant MRP1-overexpressing
cells to flavonoids and derivatives through GSH efflux
Doriane Lorendeau†, Lauriane Dury†, Estelle Genoux-Bastide‡, Florine Lecerf-Schmidt‡,
Claudia Simões-Pires¥, Pierre-Alain Carrupt¥ Raphaël Terreux§, Sandrine Magnard†, Attilio
,
Di Pietro†, Ahcène Boumendjel‡£and Hélène Baubichon-Cortay†£
†Equipe Labellisée Ligue 2012, Université Lyon 1, Univ Lyon, France; CNRS, UMR 5086 ;
BMSSI, IBCP,
7
passage du Vercors, F-69367, France, doriane.lorendeau@ibcp.fr,
lauriane.dury@ibcp.fr, s.magnard@ibcp.fr, a.dipietro@ibcp.fr, h.cortay@ibcp.fr
‡Département de Pharmacochimie Moléculaire Université Grenoble-Alpes, CNRS UMR
5063,
,
Grenoble,
France,
ahcene.boumendjel@ujf-grenoble.fr,
florine.lecerfschmidt@gmail.com
§Equipe “Bioinformatique : structures et interactions”, Université Lyon 1, Univ Lyon, France;
CNRS, UMR5086; Bases Moléculaires et Structurales des Systèmes Infectieux, IBCP 7
passage du Vercors, F-69367, France, r.terreux@ibcp.fr,
¥ Pharmacochimie, Section des Sciences Pharmaceutiques , Université de Genève, Université
de Lausanne , Genève, Suisse, pierre-alain.carrupt@unige.ch, claudia.avello@unige.ch,
£ Both senior investigators equally contributed to the work
Corresponding author information:
Hélène Baubichon-Cortay
Phone: (+33) 47272 2691
Fax (+33) 47272 2604
E-mail: hcortay@ibcp.fr
IBCP, UMR 5086 , 7 passage du Vercors, F-69367, France
Page 1 of 41

Abbreviations
ABC: ATP binding cassette, MDR: Multidrug resistance, MRP1 and MRP2: Multidrug
resistance protein 1 and 2, P-gp: P-Glycoprotein, BCRP: Breast cancer resistance protein,
BHK-21: Baby Hamster Kidney 21, GSH: reduced glutathione, GSSG: oxidized glutathione,
SR-verapamil: racemic verapamil, MTT: 3-[4,5-dimethylthiazol-2yl]-2,5-diphenyltetrazolium
bromide, NAC: N-AcetylCysteine, IC50: concentration required to obtain 50% of the maximal
effect, SR: Selective ratio, G418: Geneticin, PBS: Phosphate buffer salt, AcOEt: Ethyl
acetate, cHex: Cyclohexane, 3D-QSAR: Three dimensional quantitative structure activity
relationship, CoMSIA: Comparative molecular similarity indices, SD: standard deviation.
Page 2 of 41

Abstract
The multidrug resistance protein 1 (MRP1) is involved in multidrug resistance of cancer cells
by mediating drug efflux out of cells, often in co-transport with glutathione (GSH). GSH
efflux mediated by MRP1 can be stimulated by verapamil. In cells overexpressing MRP1, we
have previously shown that verapamil induced a huge intracellular GSH depletion which
triggered apoptosis of the cells. That phenomenon takes place in the more global anticancer
strategy called “collateral sensitivity” and could be exploited to eradicate some
chemoresistant cancer cells. Seeking alternative compounds to verapamil, we screened a
library of natural flavonoids and synthetic derivatives. A large number of these compounds
stimulate MRP1-mediated GSH efflux and the most active ones have been evaluated for their
cytotoxic effect on MRP1-overexpressing cells versus parental cells. Interestingly, some are
highly and selectively cytotoxic for MRP1-cells, leading them to apoptosis. However, some
others do not exhibit any cytotoxicity while promoting a strong GSH efflux, indicating that
GSH efflux is necessary but not sufficient for MRP1-cells apoptosis. In support to this
hypothesis, structure activity relationships show that the absence of a hydroxyl group at
position 3 of the flavonoid C ring is an absolute requirement for induction of MRP1-cells
death, but is not for GSH efflux stimulation. Chrysin (compound 8) and its derivatives,
compounds 11 and 22, exhibit a high selectivity toward MRP1-cells with a IC50 value of 4.1
µM for compound 11 and 4.9 µM for chrysin and compound 22, making them among the best
described selective killer compounds of multidrug ABC transporter-overexpressing cells.
Keywords
Multidrug resistance; Multidrug Resistance Protein 1; glutathione; flavonoids; collateral
sensitivity.
Page 3 of 41

Chemical compounds studied in this article: Chrysin (PubChem CID: 5281607); Galangin
(PubChem CID: 5281616), Kaempferol (PubChem CID: 5280863)
1. Introduction
Resistance of tumours to multiple structurally-unrelated anticancer drugs is a common clinical
problem that limits the curative potential of chemotherapy in clinical oncology. Eradication of
resistant cells in tumours would then have a considerable impact for curing cancer by
chemotherapy. The multidrug resistance (MDR) phenotype is in part due to the
overexpression within plasma membranes of drug transporters able to bind and extrude a wide
variety of compounds. Three human transporters belonging to the ATP-binding cassette
(ABC) proteins superfamily have been described, which can actively extrude anticancer drugs
from the cell at the expense of ATP hydrolysis: P-glycoprotein (P-gp or ABCB1), multidrug
resistance protein 1 (MRP1 or ABCC1) and breast cancer resistance protein (BCRP, or
ABCG2)[1]. MRP1 preferentially transports negatively-charged substrates, i.e. organic
anions, generally present as either glutathione-, glucuronate- or sulfate- conjugates. Its main
endogenous substrate is leukotriene C4, but it also transports reduced glutathione (GSH) and
its oxidized form (GSSG). In addition, some hydrophobic drugs, such as vinblastine and
vincristine, are also transported by MRP1 by symport with GSH[2]. MRP1 modulation would
be relevant since its expression is correlated with cancer incidence or prognostic impact[3].
Considering that MRP1 co-transports chemotherapeutic drugs with GSH, GSH depletion
could become an experimental approach to sensitize tumor cells to therapy[4].
Previously, we discovered that verapamil stimulates MRP1-mediated glutathione transport out
of the cell after binding to the transporter[5]. The fast and huge glutathione extrusion triggers
a selective apoptosis of cells overexpressing MRP1[6]. Hence, it is conceivable to propose the
Page 4 of 41

set-up of a novel and original strategy, i.e. a new therapeutic approach using ligands of MRP1
that act as stimulators of its GSH transport activity and that trigger apoptosis of cancer cells
overexpressing MRP1 without affecting normal cells. This strategy takes place in a specific
new therapeutic approach based on collateral sensitivity of cancer cells overexpressing
multidrug ABC transporters, where the physio-pathological changes associated with their
overexpression can be considered as an Achilles' heel[7], (Szakacs et al., Chem. Rev., 2014;
in press). In this research field, experiments have been first carried out on MDR cells
overexpressing P-glycoprotein and a number of compounds have been shown to induce
selective cytotoxicity towards MDR cells, such as NSC73306[8-12]. Recently, the orphan
drug tiopronin demonstrated collateral sensitivity toward various MDR cell lines
overexpressing P-glycoprotein or MRP1[13]. Successful targeting of multidrug-resistant cells
would reduce the tumor burden, and indirect problems observed with the classical use of
multidrug ABC transporters inhibitors.
When using verapamil as a sensitizer, we have showed that the S-isomer, but not the R-
isomer, of verapamil is the causative agent of MRP1-mediated GSH efflux [5], indicating a
sharp structure activity relationship. The major known drawback of verapamil is its
cardiotoxicity, which excludes any clinical development in cancer treatment. In order to find
new MRP1 ligands, less toxic and more potent than S-verapamil, we investigated flavonoids
analogs with a special focus on derivatives that have been shown to bind to ABC transporters.
On a series of flavonoids derivatives, we evaluated both the stimulation of GSH efflux and the
induction of cell death. We found very active compounds on both activities either in
transfected cells and selected cells and cell death was identified as apoptosis. Some other
flavonoids are only active on GSH efflux stimulation, which demonstrates that GSH efflux is
necessary but not sufficient to trigger cell death of MRP1-cells. Only those which do not have
an hydroxyl group at position 3 of the C ring are active for the induction of MRP1-cells death.
Page 5 of 41

