- Email: [email protected]
ABCG2 confers resistance to indolocarbazole compounds by ATP-dependent transport
BBRC Biochemical and Biophysical Research Communications 299 (2002) 669–675 www.academicpress.com
ABCG2 confers resistance to indolocarbazole compounds by ATP-dependent transportq Rinako Nakagawa, Yoshikazu Hara, Hiroharu Arakawa, Susumu Nishimura, and Hideya Komatani* Banyu Tsukuba Research Institute in Collaboration with Merck Research Laboratories, 3 Okubo, Ibaraki 300-2611, Japan Received 1 November 2002
Abstract The ABC half-transporter, ABCG2, is known to confer resistance to chemotherapeutic agents including indolocarbazole derivatives. MCF7 cells were introduced by either wild type ABCG2 (ABCG2-482R) or mutant ABCG2 (-482T), whose amino acid at position 482 is substituted to threonine from arginine, and their cross-resistance pattern was analyzed. Although this amino acid substitution seems to affect cross-resistance patterns, both 482T- and 482R-transfectants showed strong resistance to indolocarbazoles, confirming that ABCG2 confers resistance to them. For further characterization of ABCG2-mediated transport, we investigated indolocarbazole compound A (Fig. 1) excretion in cell-free system. Compound A was actively transported in membrane vesicles prepared from one of the 482T-transfectants and its uptake was supported by hydrolysis of various nucleoside triphosphates. This transport was inhibited completely by the other indolocarbazole compound, but not by mitoxantrone, implying that the binding site of mitoxantrone or the transport mechanisms for mitoxantrone is different from those of indolocarbazoles. These results showed that ABCG2 confers resistance to indolocarbazoles by transporting them in an energy-dependent manner. Ó 2002 Elsevier Science (USA). All rights reserved. Keywords: ABC transporter; ABCG2; MXR; BCRP; Drug resistance; Indolocarbazole; Vesicle transport
The development of multidrug resistance in tumor cells is frequently associated with overexpression of the ABC transporter family. Well-characterized ABC transporter members, such as P-gp (ABCB1) and MRP1 (ABCC1), have been shown to be involved in excretion of certain anticancer drugs by active transport [1]. The second member of the ATP-binding cassette transporter G subfamily (ABCG2), which is also denoted as BCRP/ABCP/MXR, is also capable of conferring drug resistance. Unlike other multidrug ABC transporters, ABCG2 is a half-transporter consisting of six putative transmembrane domains and a single
q
Abbreviations: ABC, ATP-binding cassette; BCRP, breast cancer resistance protein; MXR, mitoxantrone resistance-associated protein; ABCP, placenta specific ABC transporter; P-gp, P-glycoprotein; MRP, multidrug resistance-associated protein; SN-38, 7-ethyl-10hydroxycamptothecin; ATPcS, adenosine 50 -O-(3-thiotriphosphate). * Corresponding author. Fax: +81-298-77-2027. E-mail address: [email protected] (H. Komatani).
