Purification of breast cancer resistance protein ABCG2 and role of arginine-482

Article abstract of DOI:10.1007/s00018-006-6159-7

Human ABCG2 was efficiently overexpressed in insect cell membranes, solubilized with 3-[(3-cholamidopropyl)dimethyl ammonio]-1-propanesulfonate, and purified through N-terminal hexahistidine tag. Its functionality was assessed by high vanadate-sensitive ATPase activity, and nucleotide-binding capacity. Interestingly, the R482T point mutation increased both maximal hydrolysis rate and affinity for MgATP, and lowered sensitivity to vanadate inhibition. Direct nucleotide binding, as monitored by quenching of intrinsic fluorescence, indicated a mutation-related preference for ATP over ADP. The R482T mutation only produced a limited change, if any, on the binding of drug substrates, indicating that methotrexate, on the one hand, and rhodamine 123 or doxorubicin, on the other hand, bound similarly to wildtype and mutant transporters whether or not they were subject to cellular transport. In addition, the characteristic inhibitors GF120918 and 6-prenylchrysin, which alter mitoxantrone efflux much better for wild-type than mutant ABCG2, bound similarly to purified ABCG2, while the highly-potent Ko143 bound in the nanomolar range also effective in inhibition of drug transport. All results indicate that the role of the arginine-482 mutation on substrate drug transport and inhibitor efficiency is not mediated by changes in drug binding. Birkhaeuser Verlag, 2006.

Full text of DOI:10.1007/s00018-006-6159-7

Cell. Mol. Life Sci. 63 (2006) 1912–1922
420-682X/06/161912-11
DOI 10.1007/s00018-006-6159-7
Birkhäuser Verlag, Basel, 2006
1
Cellular and Molecular Life Sciences
©
Research Article
Purification of breast cancer resistance protein ABCG2 and
role of arginine-482
A. Pozzaa, †, J. M. Perez-Victoriab, †,A. Sardo ,A.Ahmed-Belkacem and A. Di Pietro *
a
a
a,
a
Institut de Biologie et Chimie des Protéines, UMR5086 CNRS-Université de Lyon et IFR128 BioSciences Lyon-
Gerland, 7 Passage du Vercors, 69367 Lyon Cedex 07 (France), Fax: +33 4 7272 2604, e-mail: a.dipietro@ibcp.fr
b
Instituto de Parasitologia y Biomedicina ‘Lopez-Neyra’, Consejo Superior de Investigaciones Cientificas, Granada
(
Spain)
Received 10 April 2006; received after revision 22 May 2006; accepted 12 June 2006
Online First 17 July 2006
Abstract. Human ABCG2 was efficiently overexpressed in methotrexate, on the one hand, and rhodamine 123 or
insect cell membranes, solubilized with 3-[(3-cholamid- doxorubicin, on the other hand, bound similarly to wild-
opropyl)dimethyl ammonio]-1-propanesulfonate, and puri- type and mutant transporters whether or not they were
fied through N-terminal hexahistidine tag. Its function- subject to cellular transport. In addition, the characteris-
ality was assessed by high vanadate-sensitive ATPase tic inhibitors GF120918 and 6-prenylchrysin, which alter
activity, and nucleotide-binding capacity. Interestingly, mitoxantrone efflux much better for wild-type than mu-
the R482T point mutation increased both maximal hy- tant ABCG2, bound similarly to purified ABCG2, while
drolysis rate and affinity for MgATP, and lowered sensi- the highly-potent Ko143 bound in the nanomolar range
tivity to vanadate inhibition. Direct nucleotide binding, also effective in inhibition of drug transport. All results
as monitored by quenching of intrinsic fluorescence, in- indicate that the role of the arginine-482 mutation on sub-
dicated a mutation-related preference for ATP over ADP. strate drug transport and inhibitor efficiency is not medi-
The R482T mutation only produced a limited change, ated by changes in drug binding.
if any, on the binding of drug substrates, indicating that
Keywords. Breast cancer resistance protein, multidrug resistance, ABC transporter, substrates, inhibitors, ATPase
activity, anticancer chemotherapy.
Introduction
Despite being a half-transporter, functioning at least as
a homodimer, ABCG2 exhibits a pattern of transported
The multidrug resistance (MDR) phenotype of cancer drugs largely overlapping with P-glycoprotein (for a re-
cells has been often related to overexpression within view see [6–8]). In contrast, only a few inhibitors, such as
plasma membranes of ATP-binding cassette (ABC)- GF120918, are efficient on both transporters [9], whereas
transporters able to efflux many chemotherapeutic drugs a few compounds, namely fumitremorgin C [10] and its
at the expense of ATP hydrolysis. More recently, in addi- potent derivative Ko143 [11], 6-prenylchrysin [12] and
tion to P-glycoprotein/ABCB1 [1] and MRP1 (multidrug novobiocin [13], appear to be ABCG2 specific (for a re-
resistance protein)/ABCC1 [2], breast cancer resistance cent review see [14]).
protein/ABCG2 has been shown to be involved [3–5]. Only limited information is available on the transporter
molecular mechanism, partially due to the lack of pu-
rified protein and three-dimensional structure, despite
some tentative extrapolations from extensive functional
*
†
Corresponding author.
These authors contributed equally to this work.

Cell. Mol. Life Sci. Vol. 63, 2006
Research Article
1913
data with P-glycoprotein and MRP1, and high-resolu- ate, rhodamine 123 and novobiocin were purchased from
tion structure of MsbA, a bacterial lipid exporter [15]. Sigma. Sodium orthovanadate was from Aldrich. Doxo-
Interestingly, a hot-spot mutation substituting arginine- rubicin, Hoechst33342 and azolectin were from Fluka.
4
82, especially by threonine or glycine, was shown to Pheophorbide a was purchased from Frontier Scientific
markedly change the spectrum for substrate drugs, by (Logan, USA). 6-Prenylchrysin and GF120918 were
abolishing methotrexate transport [16, 17] and, on the obtained as previously described [12], and Ko143 was
contrary, allowing transport of both rhodamine 123 and kindly provided by Dr. A.H. Schinkel (The Netherlands
doxorubicin [18–21], or increasing that of mitoxantrone Cancer Institute, Amsterdam, The Netherlands).
[
20, 21]. Transport of topotecan, SN-38, Hoechst33342
and pheophorbide a appeared much less affected [18, cDNA cloning and site-directed mutagenesis. The
0–22] (for a review, see [8]). A number of other muta- pcDNA3 plasmid containing R482T-ABCG2 cDNA, pro-
2
tions at the same position have shown that the positive vided by Dr. D. D. Ross, was used for subcloning into the
charge of arginine is not alone a decisive parameter [19, pTriex-4-Neo plasmid (Novagen, VWR, Fontenay-sous-
2
3, 24]. Interestingly, the mutation was also found to al- Bois, France). The R482T cDNA was mutated to obtain
ter the efficiency of some inhibitors, such as novobiocin R482 cDNA by site-directed mutagenesis using a Quick-
against BODIPY-prazosin and topotecan transport [21], Change Site-directed mutagenesis Kit (Stratagene, La
or GF120918 and 6-prenylchrysin against mitoxantrone Jolla). The primers used were: 5′-CTGTTATCTGATT-
transport [12]. In addition, the ATPase activity of in- TATTACCCATGAGGATGTTACCAAGTATTATATT-
sect cell membranes enriched with recombinant human TACC-3′ and 5′-GGTAAATATAATACTTGGTAACAT-
ABCG2 appeared to be dependent on mutation at posi- CCTCATGGGTAATAAATCAGATAACAG-3′.
tion 482 [12, 20, 25].