The best one, 3-methoxy-chrysin (compound 11) (IC50 of 4.1 µM on MRP1-cells), exhibits a
high selective ratio upper than 24.5.
2. Materials and Methods
2.1. Chemical compounds
Compounds 1-8, 10 and 20-21 shown in table 1 were purchased from Sigma-Aldrich
company (Saint Quentin Fallavier, France). Compounds 9, 11-19 were prepared according to
a previous report[14]. Flavones 22-38 were prepared by condensation of 4’,6’-dimethoxy-2’-
hydroxyacetophenone (Sigma-Aldrich, Saint Quentin Fallavier, France) with substituted
benzaldehyde (Sigma-Aldrich, Saint Quentin Fallavier, France) ( in the presence of KOH in a
mixture of EtOH:H2O which affords chalcones intermediates. The chalcones intermediates
were treated in oxidative cyclization conditions with iodine in DMSO to afford 5,7-
dimethoxyflavones (24, 26 and 28-32[15]. During the oxidation step and only in one case, we
observed the incorporation of iodine into the A-aromatic ring at the C-6 carbon (compound
30). 5-Hydroxyflavones and 5,7-dihydroxyflavones (22, 23, 25 and 27) were obtained by
treating their methoxylated derivatives with boron tribromide (Sigma-Aldrich, Saint Quentin
Fallavier, France) in CH2Cl2 to afford either totally demethylated or a mixture of mono and
demethylated derivatives.
4’-Aminoflavone 32 was prepared by reduction of 4’-nitroflavone 31 with SnCl2 (Sigma-
Aldrich, Saint Quentin Fallavier, France) in HCl according to conventional procedures[16].
4’-benzamide flavones (33, 38) were obtained by treatment of 32 with substituted benzoyl
chlorides.
NMR spectra were recorded on a Brüker AC-400 instrument (400 MHz). Electrospray
ionization ESI mass spectra were acquired by the Analytical Department of Grenoble
University on an Esquire 300 PlusBrukerDaltonis instrument with a nanospray inlet.
Page 6 of 41

2.2. Antioxidant and radical scavenger activities measurements
All chemicals were purchased from Sigma-Aldrich company, Saint Quentin Fallavier, France.
2.2.1. Oxygen radical absorbance capacity (ORAC) assay.
Antioxidant activities of the tested compounds were determined by their ability to preserve
from oxygen radical species the fluorescence of fluorescein exposed to peroxyl radicals
generated by 2,2’-azobis(2-methylpropionamidine) dihydrochloride (AAPH)[17]. Black
polypropylene 96-well plates were filled with 60 nM fluorescein (in glycine buffer pH 8.3),
together with test samples or vehicle (in 2% DMSO), and pre-incubated at 40°C for 15 min.
The oxidative reaction was obtained by adding 5 mM AAPH (in glycine buffer pH 8.3) to
wells containing samples, positive control (here quercetin, which does not induce GSH efflux)
and oxidized fluorescein control. Non-oxidized fluorescein controls were added with the same
volume of assay buffer. The plate was incubated at 40°C for 90 min with continuous shaking
at 150 rpm, and cooled down at room temperature (5 min) prior to fluorescence reading at
485/528 nm. Pure compounds were first tested at 10 µM according to pre-established
screening criteria[18]. For each compound with a percentage of fluorescein protection higher
than 50%, the EC50 (concentration required to obtain 50% of the maximal effect) value was
determined by a dose-response curve of at least 6 concentrations.
2.2.2. DPPH radical scavenging assay.
Capacities of samples to scavenge the stable radical 2,2-diphenyl-1-picrylhydrazyl (DPPH)
were determined spectrophotometrically by measuring the loss of absorbance of DPPH at 515
nm [19]. 96-well microplates were filled with test sample solution (in ethanol containing up to
2% DMSO) or vehicle for the DPPH control. The reaction was initiated by the addition of 80
M DPPH (in ethanol). The decrease in absorbance at 515 nm was monitored at room
temperature after 10 min in order to determine the percentage of scavenged radical. Samples
Page 7 of 41

were tested at 40 µM for screening purposes to provide a concentration ratio of compound
over radical of 0.5 (R=0.5), as previously defined [18].For each compound with a percentage
of scavenged DPPH higher than 20%, the ER50 (ratio between EC50 and radical concentration)
values were determined by on dose-response curves of at least 6 concentrations by calculating
as ratios between compound concentration and radical concentration.
2.3. Molecular Modeling
Compounds were analysed by the Sybyl X 1.3 molecular modelling suite software[20], and
minimized by the MMFF94 force field[21] with a dielectric constant of 80 and an electrostatic
cut-off of 16Å. Minimized molecules were aligned on the central common core and put in a
database; lateral chains of molecules were manually checked and aligned on a common
position, and the modified conformation was minimized. The differences in internal energy
between the two conformations must be lower than 20 kcal.mol-1 in order to validate the
aligned conformations. Using the aligned molecules, a 3DQSAR study was initiated with the
CoMFA[22] program. Grids of electrostatic and steric potential fields were computed using
the C3 atom with a charge of +2 as a probe. Grids were filtered with 4 kcal.mol-1 as the
minimal variation to select probes, and the validation by leave one out (LOO) method was
chosen.[23]
2.4. Cell lines
All cell lines were grown at 37°C in 5 % CO2. BHK-21 (Baby Hamster Kidney-21) cells stably
transfected with wild-type mrp1 or K1333L mrp1 mutant have been previously described[24,
25]. They were grown in DMEM/F-12, GlutaMAX Supplement culture medium (Gibco-Life
Technologies, Saint Aubin, France) supplemented with 15 mM HEPES,
1
%
penicillin/streptomycin (PAA, GE Healthcare Life sciences, Velizy-Villacoublay, France) and 5
Page 8 of 41

% of heat-inactivated fetal bovine serum (PAA, GE Healthcare Life sciences, Velizy-
Villacoublay, France), in the presence of 200 µM methotrexate for transfected cells. The
small cell lung cancer cell lines NCI-H69 and H69AR (from ATCC number HTB-119 and
CRL-1135, respectively), which are human lung epithelial carcinoma, have been previously
described[26]. The H69AR line is derived from the NCI-H69 line after 14 months selection
with increased doxorubicin concentrations. They were grown in RPMI 1640 medium with
HEPES and L-glutamine (PAA, GE Healthcare Life sciences, Velizy-Villacoublay, France)
supplemented with 1 % penicillin/streptomycin, 10 % of heat-inactivated fetal bovine serum and
1mM sodium pyruvate (PAA, GE Healthcare Life sciences, Velizy-Villacoublay, France).
H69AR were monthly selected with doxorubicine at 0.8 µM (Sigma-Aldrich, Saint Quentin
Fallavier, France). The other cell lines were grown in DMEM High Glucose culture medium
with L-Glutamine (PAA, GE Healthcare Life sciences, Velizy-Villacoublay, France).
supplemented with 1 % penicillin/streptomycin and 10 % of heat-inactivated fetal bovine
serum. The NIH3T3/P-gp (ATCC)[27] grown medium was also supplemented with 60 ng/mL
of colchicine (Sigma-Aldrich company, Saint Quentin Fallavier, France). Human MDR1-
G185 P-glycoprotein-transfected (Pgp-expressing) HEK293 cells, a gift from Dr S. Bates[28],
and ABCG2-transfected HEK293 cells, obtained by transfection of cells with pcDNA3.1-
ABCG2 using lipofectamine (Invitrogen, Life technologies, Saint Aubin, France), were
respectively grown with 2 and 1 mg/mL of G418 (Invitrogen, Life technologies, France).
MDCKII (Madin-Darby Canine kidney) transfected with mrp2 (MCDKII-MOAT17)[29] and
the MDCKII parental cell line were a gift from Dr E. Georges (McGill University, Montreal,
Canada). MDCKII-MOAT17 was maintained with 800 µg/mL of G418.
Page 9 of 41