ATP-binding domain [2–4]. The ABCG subfamily contains the Drosophila white/brown/scarlet genes, which are known to transport eye pigment, and the human homologue of the Drosophila white gene, ABCG1, which was recently proved to be involved with lipophilic xenobiotics [5]. The physiological role of ABCG2 is not yet understood. However, high expression of ABCG2 was observed in the placenta, colon, and small intestine, suggesting protective functions on the fetus and gastrointestinal tract [6,7]. Extensive studies using drug-selected cell lines and cells with enforced ABCG2 expression demonstrated that this protein is localized in the plasma membrane [8,9] and confers high levels of resistance to multiple chemotherapeutic drugs, including mitoxantrone [2], anthracyclines [2], the camptothecins (i.e., topotecan and active form of irinotecan, SN-38) [10,11], and indolocarbazole analogues [12]. Originally this gene was cloned by different groups with a different amino acid at codon 482 [2–4]. We recently proposed that the
0006-291X/02/$ - see front matter Ó 2002 Elsevier Science (USA). All rights reserved. PII: S 0 0 0 6 - 2 9 1 X ( 0 2 ) 0 2 7 1 2 - 2
670
R. Nakagawa et al. / Biochemical and Biophysical Research Communications 299 (2002) 669–675
gene carrying arginine at this position is the naturally expressed form, whereas the genes holding threonine or glycine are mutated forms [12]. It has also been proved that this amino acid substitution alters the cross-resistance pattern to rhodamine 123 and adriamycin [13]. To date, Nakatomi et al. [14] reported that ABCG2 functions as an ATP-dependent efflux pump for SN-38 using plasma membrane vesicles obtained from SN-38 selected cells. However, ABCG2-mediated export systems have not been characterized in detail and are still largely unknown. In the present study, we examined drug resistance phenotype of ABCG2-482T and -482R introduced MCF7 cells. Although the amino acid substitution seems to affect cross-resistance patterns, both types of cells are highly resistant to indolocarbazoles, confirming that ABCG2 confers resistance to indolocarbazole compounds. In order to characterize mechanisms of resistance to indolocarbazoles, we investigated transport activities of ABCG2 in a plasma membrane vesicle from one of the transfectants. Here we provide the direct evidence about ATP-dependent export of the indolocarbazole, by which ABCG2 confers great resistance to the compound.
Materials and methods Reagents. Indolocarbazole compound A (Fig. 1), indolocarbazole compound B, and topotecan were synthesized in our institute as previously described [12]. 14 C-labeled compound A was from Daiichi Pure Chemical (Tokyo, Japan). Mitoxantrone, camptothecin, ADP, ATPcS, and creatine phosphokinase were purchased from Sigma Chemical (St. Louis, MO). ATP, GTP, CTP, and UTP were obtained from Amersham Pharmacia Biotech (Piscataway, NJ). Creatine phosphate was from Wako Pure Chemicals (Osaka, Japan). ABCG2 transfectants. MCF7 human breast carcinoma cells were transfected with pcDNA3.1/V5-His-TOPO containing full-length ABCG2 cDNA (482T variant) [2] or pcDNA3.1/V5-His-TOPO using
Effectene Transfection Reagent (Qiagen, Valencia, CA), selected by culture with geneticin (G418, 0.9 mg/ml), and isolated with cloning rings. Clones with high-level expression of ABCG2 were selected by Northern blots. All cell lines were maintained in RPMI 1640 medium supplemented with 10% FCS, 100 lg/ml gentamicin, and selected concentrations of G418. Northern blotting. Total RNA was extracted from ABCG2-transfectants using an RNeasy Kit (Qiagen). Samples of 8 lg total RNA were separated on 1% agarose gel and transferred to nylon membranes. The membranes were probed with a 32 P-labeled 269-bp fragment of human ABCG2 cDNA (161–429 of AF103796). Cytotoxicity assay. Samples of 1500 cells/well were plated one day before adding a drug and grown for a further 3 days with various concentrations of the drugs. The sensitivity of MCF7 transfectants to chemotherapeutic agents was determined by