Since arginine-482 is assumed to be located at the in- Generation of recombinant baculoviruses. Recom-
ner edge of the third membrane span of ABCG2 trans- binant baculoviruses carrying either R482T- or R482-
membrane domain, it might occupy a strategic position ABCG2 human cDNA were generated with a BacVector
towards either binding or transport of substrate drugs, or transfection kit (Novagen), according to manufacturer’s
energy coupling. To decide between these alternatives instructions. Individual clones expressing the protein
and establish structure-function relationships, the aim were isolated by plaque staining with neutral red, and
of this study was to overexpress and purify both wild- the clone producing the recombinant protein at the high-
type (R482) and mutant (R482T) ABCG2 to determine est yield was selected by Western blot (see below). The
parameters ofATPase activity and measure direct binding selected recombinant baculoviruses were amplified with
of ligands, i.e. nucleotides, substrate drugs and inhibi- Sf9 insect cells.
tors. Changes in intrinsic fluorescence have been used to
monitor direct ligand interaction. Each ABCG2 mono- Protein overproduction and inverted membrane ves-
mer contains a total of five tryptophan residues, and the icle preparation. High Five insect cells were infected
intrinsic fluorescence of various ABC transporters have with recombinant baculoviruses at a multiplicity of 20
been found to be quite sensitive to quenching upon inter- for 2 h. After 3 days, the cells were harvested, and their
action with many ligands in both recombinant nucleotide- membranes isolated as described [33].
binding domains (mouse P-glycoprotein [26–28], human
MRP1 [29] or parasite Ltmdr1[30]) and full transporters SDS-PAGE and Western blot. Protein concentration
(
rodent P-glycoprotein [27, 31] or yeast Pdr5p [32]).
of inverted membrane vesicles was determined by the
The present results show that the R482T mutation repre- modified Lowry method with bovine serum albumin as
sents a gain-of-function for ATPase activity by increasing a standard. Samples were not heated before loading, and
both maximal rate of hydrolysis and affinity for MgATP. protein contents were analyzed by SDS-PAGE in 8% or
The interaction with any nucleotide, substrate drug or 10% polyacrylamide gels. The gels were either stained
inhibitor induces a quenching of intrinsic fluorescence, with Coomassie brilliant blue G-250, or used for protein
allowing the binding of the different ligands to be charac- electrotransfer (at 90 mA for 1 h) and Western blot onto
terized. The mutation does not alter the binding of either 0.45- μm pore nitrocellulose filters (Bio-Rad). Proteins
substrate drugs or inhibitors, suggesting a role of argi- of interest were detected by the anti-ABCG2 antibody
nine-482 in the transport of substrate drugs.
BXP-21 (1:1000 dilution, for 1 h at room temperature) in
accordance with manufacturer’s instructions (Alexis); the
secondary antibody was a goat anti-mouse IgG (H + L)-
alkaline phosphatase (AP) conjugate (Bio-Rad) used at
Materials and methods
1
:2300 dilution for 1 h at room temperature. Detection
Chemical compounds. ATP, ADP, 1,4-dithio-DL-threitol, was performed using the AP conjugate substrate kit (Bio-
protease inhibitors cocktail, mitoxantrone, methotrex- Rad) according to manufacturer’s instructions.

1914
A. Pozza et al.
Ligand binding to purified ABCG2
Protein purification. Inverted membrane vesicles of Malvern Instruments, Worcestershire, UK) was used to
High Five cells infected by baculovirus vector encoding monitor the absence of aggregation of purified ABCG2
the R482 or R482T transporter were treated with solu- in detergent solution, either before freezing or after freez-
bilization buffer {50 mM HEPES/NaOH, pH 8, 18 mM ing and thawing, taking into account the viscosity and
3
-[(3-cholamidopropyl) dimethylammonio]-1-propane- refraction index of the buffer. All steps were performed
sulfonate (CHAPS), 0.5 M NaCl, 10 mM imidazole, 20% in the presence of 1,4-dithio-DL-threitol (5 mM) and pro-
glycerol} at a final protein concentration of 2 mg/ml, for tease inhibitors (1 μl cocktail/ml buffer).
3
0 min with gently shaking at 4 °C. The sample was then
centrifuged at 15 000 g for 30 min at 4 °C. The super- ATPase activity measurements. The ATPase assay was
natant containing solubilized protein was immediately based on a colorimetric ascorbic acid/ammonium molyb-
applied onto a Ni-NTA agarose (Qiagen) resin, which date assay for measuring the release of inorganic phos-
was pre-equilibrated in the solubilization buffer, for 2 h phate as described [34] with minor modifications. The
at 4 °C with gently shaking. The Ni-NTA agarose resin reaction mixture contained 50 mM HEPES/NaOH, pH 8,
with bound protein was washed with ten column volumes 5 mM 1,4-dithio-DL-threitol, anATP-regenerating system
of wash buffer (50 mM HEPES/NaOH, pH 8, 18 mM (4 mM phosphoenolpyruvate, 60 μg/ml pyruvate kinase),
CHAPS, 0.5 M NaCl, 30 mM imidazole and 20% glyc- 0.5 μg of purified protein and 50 μg azolectin. The reac-
erol). The protein was eluted with two column volumes tion was started by addition of 5 mM ATP/20 mM MgCl2
of elution buffer (50 mM HEPES/NaOH, pH 8, 18 mM and proceeded for 40 min at 37 °C. Orthovanadate was
CHAPS, 0.5 M NaCl, 250 mM imidazole and 20% glyc- prepared as described [35]. Each experiment was per-
erol). Imidazole was removed by gel filtration (Econo-Pac formed two times in triplicate, and experimental data
1
0DG Columns Bio-Rad), and the purified protein was were analyzed with SigmaPlot.
quickly frozen and stored in liquid nitrogen. Protein con-
centration was determined before freezing with Bio-Rad Fluorescence quenching by nucleotides and drugs.