2.5. Cell proliferation analysis.
The MTT colorimetric assay was used to assess the sensitivity of parental cells and MRP1-
cells to the different compounds (by determination of IC50: concentration of compound where
50 % of the growth inhibition are observed)[30]. Briefly, growth inhibition assays were
performed by plating cells in 96-well plates at density 2.0 × 104 cells/well for NCI-H69 and
H69AR, at 3.2 ×103 cells/well for BHK-21-wt and 4 × 103 cells/well for BHK-21-MRP1 cell
lines. Cells were cultured for 17 h before addition of compounds diluted in complete culture
medium, with a fixed final DMSO concentration of 0.5 %. The cells were then incubated for
24 or 72 h in a humidified tissue-culture chamber (37 °C, 5 % CO2). Surviving cells were
detected by the bromo-3(4,5-dimethyl-2-thiazoyl) 2,5-diphenyltetrazolium (MTT) assay. The
MTT solution was added to a final concentration of 0.5 mg/ml and cells were then incubated
for 4 h at 37 °C. For NCI-H69 and H69AR cell lines, the 96-well plates were thereafter
centrifuged at 700 g for 10 min to pull down floating cells, the MTT dye was carefully
aspirated and 200 µL of 50 % ethanol/50 % DMSO mixing solution was added to each well to
solubilize the reduced formazan dye under shaking. The absorbance of formazan was
determined by spectrophotometry at 570 nm. The effect of drugs on cell growth was
determined from differences in absorbance between drug-treated cells versus untreated cells
or solvent control. IC50 values were calculated from dose response curves obtained from
triplicate independent experiments. For cell rescue experiments, BHK-21-wt and BHK-21-
MRP1 cells were incubating with flavonoids at 20 µM with increasing concentrations of N-
acetylcysteine (NAC) (purchased from Sigma-Aldrich) for 72h and with increasing
concentrations of Z-VAD-FMK, a generic pan-caspase inhibitor (purchased from
Calbiochem, Merck Chemicals, UK) for 24h.
Page 10 of 41

2.6. Total cellular glutathione determination.
All chemicals were purchased from Sigma-Aldrich, Saint Quentin Fallavier, France
Total cellular glutathione contents (GSH + GSSG) were measured using the enzymatic
method described by Tiezte[31] and modified by Anderson[32] to the microtiter plate. Cells
were seeded into 96-well plates at a density of 1.0 × 105 cells/well and cultured for 17 h
before addition of compounds at 20 µM. DMSO concentration was fixed at 0.5 %. After a 3h-
incubation, cells were washed twice with 100 µL cold PBS, and then lysed by 100 µL PBS
supplemented with 0.1 % Triton X-100. After homogenization, lysis was performed by
incubation 1 h at 4° C under strong shaking prior to freezing. Then 25 µL of lysate were
transferred to a new plate for GSH titration, and 10 µL to another plate for protein
quantitation. GSH determination was performed by adding 100 µL of reaction buffer
containing 266.6 µM NADPH, 555 µM 5,5'-dithiobis-(2-nitrobenzoic acid), 10 U/mL GSH
reductase and PBS-Triton 0.1 %. The plate was then placed in a microplate reader and the
absorbance was read at 412 nm every 30 s for 2 min. The content of total cellular glutathione
was quantified by comparison with known glutathione standards. The slope was determined
for each sample and led to the determination of the GSH contents. Protein titration was
performed on the remaining 10 µL by the Bicinchoninic Acid method assay. The measured
total glutathione was expressed in nmol per mg protein, and the results were obtained from
triplicates.
3. Results
3.1. Flavonoids and derivatives as inducers of a strong intracellular GSH depletion
Page 11 of 41

Flavonoids are natural compounds extensively studied for their anti- and pro-oxidant
properties. They have a broad range of biological activities by interacting with various
proteins and their use in cancer chemoprevention and chemotherapy is of great interest[33,
34]. In order to circumvent the MDR phenotype of cancer cells due to multidrug ABC
transporters, flavonoids and derivatives were shown to inhibit drug transport mediated by P-
glycoprotein[35-38], MRP1[39, 40] and BCRP[41, 42]. A new type of activity has been
observed for some flavonoids that are effective in depleting the glutathione content of tumor
cells overexpressing MRP1, thanks to MRP1-mediated GSH efflux[33, 43-45].
In 2004, we reported for the first time collateral sensitivity of MRP1-cells. Indeed, we showed
that verapamil triggered selective apoptosis of MRP1-overexpressing cells after a strong
cellular GSH depletion[6], resulting from the stimulation of GSH efflux. This efflux
stimulated by verapamil was demonstrated to be mediated by MRP1[46]. Afterwards, our
results were confirmed by Laberge et al. who studied both verapamil and the flavonoid
apigenin[47], this last one being only active on H69AR cells but not on transfected HelaABCC1
cells.
Here, we further prospect for compounds with the same desirable properties by searching the
requested structural requirements. This could lead to the elaboration of novel anticancer
agents that would act as collateral sensitivity agents. Natural and synthetic flavonoids (Table
1) were submitted to a first screen to assess their ability to modify the intracellular levels of
GSH in two cell lines: Baby Hamster Kidney-21 wild-type parental cells (BHK-21-wt)
(control cells) and BHK-21 cells transfected by the MRP1 gene (BHK-21-MRP1), which only
express MRP1[24], previously used to study collateral sensitivity to verapamil[6].
The total cellular glutathione levels were measured in both control cells and MRP1-
transfected cells treated or not with each compound at 20 µM and expressed in percentage
(untreated cells glutathione levels considered as 100 %). The 20 µM concentration was
Page 12 of 41

selected from previous works describing the effects of flavonoids on cellular GSH depletion
[44, 45]. Percentages of depletion caused by tested compounds were calculated for each cell
line and the effect on the MRP1-transfected cells was then corrected from effects on control
cells to get the “net GSH depletion". This "net depletion" can be assumed as a "net GSH
efflux" since it is specifically induced by stimulated MRP1, as already described on
membrane vesicles for verapamil and for bioflavonoids[45, 46]. In addition, an increase of
extracellular GSH was measured in a concomitant way with the decrease of intracellular GSH
in BHK-21-MRP1 cells treated with verapamil (unpublished results).
The values of resulting net efflux of cellular GSH are shown in Table 1. Our hit compound
previously identified was the unsubstituted flavone (compound 1) which induced a 42 % net
efflux of GSH. From our previous investigations dealing with flavonoids interactions with
ABC proteins[35-42, 48], we learned that the hydroxylation pattern of the A-ring was
decisive. In the present study, the hydroxylation at C-5 or C-7 (compounds 2 and 6) induces a
significant increase in GSH efflux, whereas the hydroxylation at C-6 (compound 4) is harmful
for the GSH depletion.
The OH-methylation of 7-hydroxyflavone (compound 7) leads to a decreased activity, which
highlights the influence of the hydrogen bond and/or hydrophobic properties. The
concomitant presence of hydroxyl groups at both C-5 and C-7 positions (compound 8,
compared to 2 and 6) does not lead to a significant increase in GSH depletion. Interestingly,
the replacement of the 5-OH or 6-OH (compounds 2 and 4) by an amino group (compounds 3
and 5) has no major impact on the activity. The presence of an additional hydroxyl group on
the B-ring (apigenin, compound 20) provides a less active derivative. Insertion of a hydroxyl
group at the 3-position of the C-ring favors the activity, leading to one of the most active
derivatives (galangin, compound 10). The positive role of the 3-OH group is confirmed by
comparing the activity of derivatives 20 and 21.
Page 13 of 41