sulforhodamine-B cytotoxicity assay. DMSO-treated cells in control wells were assigned a value of 100% and IC50 values were defined as the dose required to inhibit growth of cells to 50% of that of the controls. Membrane vesicle preparation. Plasma membrane vesicles were prepared as described previously [15]. Briefly, cells (3:0 108 ) were harvested by scraping and washed once in ice-cold phosphate-buffered saline (PBS). The cell pellet was homogenized in buffer containing 0.5 mM Na-phosphate (pH 7.0), 0.1 mM EDTA, and protease inhibitors (Roche, Mannheim, Germany) using a Potter–Elvehjem homogenizer and then the cell lysate was centrifuged at 100,000g for 30 min. The crude membrane fraction was layered over 38% sucrose solution and centrifuged at 100,000g for 30 min. The interface was collected, suspended in 0.25 M sucrose containing 10 mM Tris–HCl (pH 7.4), and centrifuged at 100,000g for 20 min. The membrane pellet was resuspended in transport buffer [0.25 M sucrose containing 10 mM Tris– HCl (pH 7.4)], frozen in liquid N2 and stored at )80 °C until use. Western blotting. Samples of 5.5 lg cellular membrane proteins were loaded onto 7.5% polyacrylamide gel and subjected to electrophoresis. Proteins were transferred to a PVDF membrane (Millipore, Bedford, MA). The membrane was blocked with Tris-buffered saline (pH 7.6) containing 0.1% (v/v) Tween 20 and 2% Block Ace (Dainippon Seiyaku, Osaka, Japan). The membranes were probed for 1.5 h with 1:700 dilution of a polyclonal anti-ABCG2 antibody, which we raised against a 15-mer peptide of ABCG2:NREEDFKATEIIEPS, followed by a secondary horseradish peroxidase-conjugated donkey anti-rabbit antibody (Amersham) (1:10,000 dilution) for 1 h. Detection was enhanced by chemiluminescence detection (Amersham) and immunoreactive bands were visualized. Vesicle transport studies. Thawed membranes were diluted in the transport buffer: 0.25 M sucrose, 10 mM Tris–HCl (pH 7.4), and passed 20-times through a 27-gauge needle using a syringe for vesicle formation. Protein concentration was measured by Bradford assay. Standard transport assays were performed at 37 °C in a 50 ll volume of mixture consisting of 20 lg vesicle protein, 30 lM [14 C]compound A (3:4 103 lCi/reaction), 10 mM MgCl2 , 4 mM ATP, and an ATP regenerating system: 10 mM creatine phosphate and 100 lg/ml creatine kinase in transport buffer. The mixtures were incubated at 37 °C for 5 min. The reactions were terminated by adding 1 ml ice-cold transport buffer and then the mixtures were filtered through glass fiber (type GF/C) filters (Millipore) using a rapid filtration device with light suction and rinsed four times with 1 ml ice-cold transport buffer. The filters were then dissolved and radioactivity was determined. Data on ATP-dependent transport were calculated by subtracting values in the absence of ATP from those in the presence of ATP [16].
Results Fig. 1. Structures of indolocarbazole compound A: 6-N-formylamino12,13-dihydro-1,11-dihydroxy-13-(b-D -glucopyranosil)5H-indolo[2,3-a] pyrrolo [3,4-c]carbazole-5,7(6H)-dione.
We introduced an expression vector containing fulllength human ABCG2 cDNA (482R and 482T variants)
R. Nakagawa et al. / Biochemical and Biophysical Research Communications 299 (2002) 669–675
671
Fig. 2. (A) Northern blot of MCF7 cells transfected with vector, ABCG2-482R- or ABCG2-482T-variant (8 lg total RNA/lane). (B) Western blot with polyclonal anti-ABCG2 antibody of membranes (5.5 lg protein/lane) from MCF7/T#8 or MCF7/pc#1. (C) Time courses of indolocarbazole compound A uptake in membranes from MCF7/T#8 and MCF7/pc#1 cells: membrane vesicles (20 lg protein/reaction) from T#8 (j,) and pc#1 (d,s) were incubated with 30 lM [14 C]compound A (3:4 103 lCi/reaction) in transport buffer [0.25 M sucrose, 10 mM Tris–HCl (pH 7.4)] with an ATP-regenerating system (10 mM creatine phosphate and 100 lg/ml creatine kinase). Closed symbols represent uptake in the presence of 4 mM ATP; open symbols represent uptake in the absence of ATP. Values are means SEM in three separate experiments.