protein assay using bovine serum albumin as a standard. Experiments were performed at 25 °C using a Photon
A dynamic light scattering apparatus (Zetasizer Nano, Technology International Quanta Master I spectrofluori-
Figure 1. Purification of recombinant wild-type ABCG2. (a) Western blot revealed with anti-ABCG2 monoclonal antibody BXP-21. (b)
SDS-PAGE Coomassie blue-stained gel. Lane 1: prestained molecular weight markers; lane 2: inverted membrane vesicles of High Five
cells infected with baculovirus vector encoding R482 or R482T ABCG2; lane 3: supernatant after solubilization; lane 4: supernatant after
centrifugation (15000 g, 30 min); lane 5: detergent-insoluble pellet after centrifugation; lane 6: supernatant after binding for 2 h; lane 7:
Ni-NTA agarose gel after binding for 2 h; lane 8: washing; lane 9: Ni-NTA agarose gel after washing; lane 10: elution; lane 11: Ni-NTA
agarose gel after elution; lane 12: after imidazole removal by gel filtration. (c) Intrinsic fluorescence spectra. Buffer-corrected fluorescence
emission spectra for purified wild-type R482 (■) or mutant R482T (■) transporter at (0.045 mg protein/ml); the fluorescence emission
spectra were recorded at 25 °C upon excitation at 295 nm. Representative spectra are shown here, similar results being obtained in many
experiments from two different preparations of both wild-type and mutant transporters.

Cell. Mol. Life Sci. Vol. 63, 2006
Research Article
1915
meter with spectral bandwidths of 4 nm. All measure- Results
ments were corrected for buffer contribution to fluores-
cence, and inner-filter effects measured on N-acetyltryp- Overexpression of human ABCG2 in insect cells,
tophanamide. Purified protein (0.045 mg/ml into 50 mM solubilization and purification. The cDNA encoding
HEPES/NaOH, pH 8, 18 mM CHAPS, 0.5 M NaCl and the hexahistidine-tagged wild-type transporter was re-
20% glycerol) was excited at 295 nm and the emission combined with baculovirus, and used to infect High Five
spectra were scanned from 310 to 370 nm. The concentra- insect cells, and cell culture was harvested after 72 h
tion-dependent quenching studies were performed at least for membrane preparation. The fractions obtained from
three times, and the most representative curve was fitted solubilization with 1.1% CHAPS and purification with
with the Grafit program (Erithacus software). For nucleo- nickel-chelate affinity chromatography (Fig. 1) were ana-
tides that produced a monophasic quenching curve the ex- lyzed by SDS-PAGE, and followed by Western blot with
perimental data were fitted to the one-site ligand binding the specific BXP-21 monoclonal antibody (Fig. 1a) and
model, as previously described [36]. For substrate drugs Coomassie blue staining (Fig. 1b). The transporter ap-
and inhibitors that showed biphasic quenching curves, peared to be efficiently overexpressed in the membrane
the experimental data were fitted according to a two-site fraction (lane 2) and recovered upon detergent solubiliza-
model as described by Doppenschmitt et al. [37].
tion (lane 4). A good purification was achieved through
Figure 2. ATPase activity and nucleotide binding. (a) Lineweaver-Burk plots of ATP hydrolysis by the purified transporters. The ATPase
activity of purified R482 (■) or R482T (■) ABCG2 was measured with 0.5 μg protein in the presence of 50 μg azolectin at 37 °C for
4
0 min. (b) Inhibition by orthovanadate of ATPase activity. The inhibition of ATPase activity of R482 (■) and R482T transporter (■) in
the presence of 5 mM MgATP was measured at increasing orthovanadate concentrations (0.001–10 mM). (c, d) nucleotide binding as
monitored by quenching of the intrinsic tryptophan fluorescence of purified R482 (■) and R482T (■) transporter at 0.045 mg/ml: ATP (c)
2
+
and ADP/Mg (d). Fluorescence emission was recorded from 310 to 370 nm at 25 °C upon excitation at 295 nm.

1916
A. Pozza et al.
Ligand binding to purified ABCG2
Table 1. Binding parameters for quenching of ABCG2 intrinsic fluorescence by nucleotides and drugs. The concentration-dependent
saturable curves obtained in Figs 2–4 with wild-type and mutant purified ABCG2 were fitted by the Grafit program, according to a one-
site model for nucleotides [29], or a two-site model for substrate drugs and inhibitors [30]. Apparent dissociation constant and maximal
quenching values are indicated. For the two-site model, the program gave ΔF and ΔFmax. ΔF2 was then calculated by the difference between
1
ΔFmax and ΔF1.
Wild-type R482
Mutant R482T
KD1
ΔF1
KD2
ΔFmax
KD1
ΔF1
KD2
ΔFmax
(
μM)
(%)
(μM)
(%)
(μM)
(%)
(μM)
(%)
Nucleotides
ATP
MgADP
81.3 ± 10.4
82.0 ± 13.5
5.42 ± 0.19
6.42 ± 0.23
36.1 ± 7.4
149 ± 22
2.99 ± 0.14
11.1 ± 0.4
Substrate drugs
Methotrexate
Rhodamine 123 0.52 ± 0.10
Doxorubicin
Mitoxantrone
Pheophorbide a 0.94 ± 0.12
4.48 ± 0.40
13.7 ± 3.4
84.2 ± 25.6 66.3 ± 3.4
24.4 ± 1.2
22.2 ± 11.1 47.6 ± 3.4
7.30 ± 0.57
0.42 ± 0.09
2.91 ± 0.60
0.80 ± 0.10
1.50 ± 0.23
1.69 ± 0.40
15.7 ± 4.7
6.11 ± 1.23
7.32 ± 2.36
4.53 ± 0.10
6.77 ± 1.75
14.0 ± 4.1
87.0 ± 29.4 84.3 ± 4.7
5.66 ± 1.21 16.6 ± 6.6
12.4 ± 3.4
4.03 ± 0.42 20.3 ± 4.3
10.0 ± 1.3
12.8 ± 2.3
15.4 ± 6.6
25.9 ± 9.0
17.8 ± 4.6
21.7 ± 6.3
23.1 ± 9.6
23.9 ± 1.2
40.7 ± 2.4
25.5 ± 0.1
38.2 ± 1.8
66.0 ± 4.1
1.63 ± 0.45
0.65 ± 0.06
16.0 ± 0.4
40.0 ± 1.3
17.7 ± 4.2
Hoechst33342
2.04 ± 0.67
28.7 ± 19.6 32.2 ± 2.3
Inhibitors
GF120918
2.71 ± 0.66
12.6 ± 5.1
8.22 ± 2.39 26.8 ± 9.2
25.7 ± 10.9 72.4 ± 5.1
6.23 ± 1.46
0.57 ± 0.07
9.6 ± 5.2
11.0 ± 2.7
30.9 ± 11.7 70.9 ± 5.2
23.4 ± 6.1 69.0 ± 2.7
6
-prenylchrysin 0.80 ± 0.11
51.8 ± 2.4
Ko143
Novobiocin
0.012 ± 0.0016 13.2 ± 7.9
2.56 ± 0.56
0.14 ± 0.11 66.8 ± 7.9
4.45 ± 2.07 23.8 ± 10.8 27.6 ± 2.1
0.012 ± 0.0005 8.63 ± 3.63 0.096 ± 0.02 91.4 ± 3.6
1.49 ± 0.33 8.75 ± 2.87 21.1 ± 10.0 41.3 ± 2.9
nickel binding (lane 7), washing with 30 mM imidazole paring the IC50 values (Fig. 2b). Even without added
lane 9), elution by 250 mM imidazole (lane 10) and lipids, the purified recombinant protein was able to
(
removal of imidazole (lane 12). The typical amount of bind nucleotides, as monitored by quenching of intrin-
purified ABCG2 was 0.7 mg protein from 55 mg to- sic fluorescence. Saturable uni-site curves were obtained
tal membrane proteins, corresponding to a 1.3% yield, as a function of increasing concentrations for both ATP
2
as prepared from four 225-cm flasks of cultured insect (Fig. 2c) and ADP (Fig. 2d). Interestingly, the R482T
cells. A significantly lower yield was obtained in Sf9, as mutation increased the binding affinity for ATP (in the
compared with High Five, insect cells as due to lower absence of magnesium ions to avoid hydrolysis), with a
overexpression (not shown here). A similar procedure shift in apparent K from 81.3 ± 10.4 to 36.1 ± 7.4 μM,
D
was used to overexpress, solubilize and purify mutant whereas it lowered the binding affinity for MgADP, the
R482T ABCG2: qualitatively similar results were ob- hydrolysis product, with a shift in K from 82.0 ± 13.5 to
D
tained as in Fig. 1a and b, but with a twofold lower yield 149 ± 22 μM (cf. Table 1).