The role of hydroxylation on the A-ring is confirmed by the screening of a series of
derivatives with methoxy groups at the 5 and 7 positions and having different substituents at
the 4’-position. As shown in table 1 (compounds 24, 26, 28-32), all derivatives are less active,
indicating once again the crucial role of the hydroxyl groups at 5 and 7 positions.
Finally, we set up 5 compounds (33-38) that have a flavonoidic skeleton and bear substituents
at 4’-position that make them verapamil-like structures. Unfortunately, we observed only very
low, if any, GSH efflux.
3.2. 3D-QSAR for GSH efflux mediated by MRP1
The 38 compounds that were screened for GSH efflux were modelled and aligned and a 3D-
QSAR study with CoMSIA [20-23] was initiated. The PLS algorithm with an optimal number
of component set at 4, gave a good correlation coefficient, q2 = 0.723, with the LOO method
and r2 = 0.952 in calibration (Figure 1).
The analysis of volumes contribution (Figure 2) reveals a negative steric contribution (yellow)
at both ends of the molecule, suggesting a narrow pocket for this type of compounds. The
volume size is well suited for flavonoid compounds with molecular backbone deprived of
heavy moieties
3.3. Induction of MRP1-cells death by selected flavonoids.
Compounds cytotoxicity at 20 µM was then investigated against BHK-21-wt cells and BHK-
21-MRP1. Figure 3 shows that the compounds that are the most cytotoxic against BHK-21-
MRP1 cells and the less against parental cells are chrysin (compound 8) and its derivatives
(compounds 11 to 15) and the two compounds 5,7-dihydroxy-4'-fluoroflavone and 5,7-
dihydroxy-4'-bromoflavone (22 and 25 respectively). Some other compounds, 5-
hydroxyflavone, 8-(1,1-DMA)-3-methoxy-chrysin and apigenin (compounds 2, 17 and 20
Page 14 of 41

respectively) exhibit a good cytotoxicity against BHK-21-MRP1 cells but are also cytotoxic
against BHK-21-wt cells to a lesser extent. Surprisingly, compounds 3, compounds 10, 16-18
(galangin and derivatives), compound 19, compound 21 (kaempferol), compounds 23, 24 and
32, while inducing a significant GSH depletion, do not trigger cell death of BHK-21-MRP1
cells. Given these observations, it can be suggested that there is no strict correlation between
both activities of GSH efflux and cell death. However, among compounds triggering selective
cytotoxicity, GSH efflux stimulation is the pre-requisite for cell death induction. Indeed, the
addition of N-acetyl-cysteine (Figure 4), a synthetic precursor of intracellular cysteine and
GSH, or GSH (data not shown) in culture medium prevents cell death induced by chrysin
(compound 8) and flavonoids 11 and 22, both strong GSH efflux inducers. This was
previously shown with verapamil[6], indicating that these flavonoids would act in a similar
way to verapamil regarding GSH depletion and cell death.
MRP1 has been demonstrated to be the causative agent of the selective cell death induced by
verapamil[6]. Flavonoids (flavonoid 8, 11 and 22 for the most active ones), like verapamil,
are only effective on BHK-21 cells overexpressing wild-type MRP1 and not on those
overexpressing MRP1 mutated in its hydrolytic site (K1333L) (Figure 5) suggesting that the
functional transporter is required to induce selective cell death.
The addition of the pan-caspase inhibitor Z-VAD-FMK in culture medium prevents the death
of MRP1-cells triggered by the natural compound chrysin and the flavonoid derivatives 11
and 22 after 24h of co-incubation (Figure 6). This result is in favor of an apoptotic cell death
pathway induced by these flavonoids.
In order to establish valuable structure-activity relationships, the IC50 of the compounds were
determined. Only the 24 compounds that induced more than 38.5 % of net GSH depletion
were tested. We included verapamil for comparison purposes and compound 38 as a negative
control that does not induce GSH depletion (as shown in Table 1). Collateral sensitivity
Page 15 of 41

(abbreviated as "SR" for selective ratio) was calculated for each compound as the ratio of IC50
for parental cells divided by IC50 for MRP1-cells. A SR greater than 2 is indicative of
significant collateral sensitivity. In order to validate our proof of concept in a human
chemoresistant cell line, we chose to determine SR on both human drug-sensitive cells NCI-
H69 (carcinoma cells from small cell lung cancer) and drug-resistant MRP1-overexpressing
cells H69AR (derived from NCI-H69 cells after 14 months of culture in the presence of
increasing adriamycin concentrations), cell lines that are of better relevance in anticipation of
in vivo studies. H69AR cells overexpress MRP1, and P-gycoprotein and BCRP were not
detected by western-blot (data not shown). They are characterized by a lower basal GSH
content compared to NCI-H69 (11 ± 5 nmol GSH/mg protein versus 39 ± 11 nmol GSH/mg
protein), which results from the overexpression of MRP1, and they behave similarly to BHK-
21-MRP1 cells toward GSH efflux in the presence of verapamil, which results in a near
complete GSH depletion (1.5 ± 0.5 nmol GSH/mg after a 3h-exposure).
As shown in Table 2, galangin (compound 10) and kaempferol (compound 21) provide only
weak activities on H69AR cells, as shown for BHK-21-MRP1, indicating that the absence of
a hydroxyl group at position 3 seems to be an absolute requirement to obtain an active
molecule. This is confirmed by results obtained with chrysin (compound 8) which differs
from galangin only by the absence of the 3-OH group and which exhibits one of the most
potent activities with a high selective ratio (SR) upper than 20.5, even higher than verapamil’s
(Table 2). The later structural requirement is also supported by comparison of compound 16
(8-(1,1-DMA)-galangin) with compound 17 (8-(1,1-DMA)-3-O-methylgalangin) where 3-OH
is substituted by 3-OMe. Compound 16 has no cytotoxic effect on MRP1-overexpressing cells
(SR = 1) while compound 17 exhibits a SR = 10. The presence of hydrophobic substituents
like prenyl and geranyl does not seem to induce any additional effect on cytotoxicity
(compound 8 versus compounds 12, 13, 14, 15). As observed for BHK-21-MRP1 cells,
Page 16 of 41

galangin and derivatives (compounds 10, 16, 18), kaempferol (compound 21) and compounds
3, 19, 23, 24 and 32 do not trigger cell death of H69AR, while inducing a GSH depletion
(Table 1 and 2 and Figure 3). In overall, when comparing both transfected cells and selected
cells models, the effects of the different compounds are quite similar, even if BHK-21-wt
cells are slightly more sensitive than the NCI-H69 cells.
Flavonoids are known to possess, for many of them, antioxidant capacities[49]. The absence
of cytotoxicity observed for example with galangin (compound 10) and kaempferol
(compound 21) could be explained by such properties, despite the strong GSH efflux
triggered by these compounds. High reactivity of the hydroxyl groups of some flavonoids is
responsible for their free radical scavenger properties [50], so they could scavenge the
oxidative consequence of this GSH efflux. We checked this hypothesis by measuring the
antioxidant properties of some flavonoids, as described in the Materials and Methods section.
Among those exhibiting oxygen radical absorbance capacity, preventing fluorescein
oxidation, some of them (compounds 8, 11, 25 and 27) induce cell death and others
(compounds 10, 21 and 32) do not (Table 3). We also determined, using the radical DPPH•,
their peroxyl radical scavenger capacity. Among compounds triggering GSH depletion but no
selective cell death (compounds 3, 10, 16, 18, 19, 21, 23, 24, 32) , only galangin (compound
10) and kaempferol (compound 21) exhibit this capacity with an ER50 (ratio between EC50
and radical concentration) of 2.44 ± 0.05 and 0.36 ± 0.02 respectively, compared to 0.28 ±
0.01 determined for quercetin. Taking into account their strong scavenging capacities, we
cannot exclude that the absence of cytotoxicity of galangin and kaempferol (compounds 10
and 21) can be related to their antioxidant properties. However, some other flavonoids also
show antioxidant capacities (table 3) but induce cell death.
Finally, the two flavonoid derivatives 11 and 22 and the natural compound chrysin
(compound 8) exhibit the highest SR (Table 2), despite antioxidant properties (Table 3).
Page 17 of 41