into MCF7, which was reported to be highly resistant to mitoxantrone upon introduction of 482T [2]. ABCG2 mRNA expression was analyzed by Northern blots in ABCG2- or vector-transfected cells (Fig. 2A). All the ABCG2-introduced clones expressed prominent amount of ABCG2 mRNA with a similar level. The mRNA of ABCG2 was not detected in a vector-transrfected cell (MCF7/pc#1). Table 1 shows the relative resistance value to various anti-cancer drugs. ABCG2-transfected cells displayed pronounced resistance to indolocarbazole compound A (26–298-fold resistance) as PC-13 cells transfected with ABCG2-482R did in an earlier study [12], confirming that ABCG2 is responsible for indolocarbazole resistance. MCF7 cells expressing 482R (MCF7/R#7 and R#59) showed no resistance to adriamycin, but showed mild resistance to mitoxantrone and topotecan (7.6–8.8-fold, 5.1–6.6-fold, respectively) and the strongest resistance to indolocarbazole compound A (106–298-fold), thereby displaying a rather indolocar-
bazole-specific resistance phenotype. On the other hand, MCF7 clones expressing 482T (MCF7/T#8 and T#79) showed modest resistance to adriamycin and topotecan (3.8–5.5-fold, 2.7–4.7-fold, respectively), strong resistance to mitoxantrone (16–40-fold) and strongest resistance, but in less magnitude than 482R-transfectants, to compound A (26–124-fold). Though there are differences in resistance magnitude between the clones with similar transcript level of 482R or 482T, these results are concordant with the previous finding [13] that alteration of amino acid 482 affects the cross-resistance patterns of ABCG2. The differences of resistance might be due to what expression level of the protein or stability of the protein is different between the clones. Transfection of both 482R- and 482T-transfectants was found to have only a modest effect on resistance to topotecan and not to affect sensitivity to camptothecin. The MCF7/T#8 was chosen out of the transfectants to investigate the mechanism of drug transport associated
Table 1 Cross-resistance profiles of ABCG2-transfected cells Drug
MCF7/pc#1 IC50 (lM)
MCF7/T#8 RRa
MCF7/T#79 RRa
MCF7/R#7 RRa
MCF7/R#59 RRa
Compound A Adriamycin Camptothecin Topotecan Mitoxantrone
0.84 0.28 0.024 0.0064 0.012 0.0018 0.018 0.0038 0.0031 0.00071
124 5.5 1.3 4.7 40
26 3.8 1.1 2.7 16
298 1.4 1.4 6.6 8.8
106 1.4 1.8 5.1 7.6
Note. IC50 values are shown as means SEM of more than five separate experiments. a RR, relative resistance was calculated by dividing the IC50 values of the MCF7 clones by those of MCF7/pc#1 cells.
672
R. Nakagawa et al. / Biochemical and Biophysical Research Communications 299 (2002) 669–675
with resistance to indolocarbazole. Western blot analysis of membrane protein from this clone showed that the cells expressed a large amount of ABCG2 protein (Fig. 2B). The plasma membrane localization of ABCG2 in MCF7/T#8 cells was confirmed by immunocytochemical analysis with a monoclonal antibody (data not shown). The time course and ATP-dependence of compound A accumulation by vesicles prepared from MCF7/T#8 cells or MCF7/pc#1 cells are shown in Fig. 2C. When membrane vesicles were incubated with 30 lM [14 C]compound A in transport buffer containing 4 mM ATP, considerable uptake of compound A into MCF7/ T#8 vesicles was observed. The rate of ATP-dependent uptake into MCF7/T#8 vesicles was almost constant for
Fig. 3. Effect of osmolarity on compound A uptake by membrane vesicles: MCF7/T#8 membrane vesicles (20 lg protein/reaction) were preincubated for 30 min in transport buffer containing sucrose (0.25, 0.35, 0.5, 0.8, and 1 M). Data are averages for duplicate determinations.