(
data not shown). Both wild-type and mutant transport-
ers exhibited a tryptophan-characteristic emission spec- Direct interaction of ABCG2 with substrate drugs.
trum of intrinsic fluorescence, with a maximal emission A number of compounds known to be transported by
wavelength around 333 nm upon excitation at 295 nm ABCG2 were found here to produce a saturable con-
(
Fig. 1c), indicating a rather hydrophobic average envi- centration-dependent quenching of intrinsic fluores-
ronment around the tryptophan residues. It is worthwhile cence of the purified transporter (Fig. 3). However,
mentioning that the R482T point mutation significantly curve fitting indicated two binding sites for each sub-
increased fluorescence intensity, by nearly 20%, giving strate drug. For methotrexate, the apparent KD1 and
rise to a more symmetrical emission spectrum.
KD2 values for wild-type ABCG2 were 4.48 ± 0.40 and
4.2 ± 25.6 μM, respectively (Fig. 3a, and Table 1). In-
8
ATPase activity and nucleotide binding. Both trans- terestingly, the R482T point mutation, reported to abol-
porters exhibited a high ATPase activity upon addi- ish methotrexate transport by whole cells and inverted
tion of lipids. The maximal rate of wild-type ABCG2 membrane vesicles, was found here to only slightly alter
was 357 ± 72 nmol ATP hydrolyzed/min/mg, with a KD1 (7.30 ± 0.57 μM), while KD2 remained unchanged
K (MgATP) of 2.0 ± 0.4 mM (Fig. 2a). Interestingly, (87.0 ± 29.4 μM), whereas maximal quenching was
m
the R482T point mutation both increased Vmax, up to higher for mutant as compared with wild-type ABCG2.
1
111 ± 79 nmol ATP hydrolyzed/min/mg, and lowered Conversely, although wild-type ABCG2 was known not
the K (MgATP) to 1.0 ± 0.1 mM. Concomitantly, the to transport rhodamine 123, in contrast to R482T mu-
m
strong inhibition by orthovanadate (80 ± 0.5%) became tant [8], displaying here apparent K and K values of
D1
D2
2
.5-fold less sensitive towards concentration when com- 0.52 ± 0.10 and 16.6 ± 6.6 μM, it in fact bound the fluo-

Cell. Mol. Life Sci. Vol. 63, 2006
Research Article
1917
Figure 3. Quenching of the intrinsic fluorescence of purified ABCG2 by substrate drugs and inhibitors. Fluorescence emission was re-
corded, as in Fig. 2c, d for R482 (■) and R482T (■) transporters, in the presence of increasing concentrations of either substrate drugs
such as methotrexate (a) and rhodamine 123 (b), or known inhibitors such as GF120918 (c) and 6-prenylchrysin (d).
rescent dye similarly (0.42 ± 0.09 and 15.4 ± 6.6 μM, re- cantly modify the intrinsic fluorescence of purified wild-
spectively) (Fig. 3b, and Table 1). Furthermore, doxorubi- type transporter, up to millimolar concentrations.
cin and mitoxantrone are known to be transported much
better by mutant as compared with wild-type transporter; Binding of known inhibitors. GF120918 and 6-pre-
the binding curves again showed that the apparent KD nylchrysin have been recently shown to lead to a better
values, 1.63 ± 0.45 and 22.2 ± 11.1 μM for doxorubicin, inhibition (22-fold and 12-fold, respectively) of the mi-
or 0.65 ± 0.06 and 20.3 ± 4.3 μM for mitoxantrone, were toxantrone efflux catalyzed by wild-type as compared
not markedly altered, if at all, by the mutation, leading to with mutant ABCG2-transfected HEK cells, as measured
the respective values of 2.91 ± 0.60 and 25.9 ± 9.0 μM for by flow cytometry [12]. In contrast, the present binding
doxorubicin and 0.80 ± 0.10 and 17.8 ± 4.6 μM for mi- experiments (Fig. 3c and d, Table 1) displayed only very
toxantrone (Table 1). Two other substrates were assayed, limited changes, if any: 2.71 ± 0.66 and 25.7 ± 10.9 μM
which bound similarly to wild-type and mutant purified as compared with 6.23 ± 1.46 and 30.9 ± 11.7 μM for
ABCG2: firstly, the natural substrate pheophorbide a with GF120918 on the one hand, 0.80 ± 0.11 and 26.8 ± 9.2
K values of 0.94 ± 0.12 and 17.7 ± 4.2 μM compared as compared with 0.57 ± 0.07 and 23.4 ± 6.1 for 6-pre-
D
with 1.50 ± 0.23 and 21.7 ± 6.3 μM and, secondly, the flu- nylchrysin on the other hand. Interestingly, a much
orescent dye Hoechst33342 with K values of 2.04 ± 0.67 higher binding affinity was observed for Ko143 (KD1
D
and 28.7 ± 19.6 μM compared with 1.69 ± 0.40 and 23.1 of 12 ± 1.6 nM or 12 ± 0.5 nM for wild-type or mutant
±
9.6 μM (Table 1). As a control, glutathione, a charged ABCG2, respectively, Table 1), which is known as a sec-
substrate of MRP1 but not of ABCG2, did not signifi- ond-generation highly potent inhibitor of both drug trans-

1918
A. Pozza et al.
Ligand binding to purified ABCG2
Figure 4. Prevention by inhibitors of mitoxantrone binding. Wild-type (a, c) or mutant (b, d) ABCG2 was preincubated for 30 min with
either 5 μM of GF120918 (a, b) or 3 μM 6-prenylchrysin (c, d) (open symbols) or without inhibitor (closed symbols), and the concentra-
tion-dependent binding of mitoxantrone was then measured under conditions of Fig. 3d.
port andATPase activity [11, 31]. Finally, the lower affin- 17.8 ± 4.6 μM (Fig. 4b). A comparable behavior was ob-
ity inhibitor novobiocin also bound similarly to wild-type served for 6-prenylchrysin with both wild-type ABCG2
and mutant transporter, with K values of 2.56 ± 0.56 (0.87 ± 0.23 and 28.6 ± 22.6 μM, Fig. 4c) and the mutant
D
and 23.8 ± 10.8 μM as compared with 1.49 ± 0.33 and transporter (0.74 ± 0.11 and 17.8 ± 8.8 μM, Fig. 4d).