Compound 11 gives the best IC50 for cytotoxicity against H69AR (4.1 µM) (Table 2), even
though it does not induce the highest GSH efflux (66%) (Table 1).
3.4. 3D-QSAR for cell death
The 25 initial compounds screened for cell death were modelled and aligned and a 3D-QSAR
study with CoMSIA [20-23] was initiated. The establishment of a standard QSAR proving to
be difficult, we chose an IC50 of 70 µM as a cut off and compounds showing a higher IC50
were considered to be inactive (and coded with a 100 value) while compounds showing a
lower value IC50 were considered to be active (and coded with a 0 value) (Table 4).
Analysis revealed that 3 molecules out of the 25 initial compounds were outliers. The PLS
algorithm, with an optimal number of 6 components, gave a good correlation coefficient: q2 =
0.803 with the LOO method and r2 = 0.975 in calibration (Figure 7). All compounds were
well predicted, positive or negative, with LOO methods.
The “yes or no” response of the cell death QSAR shows the difficulty to predict cellular cell
death which is often multifactorial and involves different pathways. The 3D-QSAR of GSH
efflux and cell death (Figures 2 and 8) reveal strong differences between the two maps of
fields, particularly regarding electrostatic fields whose contributions are negative only for
QSAR of cell death. This suggests that the target of offending flavonoids might not be unique.
The lack of direct correlation between GSH efflux and selective cell death emphasizes this
hypothesis.
3.5. Specificity of "cytotoxic" flavonoids for MRP1 as compared to other ABC drug
transporters
We checked the specificity of the active flavonoid derivative 11 and chrysin (compound 8) for
MRP1. Thus, we tested their effect on resistant cell lines overexpressing either P-gp
Page 18 of 41

(NIH3T3-P-gp and HEK-293-MDR1), MRP2 (MDCKII-MRP2) and BCRP (HEK-293-
BCRP) in comparison with their respective parental cell lines (Table 5). None of these other
MDR cell lines are hypersensitive to the two compounds, indicating that they selectively act
on MRP1-cells.
4. Discussion
Cancer cells are usually treated with chemotherapeutic drugs, which unfortunately also induce
cytotoxicity for non cancer cells. Many strategies are currently in development to act
specifically against cellular targets only present in cancer cells. To overcome the MDR
phenotype of cancer cells related to multidrug ABC transporters (P-glycoprotein, MRP1 and
BCRP), different approaches are being studied. The classical one consists of using inhibitors
in order to chemosensitize resistant cancer cells to chemotherapeutic drugs. A more promising
trend is to exploit the phenomenon of collateral sensitivity of resistant cancer cells specifically
due to the overexpression of the MDR transporters[51]. This overexpression can indeed be
considered as an Achilles’ heel that can be selectively targeted through this new therapeutic
approach. This strategy, in addition to improving response to chemotherapy, would become a
novel and specific anticancer therapy since the compounds would target resistant cancer cells
that overexpress either P-glycoprotein, MRP1 or BCRP[12, 52] and (Szakacs et al., Chem.
Rev., 2014; in press).
We describe in this work new findings about flavonoids which can be used as a new class of
modulators able to selectively target chemoresistant cells overexpressing MRP1. We
demonstrate for some of them a "killer" potency instead of the more traditional effects in
cancer prevention. The flavonoids chrysin (compound 8), 3-methoxy-chrysin (compound 11)
and 5,7-dihydroxy-4'-fluoroflavone (compound 22) are more effective (SR > 20 to 24.5) than
verapamil and show IC50 of about 4 µM, which is 3-fold lower compared to that of verapamil.
Page 19 of 41

Compounds 8 and 22 trigger apoptosis within MRP1-cells as shown by prevention of
BHK21-MRP1 cells death using the pan-caspase inhibitor Z-VAD-FMK. These compounds
would be expected to be less toxic than verapamil since polyphenolic compounds are common
in the diet and have been suggested to have a number of beneficial health effects including
prevention of cancer, cardiovascular disease, diabetes, and others.
Some studies showed that both VP-16 cell line, which overexpresses MRP1, and KB-V1 cell
line, which overexpresses P-glycoprotein, develop a relative sensitivity to the sensitizer
tiopronin of respectively 42.5-fold and 51-fold compared to the parental cell line. However, in
both cases, collateral sensitivity exhibited by these MDR cell lines was not linked to
interaction of tiopronin with MRP1 or P-gp, but seemed to require alterations associated with
the development of MDR[13]. In another work using indomethacin, the doxorubicin-resistant
MRP1-overexpressing human SCLC cell line GLC4-Adr was shown to be highly sensitive to
the compound when compared to the parental line GLC4. Although GSH level was decreased
by 31.5 % after 2 hours of incubation, the reduced cellular GSH level seemed not to be the
primary cause of indomethacin-induced apoptosis[53]. On the contrary, we previously
demonstrated that a strong GSH level diminution induced upon activation of MRP1 transport
activity by verapamil[46], triggered apoptosis of MRP1-overexpressing cells[6]. When using
cells overexpressing an inactive mutant of MRP1 defective for transport activity, neither GSH
efflux (data known shown) nor concomitant cell death (figure 5) are observed with
flavonoids, as already shown with verapamil[6], indicating an effective MRP1-dependent
mechanism. This GSH efflux inducing cell death is directly dependent of MRP1-
overexpression since MRP1-transfected cells (BHK21-MRP1 cells) show the same profile of
selective sensitivity than selected MRP1-overexpressing cells (H69AR), although the two cell
lines are strongly different. Furthermore, incubation with GSH or N-acetyl-cystein prevents
Page 20 of 41

cell death (figure 4) showing a direct link between intracellular GSH level decrease and cell
death.
Cancer cells are in a permanent oxidative stress state and one mechanism they use for
hindering oxidative stress is to increase their intracellular GSH level and regulate the different
components of the GSH synthesis system. Furthermore, beyond antioxidant properties of
GSH, there is a strong relationship between apoptosis and GSH[54]. In this regard, several
agents that target the GSH system have been developed in an attempt to improve cancer
chemotherapy[55]. Our work lies with this approach. However, whereas flavonoids were used
in addition to chemotherapeutic drugs to potentialize their effect[43], we describe for the first
time their direct effect on resistant MRP1-cells as inducers of apoptosis. Cells overexpressing
MRP1 (BHK21-MRP1 and H69AR) present a dramatically decreased intracellular GSH level
in comparison to parental cells (BHK-21-wt and NCI-H69)[6]. Indeed, in the presence of
verapamil, GSH content of H69AR is further reduced to a very low level, as shown in this
work (results § 3.3) and as described[47], which probably induces a strong oxidative stress in
the cell or acts as a critical regulator of apoptosis[54]. Modulation via MRP1 of cellular GSH
levels would be therefore a powerful approach for cancer therapy[56]. Although depletion of
intracellular GSH via stimulation of MRP1 transport activity by bioflavonoids was previously
reported [45], we report here the first study about the structural requirements among flavonoid
derivatives as selective cancer agents. Apigenin, already shown to be effective on
H69AR[47], also leads to strong GSH efflux and cell death of BHK21-MRP1 cells (Table 1
and Figure 3) . Chrysin is the best natural compound with a SR higher than 20.5 and an IC50
of 4.9 ± 3.9 µM. Chrysin has already been used for its strong stimulation of GSH efflux in
MRP1-overexpressing cells, in order to potentiate the doxorubicin effect in human lung
epithelial cancer[33]. From our results, this could be explained by the own cytotoxicity of
chrysin against MRP1-cells. Potentiation of prooxidants toxicities towards cancer cells using
Page 21 of 41

GSH depletion, induced for example by flavonoids, is already proposed as a strategy to
sensitize cancer cells[43, 56] and we propose here that chrysin and derivatives could be
directly used as selective anticancer agents against resistant cancer cells, their low range IC50
being compatible with biologically achievable plasma level measured for a similar flavonoid,
quercetin [57].
On the other hand, a hydroxyl group at position 3 of the flavonoid skeleton is incompatible
with cytotoxicity against MRP1-overexpressing cells since galangin or kaempferol only
induce GSH efflux but do not trigger cytotoxicity. This structural element is greatly valuable
for further investigations of flavonoid derivatives. A strong GSH efflux, though necessary, is
not sufficient for cell death induction, as also highlighted for indomethacin[53] and xanthones
we previously studied[58], questioning the mechanism through which GSH depletion triggers
the downstream events of apoptosis[59]. Free radical scavenger properties of galangin
(compound 10) and kaempferol (compound 21) and probably those of galangine derivatives
(compounds 16 and 18) are strong. However, this does not totally explain their lack of
selective cytotoxicity since other compounds exhibiting similar antioxidant capacities are able
to induce MRP1-cells death. Furthermore, flavonoids 3, 19, 21, 23, 24 and 32 also are not
cytotoxic for MRP1-cells whereas they induce GSH efflux. Therefore, it can be supposed that
there would be different targets, the first one being MRP1 that leads to a huge cellular GSH
depletion, while other targets, probably involving additional pathways, would lead to cell
death solely when cells are enough depleted in their GSH content. The “yes or no” response
of the QSAR of cell death (Figure 7) emphasizes this hypothesis. The implication of other
cellular pathways in the induction of apoptosis is under investigation.
Among the flavonoid derivatives tested, compounds 11 and 22 are highly selective for cells
overexpressing MRP1 as illustrated by their high SR (> 24.5 and > 20.5 respectively). The
exact value of these SR could not be determined since the tested concentrations were
Page 22 of 41

technically limited to 100 µM, concentration which is not toxic for control cells. They induce
a collateral sensitivity towards cells overexpressing MRP1 at low concentrations (IC50 of 4.1
and 4.9 µM respectively) and are much more powerful than tropionin (IC50 of 290 µM)[13] or
indomethacin (IC50 30 µM)[53], and may be considered among the best compounds described
for their collateral sensitivity towards MDR cells that overexpress MRP1[13, 53, 58, 60].
In conclusion, our work brings a significant framework to consolidate the novel approach
called collateral sensitivity as an emerging strategy in cancer treatment. The structural
elements found here will allow to design new anticancer agents that target selectively MDR
tumors overexpressing MRP1 through the induction of a strong GSH efflux, probably leading
to oxidative stress. The in vivo efficacy of our best compounds is currently being assessed
using H69AR xenografts in mice. Whereas safety is expected, as often described using dietary
bioflavonoids, their cytotoxicity is being evaluated. Depending on the results, we could
modify flavonoid structure while maintaining structural requirements needed for cell death
induction described in this work. The relevance of this new strategy can be emphasized on the
basis that the recurrent tumors often acquire MDR phenotype and that this novel anticancer
therapy, which is selective for MDR tumors, can be applied upon failure of conventional
chemotherapies, while being non toxic for non MRP1-overexpressing cells.
Acknowledgments
We thank la Région Rhône-Alpes for the doctoral fellowship accorded to E. G-B. F. L-S is
thankful to Région Rhône-Alpes for the financial support during her joint PhD thesis between
the University Joseph Fourier and the University of Geneva. We thank the French Ministère
de l’Enseignement Supérieur et de la Recherche and the Ligue Nationale contre le Cancer for the
doctoral fellowships accorded to Doriane Lorendeau and Lauriane Dury. A part of the
financial support of this work was provided by grants from the French ANR and Hungarian
Page 23 of 41

NIH (2010-INT-1101-01) and from the Ligue Nationale contre le Cancer (Equipe labellisée Ligue
2012). Authors are grateful to ANR (Agence Nationale pour la Recherche) for financial support
(Labex Arcane (ANR-11-LABX-0003-01).
Figure captions
Figure 1. 3D-QSAR study based on GSH efflux at compound concentration of 20 µM.
Figure 2. CoMSIA volumes contribution of the 38 molecules submitted to the study of 3D-
QSAR for GSH efflux mediated by MRP1. Volumes were plotted with 30 for negative
contribution (yellow for steric and red for electrostatic fields) and 70 for positive contribution
(green for steric and blue for electrostatic fields).
Figure 3. Cell survival of parental cells (BHK-21-wt) and MRP1-overexpressing cells (BHK-
21-MRP1) in presence of 20 µM of flavonoid derivatives. Error bars represent standard
deviation from three different experiments done in triplicate.
Figure 4. Prevention of BHK-21-MRP1 cell death induced by compounds 8, 11 and 22 and
racemic verapamil (SR-verapamil) by extracellular addition of N-acetyl-cystein. Error bars
represent standard deviation from three different experiments done in triplicate.
Figure 5. Cell survival of parental cells (BHK-21-wt), MRP1-overexpressing cells (BHK-21-
MRP1) and mutated MRP1-overexpressing cells (BHK-21-MRP1m) in presence of racemic
verapamil (SR-verapamil) and compounds 8, 11 and 22.
Figure 6. Prevention of BHK-21-MRP1 cell death induced by compounds 8, 11 and 22 and
racemic verapamil (SR-verapamil) by extracellular addition of the pan-caspase inhibitor Z-
VAD-FMK. Error bars represent standard deviation from three different experiments done in
triplicate.
Page 24 of 41

Figure 7. 3D-QSAR of cell death with attributed coded values (0 active/100 inactive).
Figure 8. CoMSIA volumes contribution of the 22 molecules submitted to the study of 3D-
QSAR for cell death. Volumes were plotted with 30 for negative contribution (yellow for steric
and red for electrostatic fields) and 70 for positive contribution (green for steric and blue for
electrostatic fields).
Figure captions for Figures 2 and 8 in black and white
Figure 2. CoMSIA volumes contribution of the 38 molecules submitted to the study of 3D-
QSAR for GSH efflux mediated by MRP1. Volumes were plotted with 30 for negative
contribution (light grey for steric fields) and 70 for positive contribution (black for steric and dark
grey for electrostatic fields).
Figure 8. CoMSIA volumes contribution of the 22 molecules submitted to the study of 3D-
QSAR for cell death. Volumes were plotted with 30 for negative contribution (dark grey for
electrostatic fields) and 70 for positive contribution (light grey for steric and black for electrostatic
fields).
References
[1]
[2]
[3]
[4]
Glavinas H, Krajcsi P, Cserepes J, Sarkadi B. The role of ABC transporters in drug resistance,
metabolism and toxicity. Curr Drug Deliv 2004;1:27‐42.
Deeley RG, Cole SP. Substrate recognition and transport by multidrug resistance protein 1
(ABCC1). FEBS Lett 2006;580:1103‐11.
Munoz M, Henderson M, Haber M, Norris M. Role of the MRP1/ABCC1 multidrug transporter
protein in cancer. IUBMB Life 2007;59:752‐7.
Traverso N, Ricciarelli R, Nitti M, Marengo B, Furfaro AL, Pronzato MA, et al. Role of
Glutathione in Cancer Progression and Chemoresistance. Oxid Med Cell Longev
2013;2013:10.
[5]
[6]
Perrotton T, Trompier D, Chang XB, Di Pietro A, Baubichon‐Cortay H. (R)‐ and (S)‐verapamil
differentially modulate the multidrug‐resistant protein MRP1. J Biol Chem 2007;282:31542‐8.
Trompier D, Chang XB, Barattin R, du Moulinet D'Hardemare A, Di Pietro A, Baubichon‐Cortay
H. Verapamil and its derivative trigger apoptosis through glutathione extrusion by multidrug
resistance protein MRP1. Cancer Res 2004;64:4950‐6.
[7]
Turk D, Szakacs G. Relevance of multidrug resistance in the age of targeted therapy. Curr
Opin Drug Discov Devel 2009;12:246‐52.
Page 25 of 41