2
1.1 ± 10.0 μM (Table 1).
Effects of preincubation with inhibitors on mitoxan- Discussion
trone binding. The differential inhibitory effects of
GF120918 and 6-prenylchrysin observed on mitoxan- This paper reports a first purification of human breast
trone efflux [12] might relate to either its binding, or its cancer resistance protein ABCG2, enabling measure-
transport, or both. Figure 4 clearly shows that preincuba- ment of direct interactions with substrates and inhibitors
tion of either wild-type or mutant ABCG2 with each in- by taking advantage of a hydrophobic tryptophan envi-
hibitor did not significantly modify further mitoxantrone ronment highly sensitive to ligand binding; this allowed
binding. Indeed, similar saturable quenching curves were a molecular characterization of the role of arginine-482
obtained for wild-type ABCG2 whether GF120918 was known to control the spectrum of transported drugs and
present (0.56 ± 0.08 and 17.3 ± 4.6 μM) or not (0.65 the related cellular multidrug resistance.
±
0.06 and 20.3 ± 4.3 μM) (Fig. 4a). Similar results
were obtained with the mutant transporter: 0.61 ± 0.07 Heterologous overexpression and purification of func-
and 19.0 ± 4.6 μM as compared with 0.80 ± 0.10 and tional wild-type or mutant ABCG2. The ability of

Cell. Mol. Life Sci. Vol. 63, 2006
Research Article
1919
baculovirus-infected insect cells to produce recombinant tated, and, therefore, might constitute a less limiting step
active human ABCG2 was clearly demonstrated by reten- in the ATPase reaction.
tion of drug-efflux activity of cells before lysis, as well as
ATP hydrolysis and transport capacity of inverted vesicles Drug interaction appears to be independent of the abil-
of enriched membranes [12, 24, 25]. It is shown here that ity to be transported and to induce cellular resistance.
reasonable amounts of purified ABCG2 can be obtained By contrast to uni-site binding curves for nucleotides, all
upon solubilization with CHAPS and one-step affinity known substrate drugs and inhibitors assayed here bound
chromatography, allowing biochemical and biophysical according to biphasic curves as a function of concentra-
studies. At least some of the five tryptophan residues tion.This could be fitted as either a one-site model with ad-
per monomer appear to be strategically located, within ditional nonspecific binding, or a two-site model [37]. The
a rather hydrophobic environment, since the interaction latter was used for P-glycoprotein for which similar bind-
with all assayed known substrates and inhibitors pro- ing curves were obtained, some of which with the same
duced a significant quenching of intrinsic fluorescence. ligands [31]. Our results shown in Table 1 were therefore
Such concentration-dependent modifications were quite analyzed according to a two-site model. Of course, a one-
convenient to compare the differential effects produced site model with additional nonspecific binding could not
by substrates and inhibitors, and the different properties be completely excluded; however, the absolute values of
of wild-type and mutant transporters. The R482T mutant KD2 were not of critical importance here, since the first
ABCG2 could be also prepared by the same procedure de- aim was to compare mutant to wild-type transporters, for
spite a twofold lower yield that might indicate some loss which nonspecific ligand binding, if any, should be similar.
in protein stability. The mutation seems to induce some Moreover, the first, high-affinity, phase of binding char-
local structural change as deduced from the differences acterized by KD1 appeared quite relevant to specific drug
observed in both the intensity of the emission fluores- binding and could be confidently used for studying the
cence spectrum (Fig. 1) and maximal quenching values mutation effects. Depending on the substrate, the higher
induced by various ligands (Figs. 2–4), whereas the ap- affinity constant, K , was in the range 0.4–7 μM, whereas
D1
parent dissociation constants were only slightly, if at all, the lower affinity one, K , was in the range 15–87 μM. A
D2
modified. The functionality of both recombinant purified similar two-site binding of most substrate drugs, including
transporters was assessed by (i) the high ATPase activity, rhodamine 123 and doxorubicin, was also reported with
sensitive to vanadate and dependent on the R482T muta- P-glycoprotein under similar conditions to those used here,
tion, (ii) the direct binding of nucleotides, in the presence contrasting with uni-site binding of nucleotides [31]. The
or absence of magnesium ions and their dependence on substrate displaying the lowest affinity was methotrexate
the mutation, (iii) the binding of most substrate drugs and (4.47 ± 0.40 and 82.4 ± 26.6 μM for wild-type purified
inhibitors in the micromolar range, consistent with their ABCG2), consistent with the relatively high concentrations
known cellular effects, and (iv) the very high-affinity required for measuring a vesicular transport [38] as com-
binding, in the nanomolar range, of Ko143 recognized as pared with other known ABCG2-transported drugs. These
the most efficient ABCG2 inhibitor.
two binding sites for methotrexate are consistent with the
positive cooperativity of transport recently reported [24].
The R482T hot-spot mutation as a gain-of-function Surprisingly, methotrexate bound nearly as efficiently to
for ATP hydrolysis. The high maximal rate of vana- purified mutantABCG2 (7.30 ± 0.57 and 87.0 ± 29.4 μM)
date-sensitive ATPase activity and millimolar affinity for despite the absence of cellular [16] or vesicular transport
MgATP of purified ABCG2 are consistent with results [17, 24] by the latter. However, other recent reports sug-
previously obtained with enriched insect cell membranes gested that methotrexate was also able to interact with mu-
[
12, 24, 25], and overexpression of the transporter in the tant ABCG2 from its capacity to inhibit both photoaffinity
range of 1–3% of total membrane proteins within these labeling within enriched membranes by a rhodamine 123
membranes. The mutation-induced threefold increase in derivative [39] and Hoechst33342 transport [40], and to
maximal rate of the purified transporter (up to 1.1 μmol induce a short-term resistance of cell growth [41].