[8]
Hall MD, Brimacombe KR, Varonka MS, Pluchino KM, Monda JK, Li J, et al. Synthesis and
structure‐activity evaluation of isatin‐beta‐thiosemicarbazones with improved selective
activity toward multidrug‐resistant cells expressing P‐glycoprotein.
2011;54:5878‐89.
J Med Chem
[9]
Hall MD, Salam NK, Hellawell JL, Fales HM, Kensler CB, Ludwig JA, et al. Synthesis, activity,
and pharmacophore development for isatin‐beta‐thiosemicarbazones with selective activity
toward multidrug‐resistant cells. J Med Chem 2009;52:3191‐204.
[10]
Ludwig JA, Szakacs G, Martin SE, Chu BF, Cardarelli C, Sauna ZE, et al. Selective toxicity of
NSC73306 in MDR1‐positive cells as a new strategy to circumvent multidrug resistance in
cancer. Cancer Res 2006;66:4808‐15.
[11]
[12]
[13]
Szakacs G, Paterson JK, Ludwig JA, Booth‐Genthe C, Gottesman MM. Targeting multidrug
resistance in cancer. Nat Rev Drug Discov 2006;5:219‐34.
Turk D, Hall MD, Chu BF, Ludwig JA, Fales HM, Gottesman MM, et al. Identification of
compounds selectively killing multidrug‐resistant cancer cells. Cancer Res 2009;69:8293‐301.
Goldsborough AS, Handley MD, Dulcey AE, Pluchino KM, Kannan P, Brimacombe KR, et al.
Collateral sensitivity of multidrug‐resistant cells to the orphan drug tiopronin. J Med Chem
2011;54:4987‐97.
[14]
Boumendjel A, Bois F, Beney C, Mariotte AM, Conseil G, Di Pietro A. B‐ring substituted 5,7‐
dihydroxyflavonols with high‐affinity binding to P‐glycoprotein responsible for cell multidrug
resistance. Bioorg Med Chem Lett 2001;11:75‐7.
[15]
[16]
Yao N, Song A, Wang X, Dixon S, Lam KS. Synthesis of flavonoid analogues as scaffolds for
natural product‐based combinatorial libraries. J Comb Chem 2007;9:668‐76.
Cholody WM, Martelli S, Konopa J. 8‐Substituted 5‐[(aminoalkyl)amino]‐6H‐v‐triazolo[4,5,1‐
de]acridin‐6‐ones as potential antineoplastic agents. Synthesis and biological activity. J Med
Chem 1990;33:2852‐6.
[17]
[18]
[19]
[20]
Huang D, Ou B, Hampsch‐Woodill M, Flanagan JA, Prior RL. High‐throughput assay of oxygen
radical absorbance capacity (ORAC) using a multichannel liquid handling system coupled with
a microplate fluorescence reader in 96‐well format. J Agric Food Chem 2002;50:4437‐44.
Nurisso A, Simoes‐Pires C, Martel S, Cressend D, Guillot A, Carrupt PA. How to increase the
safety and efficacy of compounds against neurodegeneration? A multifunctional approach.
Chimia (Aarau) 2012;66:286‐90.
Ancerewicz J, Migliavacca E, Carrupt P‐A, Testa B, Brée F, Zini R, et al. Structure–Property
Relationships of Trimetazidine Derivatives and Model Compounds as Potential Antioxidants.
Free Radic Biol Med 1998;25:113‐20.
Allen MS, Tan YC, Trudell ML, Narayanan K, Schindler LR, Martin MJ, et al. Synthetic and
computer‐assisted analyses of the pharmacophore for the benzodiazepine receptor inverse
agonist site. J Med Chem 1990;33:2343‐57.
[21]
[22]
[23]
[24]
[25]
Cramer RD, Patterson DE, Bunce JD. Comparative molecular field analysis (CoMFA). 1. Effect
of shape on binding of steroids to carrier proteins. J Am Chem Soc 1988;110:5959‐67.
Cramer RD, 3rd, Patterson DE, Bunce JD. Recent advances in comparative molecular field
analysis (CoMFA). Prog Clin Biol Res 1989;291:161‐5.
Marshall GR, Cramer RD, 3rd. Three‐dimensional structure‐activity relationships. Trends
Pharmacol Sci 1988;9:285‐9.
Chang XB, Hou YX, Riordan JR. ATPase activity of purified multidrug resistance‐associated
protein. J Biol Chem 1997;272:30962‐8.
Hou Y, Cui L, Riordan JR, Chang X. Allosteric interactions between the two non‐equivalent
nucleotide binding domains of multidrug resistance protein MRP1. J Biol Chem
2000;275:20280‐7.
[26]
Mirski SE, Gerlach JH, Cole SP. Multidrug resistance in a human small cell lung cancer cell line
selected in adriamycin. Cancer Res 1987;47:2594‐8.
Page 26 of 41

[27]
[28]
[29]
Cardarelli CO, Aksentijevich I, Pastan I, Gottesman MM. Differential effects of P‐glycoprotein
inhibitors on NIH3T3 cells transfected with wild‐type (G185) or mutant (V185) multidrug
transporters. Cancer res 1995;55:1086‐91.
Robey RW, Honjo Y, Morisaki K, Nadjem TA, Runge S, Risbood M, et al. Mutations at amino‐
acid 482 in the ABCG2 gene affect substrate and antagonist specificity. Br J Cancer
2003;89:1971‐8.
Evers R, Kool M, van Deemter L, Janssen H, Calafat J, Oomen LC, et al. Drug export activity of
the human canalicular multispecific organic anion transporter in polarized kidney MDCK cells
expressing cMOAT (MRP2) cDNA. J Clin Invest 1998;101:1310‐9.
[30]
[31]
Mosmann T. Rapid colorimetric assay for cellular growth and survival: application to
proliferation and cytotoxicity assays. J Immunol Methods 1983;65:55‐63.
Tietze F. Enzymic method for quantitative determination of nanogram amounts of total and
oxidized glutathione: applications to mammalian blood and other tissues. Anal Biochem
1969;27:502‐22.
[32]
[33]
Anderson ME. Determination of glutathione and glutathione disulfide in biological samples.
Methods Enzymol 1985;113:548‐55.
Brechbuhl HM, Kachadourian R, Min E, Chan D, Day BJ. Chrysin enhances doxorubicin‐
induced cytotoxicity in human lung epithelial cancer cell lines: The role of glutathione.
Toxicol Appl Pharm 2012;258:1‐9.
[34]
[35]
Chahar MK, Sharma N, Dobhal MP, Joshi YC. Flavonoids: A versatile source of anticancer
drugs. Pharmacogn Rev 2011;5:1‐12.
Boumendjel A, Di Pietro A, Dumontet C, Barron D. Recent advances in the discovery of
flavonoids and analogs with high‐affinity binding to P‐glycoprotein responsible for cancer cell
multidrug resistance. Med Res Rev 2002;22:512‐29.
[36]
[37]
[38]
[39]
[40]
[41]
[42]
Conseil G, Baubichon‐Cortay H, Dayan G, Jault JM, Barron D, Di Pietro A. Flavonoids: a class of
modulators with bifunctional interactions at vicinal ATP‐ and steroid‐binding sites on mouse
P‐glycoprotein. Proc Natl Acad Sci U S A 1998;95:9831‐6.
Conseil G, Decottignies A, Jault JM, Comte G, Barron D, Goffeau A, et al. Prenyl‐flavonoids as
potent inhibitors of the Pdr5p multidrug ABC transporter from Saccharomyces cerevisiae.
Biochemistry 2000;39:6910‐7.
Di Pietro A, Conseil G, Perez‐Victoria JM, Dayan G, Baubichon‐Cortay H, Trompier D, et al.
Modulation by flavonoids of cell multidrug resistance mediated by P‐glycoprotein and related
ABC transporters. Cell Mol Life Sci 2002;59:307‐22.
Boumendjel A, Baubichon‐Cortay H, Trompier D, Perrotton T, Di Pietro A. Anticancer
multidrug resistance mediated by MRP1: recent advances in the discovery of reversal agents.
Med Res Rev 2005;25:453‐72.
Trompier D, Baubichon‐Cortay H, Chang XB, Maitrejean M, Barron D, Riordon JR, et al.
Multiple flavonoid‐binding sites within multidrug resistance protein MRP1. Cell Mol Life Sci
2003;60:2164‐77.
Ahmed‐Belkacem A, Pozza A, Munoz‐Martinez F, Bates SE, Castanys S, Gamarro F, et al.
Flavonoid structure‐activity studies identify 6‐prenylchrysin and tectochrysin as potent and
specific inhibitors of breast cancer resistance protein ABCG2. Cancer Res 2005;65:4852‐60.
Arnaud O, Boumendjel A, Geze A, Honorat M, Matera EL, Guitton J, et al. The acridone
derivative MBLI‐87 sensitizes breast cancer resistance protein‐expressing xenografts to
irinotecan. Eur J cancer 2011;47:640‐8.
[43]
[44]
Kachadourian R, Day BJ. Flavonoid‐induced glutathione depletion: potential implications for
cancer treatment. Free Radic Biol Med 2006;41:65‐76.
Kachadourian R, Leitner HM, Day BJ. Selected flavonoids potentiate the toxicity of cisplatin in
human lung adenocarcinoma cells: a role for glutathione depletion. Int J Oncol 2007;31:161‐
8.
[45]
Leslie EM, Deeley RG, Cole SP. Bioflavonoid stimulation of glutathione transport by the 190‐
kDa multidrug resistance protein 1 (MRP1). Drug Metab Dispos 2003;31:11‐5.
Page 27 of 41