ATP hydrolyzed/min/mg) is consistent with results of Conversely, rhodamine 123 and doxorubicin are known
ABCG2-enriched membranes with various mutations of to be effluxed and to induce cellular resistance in the case
arginine-482, demonstrating a gain-of-function for ATP of mutant, but not of wild-type, transporter [18, 19, 21].
hydrolysis capacity [8, 24]. It is further shown here that The present results show that the two drugs bound to puri-
the gain-of-function induced by the R482T mutation is fied wild-type ABCG2 and to the mutant transporter, in
also characterized by a twofold increase in affinity for agreement with photoaffinity labeling of both transporter
MgATP, concomitant with a better binding of substrate variants by an iodoaryl-azido derivative of rhodamine
ATP and a better release of product ADP. In addition, the 123, and prevention of the photolabeling by doxorubicin
lowered affinity for inhibitory vanadate suggests that, as [39]. In addition, although mitoxantrone is also known
for ADP, the release of product phosphate might be facili- to be better transported and to confer stronger resistance

1920
A. Pozza et al.
Ligand binding to purified ABCG2
to cells expressing mutant as compared with wild-type [43], or vinblastine, paclitaxel and Hoechst33342 sub-
ABCG2 [21], it was shown here to bind equally to both strate sites toward GF120918 and nicardipine regula-
purified transporters. The overall results strongly suggest tory sites [44]. An alternative explanation would be that
that all drugs that are able to bind to wild-type ABCG2 GF120918 and 6-prenylchrysin bind to another substrate
also bind similarly to the mutant transporter, whether they drug site, the saturation of which by a different substrate
are transported or not, and therefore whatever the residue does not antagonize mitoxantrone transport, as reported
at position 482.
in Xenopus laevis oocytes for rhodamine 123, topote-
can and daunorubicine against mitoxantrone, or for any
Mutation-independent binding of substrate drugs tested substrate against flavopiridol [45]. A similar lack
and inhibitors, suggesting a role of arginine-482 on of antagonism between several drug substrates was also
drug transport. In the case of Ko143, not only substrate reported for P-glycoprotein [46].
drugs but also inhibitors bound similarly to both variants In conclusion, the R482T mutation, which occurs in vitro
of purified ABCG2, with an extremely high affinity in on cell cultures submitted to high drug pressure, induces
the nanomolar range, consistent with its strong potency a gain-of-function for ATPase activity and for substrate-
to inhibit both drug transport and related cell growth re- drug transport, but not drug binding. The expected lo-
sistance, as well as ATPase activity [11, 42]. Although cation of arginine-482 within the third transmembrane
novobiocin [21], on the one hand, and GF120918 and 6- span of the ABCG2 transmembrane domain and the
prenylchrysin [12], on the other hand, have been shown to drastically different effects on various substrate drugs
inhibit the efflux of BODIPY-prazosin and mitoxantrone, suggest a direct role on drug transport. By contrast, the
respectively, much more efficiently, in cells transfected by stimulation of ATPase activity might be related to dis-
wild-type as compared with mutant ABCG2, they bound tant, indirect, effects mediated through conformational
here similarly to both variants of the purified transporter. changes possibly associated to energy coupling between
This clearly indicates that the differential inhibitory ef- ATP hydrolysis and drug transport. The absence of muta-
fects observed when comparing wild-type to mutant tion at this position in vivo, despite a number of reported
ABCG2-transfected cells were not due to differences in polymorphism examples [47], might indicate deleterious
affinity of inhibitor binding, but more likely to the dif- effects produced by altered transport selectivity of natural
ferences in substrate drug transport, as discussed above. substrates. Such a critical role played by arginine-482 in
Such a lack of effect of a single point-mutation on drug transport control is consistent with the residue conserva-
binding is not so surprising for a multispecific transporter tion during evolution.
involved in cell detoxification. It was previously charac-
Acknowledgements. We warmly thank Drs. A. H. Schinkel (Am-
terized as having a wide spectrum for drug binding, as
sterdam, The Netherlands) for providing Ko143, A. Boumendjel
shown by the ability of many compounds to antagonize
(
Grenoble, France) for synthesizing 6-prenylchrysin, O. Coulom-
photoaffinity labeling of the rhodamine 123 binding site bel and R. Haser (Lyon, France) for giving access to refractometer
and viscosimeter, and to a dynamic-light scattering apparatus, re-
[
39] within enriched membranes.The K values obtained
D1
spectively. This work was supported by grants from the Association
pour la Recherche sur le Cancer (ARC 3519), the Ligue Nationale
contre le Cancer (comités de Loire, Drôme et Savoie), the Région
here for inhibitors correlate well with previously reported
IC values, e.g. in the 5–10 nM range for Ko143 towards
5
0
ATPase activity, Hoechst33342 transport and sensitiza- Rhône-Alpes (thématique prioritaire Cancer) and French-Spanish
PAI-Picasso interministerial agreements (SAF-2003-04200-CO2-
tion to mitoxantrone [42], or the submicromolar range
for 6-prenylchrysin toward mitoxantrone transport and
sensitization to mitoxantrone [12].The micromolar range
0
1). A.P. is a recipient of a doctoral fellowship from the Ligue Na-
tionale contre le Cancer (comité du Rhône), and A.A.-B. from the
Association pour la Recherche sur le Cancer. A.S. is a recipient
obtained for novobiocin binding confirms problems of of a Leonardo da Vinci Program fellowship (Unipharma-Graduates
Project), and J.M.P.-V. of a Marie-Curie post-doctoral fellowship
from the European Commission (HMPF-CF-2001-01244).
cellular accessibility [21] likely due to low membrane
penetration [13]. Comparable binding for substrates and
inhibitors were obtained with P-glycoprotein [31], with
1
Juliano, R. L. and Ling, V. (1976)A surface glycoprotein modu-
lating drug permeability in Chinese hamster ovary cell mutants.
Biochim. Biophys. Acta 455, 152–162.
monophasic binding curves for nucleotides and biphasic
ones for most substrate drugs and inhibitors. Rhodamine
also bound in the micromolar range, as did a number of
other drugs.
The present lack of prevention of mitoxantrone binding to
purified ABCG2 by preincubation with either GF120918
or 6-prenylchrysin suggests essentially distinct binding
sites for substrate drug and inhibitors. This was indeed
also proposed in the case of P-glycoprotein, by consid-
ering either rhodamine 123 and Hoechst33342 substrate
sites towards prazosin and progesterone regulatory sites
2 Cole, S. P., Bhardwaj, G., Gerlach, J. H., Mackie, J. E., Grant,
C. E., Almquist, K. C., Stewart, A. J., Kurz, E. U., Duncan, A.
M. and Deeley, R. G. (1992) Overexpression of a transporter
gene in a multidrug-resistant human lung cancer cell line. Sci-
ence 258, 1650–1654.
3 Allikmets, R., Schriml, L. M., Hutchinson, A., Romano-Spica,
V. and Dean, M. (1998) A human placenta-specific ATP-bind-
ing cassette gene (ABCP) on chromosome 4q22 that is involved
in multidrug resistance. Cancer Res. 58, 5337–5339.