[46]
[47]
Loe DW, Deeley RG, Cole SP. Verapamil stimulates glutathione transport by the 190‐kDa
multidrug resistance protein 1 (MRP1). J Pharmacol Exp Ther 2000;293:530‐8.
Laberge RM, Karwatsky J, Lincoln MC, Leimanis ML, Georges E. Modulation of GSH levels in
ABCC1 expressing tumor cells triggers apoptosis through oxidative stress. Biochem
Pharmacol 2007;73:1727‐37.
[48]
[49]
[50]
[51]
Boumendjel A, Macalou S, Valdameri G, Pozza A, Gauthier C, Arnaud O, et al. Targeting the
multidrug ABCG2 transporter with flavonoidic inhibitors: in vitro optimization and in vivo
validation. Curr Med Chem 2011;18:3387‐401.
Nijveldt RJ, van Nood E, van Hoorn DE, Boelens PG, van Norren K, van Leeuwen PA.
Flavonoids: a review of probable mechanisms of action and potential applications. Am J Clin
Nutr 2001;74:418‐25.
van Acker SA, van den Berg DJ, Tromp MN, Griffioen DH, van Bennekom WP, van der Vijgh
WJ, et al. Structural aspects of antioxidant activity of flavonoids. Free Radic Biol Med
1996;20:331‐42.
Szakacs G, Annereau JP, Lababidi S, Shankavaram U, Arciello A, Bussey KJ, et al. Predicting
drug sensitivity and resistance: profiling ABC transporter genes in cancer cells. Cancer Cell
2004;6:129‐37.
[52]
[53]
Hall MD, Handley MD, Gottesman MM. Is resistance useless? Multidrug resistance and
collateral sensitivity. Trends Pharmacol Sci 2009;30:546‐56.
de Groot DJ, van der Deen M, Le TK, Regeling A, de Jong S, de Vries EG. Indomethacin induces
apoptosis via a MRP1‐dependent mechanism in doxorubicin‐resistant small‐cell lung cancer
cells overexpressing MRP1. Br J Cancer 2007;97:1077‐83.
[54]
[55]
[56]
[57]
Franco R, Cidlowski JA. Apoptosis and glutathione: beyond an antioxidant. Cell Death Differ
2009;16:1303‐14.
Singh S, Khan AR, Gupta AK. Role of glutathione in cancer pathophysiology and therapeutic
interventions. J Exp Ther Oncol 2012;9:303‐16.
Leitner HM, Kachadourian R, Day BJ. Harnessing drug resistance: Using ABC transporter
proteins to target cancer cells. Biochem Pharmacol 2007;74:1677‐85.
Manach C, Morand C, Crespy V, Demigne C, Texier O, Regerat F, et al. Quercetin is recovered
in human plasma as conjugated derivatives which retain antioxidant properties. FEBS Lett
1998;426:331‐6.
[58]
Genoux‐Bastide E, Lorendeau D, Nicolle E, Yahiaoui S, Magnard S, Di Pietro A, et al.
Identification of Xanthones as Selective Killers of Cancer Cells Overexpressing the ABC
Transporter MRP1. ChemMedChem 2011;6:1478‐84.
[59]
[60]
Coppola S, Ghibelli L. GSH extrusion and and the mitochondrial pathway of apoptotic
signalling. Biochem Soc trans 2000;28:56‐61.
Barattin R, Perrotton T, Trompier D, Lorendeau D, Di Pietro A, d'Hardemare Adu M, et al.
Iodination of verapamil for a stronger induction of death, through GSH efflux, of cancer cells
overexpressing MRP1. Bioorg Med Chem 2010;18:6265‐74.
Page 28 of 41

Table 1. Structures of studied compounds and effect on cellular GSH efflux in BHK-21 cells.
R1
R1
R2
R2
3'
D
H
(1,1-DMA): 1,1-dimethylallyl
H
N
2'
4'
5'
N
4'
R3
R3
B
O
6'
B
8
8 1'
2
9
O
7
C
O
MeO
5'
O
C
O
O
MeO
A
O
7
6
A
3
6
geranyl
prenyl
10
3
4
5
5
O
OMe
O
O
OMe
flavone 4'-N-benzamides
Flavone 4'-N-benzamide
(derivatives 30 - 35)
flavone
Flavone
(derivatives 33-38)
Net GSH efflux
at 20 µM (%)
Compound
Compound name
flavone
CLogP
1
3.74
3.47
3.17
3.23
2.79
3.23
3.77
2.94
3.85
2.65
2.93
5.19
7.04
5.19
7.04
4.84
5.12
4.9
42.4
63.8
61.3
16.7
11.7
76.3
42.2
76.5
39.3
84.6
65.6
46.1
58.0
46.2
52.1
59.2
43.2
39.9
57.6
43
2
5-hydroxyflavone
5-aminoflavone
6-hydroxyflavone
6-aminoflavone
7-hydroxyflavone
7-methoxyflavone
3
4
5
6
7
8
chrysin (5,7-dihydroxyflavone)
5-hydroxy-7-ethoxyflavone
galangin (3,5,7-trihydroxyflavone)
3-methoxy-chrysin
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
6-prenyl-chrysin
6-geranyl-chrysin
8-prenyl-chrysin
8-geranyl-chrysin
8-(1,1-DMA)-galangin
8-(1,1-DMA)-3-methoxy-chrysin
8-prenyl-galangin
8-prenyl-3-methoxy-chrysin
5.18
2.46
2.17
3.11
3.64
3.92
3.75
3.65
4.65
5.46
5.27
6.28
3.71
apigenin (5,7,4’-trihydroxyflavone)
kaempferol (3,5,7,4’-tetrahydroxyflavone)
5,7-dihydroxy-4'-fluoroflavone
72.5
74.5
42.8
43.8
47.4
26.7
38.5
0
5-hydroxy-7-methoxy-4'-fluoroflavone
5,7-dimethoxy-4’-fluoroflavone
5,7-dihydroxy-4’-bromoflavone
5,7-dimethoxy-4’-ethoxyflavone
5,7-dihydroxy-4’-tertiobutylflavone
5,7-dimethoxy-4’-tertiobutylflavone
5,7-dimethoxy-4’-isopropylflavone
5,7-dimethoxy-6-iodo-4’-isopropylflavone
5,7-dimethoxy-4’-nitroflavone
12.4
15
0
Page 29 of 41