4
Doyle, L. A., Yang, W., Abruzzo, L. V., Krogmann, T., Gao,
Y., Rishi, A. K. and Ross, D. D. (1998) A multidrug resistance

Cell. Mol. Life Sci. Vol. 63, 2006
Research Article
1921
transporter from human MCF-7 breast cancer cells. Proc. Natl. 20 Ozvegy, C., Varadi, A. and Sarkadi, B. (2002) Characteriza-
Acad. Sci. USA 95, 15665–15670.
tion of drug transport, ATP hydrolysis, and nucleotide trapping
by the human ABCG2 multidrug transporter. Modulation of
substrate specificity by a point mutation. J. Biol. Chem. 277,
47980–47990.
5
Miyake, K., Mickley, L., Litman, T., Zhan, Z., Robey, R., Cris-
tensen, B., Brangi, M., Greenberger, L., Dean, M., Fojo, T. and
Bates, S. E. (1999) Molecular cloning of cDNAs which are highly
overexpressed in mitoxantrone-resistant cells: demonstration of
homology to ABC transport genes. Cancer Res. 59, 8–13.
Bates, S. E., Robey, R., Miyake, K., Roa, K., Ross, D. D. and
Litman, T. (2001) The role of half-transporters in multidrug re-
sistance. J. Bioenerg. Biomembr. 33, 503–511.
21 Robey, R. W., Honjo, Y., Morisaki, K., Nadjem, T. A., Runge,
S., Risbood, M., Poruchynsky, M. S. and Bates, S. E. (2003)
Mutations at amino-acid 482 in the ABCG2 gene affect sub-
strate and antagonist specificity. Br. J. Cancer 89, 1971–1978.
22 Robey, R. W., Steadman, K., Polgar, O., Morisaki, K., Blayney,
M., Mistry, P. and Bates, S. E. (2004) Pheophorbide a is a spe-
cific probe forABCG2 function and inhibition. Cancer Res. 64,
1242–1246.
6
7
8
9
Doyle, L. A. and Ross, D. D. (2003) Multidrug resistance medi-
ated by the breast cancer resistance protein BCRP (ABCG2).
Oncogene 22, 7340–7358.
Sarkadi, B., Ozvegy-Laczka, C., Nemet, K. and Varadi, A.
23 Miwa, M., Tsukahara, S., Ishikawa, E., Asada, S., Imai, Y. and
Sugimoto, Y. (2003) Single amino acid substitutions in the
transmembrane domains of breast cancer resistance protein
(BCRP) alter cross resistance patterns in transfectants. Int. J.
Cancer 107, 757–763.
24 Ozvegy-Laczka, C., Koblos, G., Sarkadi, B. and Varadi, A.
(2005) Single amino acid (482) variants of the ABCG2 mul-
tidrug transporter: major differences in transport capacity and
substrate recognition. Biochim. Biophys. Acta 1668, 53–63.
(
1
2004) ABCG2 – a transporter for all seasons
16–120.
.
FEBS Lett. 567,
De Bruin, M., Miyake, K., Litman, T., Robey, R. and Bates,
S. E. (1999) Reversal of resistance by GF120918 in cell lines
expressing the ABC half-transporter, MXR. Cancer Lett. 146,
1
17–126.
1
1
0 Rabindran, S. K., Ross, D. D, Doyle, L. A.,Yang, W. and Green-
berger, L. M. (2000) Fumitremorgin C reverses multidrug re-
sistance in cells transfected with the breast cancer resistance 25 Ozvegy, C., Litman, T., Szakacs, G., Nagy, Z., Bates, S. E.,
protein. Cancer Res. 60, 47–50.
Varadi, A. and Sarkadi, B. (2001) Functional characterization
of the human multidrug transporter, ABCG2, expressed in in-
sect cells. Biochem. Biophys. Res. Commun. 285, 111–117.
26 Baubichon-Cortay, H., Baggetto, L. G., Dayan, G. and Di Pietro,
A. (1994) Overexpression and purification of the carboxyl-ter-
minal nucleotide-binding domain from mouse P-glycoprotein.
Strategic location of a tryptophan residue. J. Biol. Chem. 269,
22983–22989.
1 Allen, J. D., van Loevezijn, A., Lakhai, J. M., van der Valk, M.,
van Tellingen, O., Reid, G., Schellens, J. H., Koomen, G. J.
and Schinkel, A. H. (2002) Potent and specific inhibition of the
breast cancer resistance protein multidrug transporter in vitro
and in mouse intestine by a novel analogue of fumitremorgin C.
Mol. Cancer Ther. 1, 417–425.
1
1
1
2 Ahmed-Belkacem, A., Pozza, A., Munoz-Martinez, F., Bates,
S. E., Castanys, S., Gamarro, F., Di Pietro, A. and Perez-Vic-
toria, J. M. (2005) Flavonoid structure-activity studies identify
27 Dayan, G., Jault, J. -M., Baubichon-Cortay, H., Baggetto, L. G.,
Renoir, J. M., Baulieu, E. E, Gros, P. and Di Pietro, A. (1997)
Binding of steroid modulators to recombinant cytosolic domain
from mouse P-glycoprotein in close proximity to the ATP site.
Biochemistry 36, 15208–15215.
6
-prenylchrysin and tectochrysin as potent and specific inhibi-
tors of breast cancer resistance protein ABCG2. Cancer Res.
5, 4852–4860.
6
3 Shiozawa, K., Oka, M., Soda, H., Yoshikawa, M., Ikegami, Y.,
Tsurutani, J., Nakatomi, K., Nakamura, Y., Doi, S., Kitazaki,
T., Mizuta,Y., Murase, K.,Yoshida, H., Ross, D. D. and Kohno,
S. (2004) Reversal of breast cancer resistance protein (BCRP/
ABCG2)-mediated drug resistance by novobiocin, a coumer-
mycin antibiotic. Int. J. Cancer 108, 146–151.
4 Ahmed-Belkacem, A., Pozza, A., Macalou, S., Perez-Victo-
ria, J. M., Boumendjel, A. and Di Pietro, A. (2006) Inhibitors
of cancer cell multidrug resistance mediated by breast cancer
28 Conseil, G., Baubichon-Cortay, H., Dayan, G., Jault, J. -M. and
Di Pietro, A. (1998) Flavonoids: a class of modulators with
bifunctional interactions at vicinal ATP- and steroid- binding
sites on mouse P-glycoprotein. Proc. Natl. Acad. Sci. USA 95,
9831–9836.
29 Trompier, D., Baubichon-Cortay, H., Chang, X. -B., Maitrejean,
M., Barron, D., Riordan, J. R. and Di Pietro, A. (2003) Multi-
ple flavonoid-binding sites within multidrug resistance protein
MRP1. Cell. Mol. Life Sci. 60, 2164–2177.
resistance protein (BCRP/ABCG2). Anti-Cancer Drugs 17, 30 Perez-Victoria, J. M., Chiquero, M. J., Conseil, G., Dayan, G.,
2
39–243.
Di Pietro, A., Barron, D., Castanys, S. and Gamarro, F. (1999)
Correlation between the affinity of flavonoids binding to the
cytosolic site of Leishmania tropica multidrug transporter and
their efficiency to revert parasite resistance to daunomycin.
Biochemistry 38, 1736–1743.
1
1
5 Reyes, C. L., Ward, A., Yu, J. and Chang, G. (2006) The struc-
tures of MsbA: Insight into ABC transporter-mediated multi-
drug efflux. FEBS Lett. 580, 1042–1048.
6 Volk, E. L., Farley, K. M., Wu, Y., Robey, R. W. and Schneider,
E. (2002) Overexpression of wild-type breast cancer resistance 31 Liu, R., Siemiarczuk, A. and Sharom, F. J. (2000) Intrinsic fluo-
protein mediates methotrexate resistance. Cancer Res. 62,
035–5040.
7 Chen, Z. S., Robey, R. W., Belinsky, M. G., Shchaveleva, I.,
Ren, X. Q., Sugimoto, Y., Ross, D. D., Bates, S. E. and Kruh,
rescence of the P-glycoprotein multidrug transporter: sensitiv-
ity of tryptophan residues to binding of drugs and nucleotides.
Biochemistry 39, 14927–14938.
5
1
32 Conseil, G., Decottignies, A., Jault, J. -M., Comte, G., Barron,
D., Goffeau, A. and Di Pietro, A. (2000) Prenyl-flavonoids as
potent inhibitors of the Pdr5p multidrug ABC transporter from
Saccharomyces cerevisia. Biochemistry 39, 6910–6917.
33 Dey, S., Ramachandra, M., Pastan, I., Gottesman, M. M. and
Ambudkar, S. V. (1998) Photoaffinity labeling of human P-gly-
coprotein: effect of modulator interaction and ATP hydrolysis
on substrate binding. Methods Enzymol. 292, 318–328.
34 Doige, C. A., Yu, X. and Sharom, F. J. (1992) ATPase activity
of partially purified P-glycoprotein from multidrug-resistant
Chinese hamster ovary cells. Biochim. Biophys. Acta 1109,
149–160.
G. D
(2003) Transport of methotrexate, methotrexate poly-
.
glutamates, and 17beta-estradiol 17-(beta-D-glucuronide) by
ABCG2: effects of acquired mutations at R482 on methotrex-
ate transport. Cancer Res. 63, 4048–4054.
1
1
8 Honjo, Y., Hrycyna, C. A., Yan, Q. -W., Medina-Perez, W. Y.,
Robey, R. W., van de Laar, A., Litman, T., Dean, M. and Bates,
S. E (2001) Acquired mutations in the MXR/BCRP/ABCP
.
gene alter substrate specificity in MXR/BCRP/ABCP-overex-
pressing cells. Cancer Res. 61, 6635–6639.
9 Allen, J. D., Jackson, S. C. and Schinkel, A. H. (2002) A mu-
tation hot spot in the Bcrp1 (Abcg2) multidrug transporter in
mouse cell lines selected for Doxorubicin resistance. Cancer
Res. 62, 2294–2299.
35 Hou, Y., Cui, L., Riordan, J. R. and Chang, X. -B. (2000) Al-
losteric interactions between the two non-equivalent nucleotide

1922
A. Pozza et al.
Ligand binding to purified ABCG2
binding domains of multidrug resistance protein MRP1. J. Biol.
Chem. 275, 20280–20287.
resistance to various hydrophilic antifolates. Cancer Res. 65,
8414–8422.
3
3
6 Divita, G., Goody, R. S., Gautheron, D. C. and Di Pietro, A.
42 Ozvegy-Laczka, C., Hegedus, T., Varady, G., Ujhelly, O.,
Schuetz, J. D., Varadi, A., Keri, G., Orfi, L., Nemet, K. and
Sarkadi, B. (2004) High-affinity interaction of tyrosine kinase
inhibitors with the ABCG2 multidrug transporter. Mol. Phar-
macol. 65, 1485–1495.
43 Shapiro, A. B., Fox, K., Lam, P. and Ling, V. (1999) Stimu-
lation of P-glycoprotein-mediated drug transport by prazosin
and progesterone. Evidence for a third drug-binding site. Eur.
J. Biochem. 259, 841–850.
(1993) Structural mapping of catalytic site with respect to al-
pha-subunit and noncatalytic site in yeast mitochondrial F1-
ATPase using fluorescence resonance energy transfer. J. Biol.
Chem. 268, 13178–13186.
7 Doppenschmitt, S., Spahn-Langguth, H., Regardh, C. G. and
Langguth, P. (1998) Radioligand-binding assay employing P-
glycoprotein-overexpressing cells: testing drug affinities to
the secretory intestinal multidrug transporter. Pharm. Res. 15,
1
001–1006.
44 Martin, C., Berridge, G., Higgins, C. F., Mistry, P., Charlton,
P. and Callaghan, R. (2000) Communication between multiple
drug binding sites on P-glycoprotein. Mol. Pharmacol. 58,
624–632.
45 Nakanishi, T., Doyle, L. A., Hassel, B., Wei, Y., Bauer, K. S.,
Wu, S., Pumplin, D. W., Fang, H. -B. and Ross, D. D. (2003)
Functional characterization of human breast cancer resistance
protein (BCRP, ABCG2) expressed in the oocytes of Xenopus
laevis. Mol. Pharmacol. 64, 1452–1462.
46 Barecki-Roach, M., Wang, E. -J. and Johnson, W. W. (2003)
Many P-glycoprotein substrates do not inhibit the transport
process across cell membranes. Xenobiotica 33, 131–140.
47 Cervenak, J., Andrikovics, H., Ozvegy-Lacska, C., Tordai, A.,
Nemet, K., Varadi, A. and Sarkadi, B (2006) The role of the
human ABCG2 multidrug transporter and its variants in cancer
therapy and toxicology. Cancer Lett. 234, 62–72.
3
3
8 Volk, E. L. and Schneider, E. (2003) Wild-type breast cancer
resistance protein (BCRP/ABCG2) is a methotrexate polyglu-
tamate transporter. Cancer Res. 63, 5538–5543.
9 Alqawi, O., Bates, S. and Georges, E. (2004) Arginine482
to threonine mutation in the breast cancer resistance protein
ABCG2 inhibits rhodamine 123 transport while increasing
binding. Biochem. J. 382, 711–716.
0 Janvilisri, T., Shahi, S., Venter, H., Balakrishnan, L. and van
Veen, H. W. (2005) Arginine-482 is not essential for transport
of antibiotics, primary bile acids and unconjugated sterols by
the human breast cancer resistance protein (ABCG2). Biochem.
J. 385, 419–426.
1 Shafran, A., Ifergan, I., Bram, E., Jansen, G., Kathmann, I., Pe-
ters, G. J., Robey, R. W., Bates, S. E. and Assaraf, Y. G. (2005)
ABCG2 harboring the Gly482 mutation confers high-level
4
4