5 min and was at least 4-fold higher than that in MCF7/ pc#1 vesicles (Fig. 2C). Moreover, the rate of transport into MCF7/T#8 vesicles increased linearly with the amount of total membrane protein up to 50 lg (data not shown). In contrast, the level of accumulation of compound A in MCF7/T#8 vesicles without ATP was as low as that in MCF7/pc#1 with or without ATP. Thus, these data demonstrated that ABCG2-mediated transport of indolocarbazole in membrane vesicles was noticeably stimulated by ATP. To confirm that uptake of compound A by vesicles reflects substrate movement across the membrane rather than random binding, we examined the effect of buffer osmolarity on uptake of compound A. Fig. 3 shows obvious osmotic sensitivity in the rate of compound A uptake by MCF7/T#8 vesicles. This result clearly indicated that the drug is actually transported into the intravesicular space. The relationship between [14 C]compound A uptake in MCF7/T#8 and various concentrations of [14 C]compound A (0.88–30 lM) and ATP (0.06–4 mM) are shown in Figs. 4A and B, respectively. Kinetic analysis revealed that the transport of compound A is consistent with the Michaelis–Menten model. Rates of ATP-dependent uptake reached a steady-state level of over 10 lM compound A (Fig. 4A). The kinetic parameters for compound A transport were calculated from double-reciprocal plots: Km ¼ 4:58 lM, Vmax ¼ 1:31 nmol mg1 min1 (Fig. 4A inset). The kinetic values for ATP were also determined from double-reciprocal plots and were Km ¼ 270 lM and Vmax ¼ 1:85 nmol mg1 min1 (Fig. 4B inset). It is probable that ATP serves as an energy source for ABCG2-mediated indolocarbazole uptake. To determine the nucleotide specificity of compound A transport and to ascertain whether ATP hydrolysis is necessary,
Fig. 4. (A) Effect of compound A concentration on compound A uptake by membrane vesicles: MCF7/T#8 vesicle membranes (20 lg protein/reaction) incubated with different concentrations of [14 C]compound A (0.88–30 lM) in the presence of 4 mM ATP. The inset shows a double-reciprocal plot (1/V vs. 1/[S]) of the transport activity. (B) Effect of ATP concentration on compound A uptake by membrane vesicles: ATP-dependent uptake of [14 C]compound A measured at various concentrations of ATP (0.06–4 mM) in the presence of 30 lM [14 C]compound A. Data up to 1.2 mM ATP are plotted. All data are expressed as means SEM of values in three separate experiments. The inset shows a double-reciprocal plot (1/V vs. 1/[S]) of the transport activity.
R. Nakagawa et al. / Biochemical and Biophysical Research Communications 299 (2002) 669–675
we measured compound A uptake in MCF7/T#8 membrane vesicles in the presence of four nucleoside triphosphates, ADP and a non-hydrolyzable ATP analogue, ATPcS. As shown in Fig. 5, ABCG2 displayed broad nucleotide specificity and higher transport in the presence of GTP and UTP than of ATP (148% and 131%, respectively). CTP was also substitutable for ATP. Nucleotide-dependent uptake was not visible when vesicles were incubated with ADP and ATPcS, indicating that hydrolysis of c-phosphate is required for ABCG2-mediated transport activity. Finally, we examined whether other cytotoxic drugs to which ABCG2 is or is not responsible for resistance could inhibit [14 C]compound A transport (Fig. 6). MCF7/T#8 membrane vesicles were incubated with uniform concentration (30-fold molar excess) of unlabeled drugs to which intact cells are highly resistant (mitoxantrone), modestly resistant (topotecan), and sensitive (camptothecin) (Table 1). The non-labeled other indolocarbazole compound, compound B, [6-N(1-hydroxymethyla-2-hydroxyl) ethylamino-12,13-dihydro-13-(b-D -glucopyranosyl)-5H-indolo[2,3-a]-pyrrolo [3,4-c]-carbazole-5,7(6H)-dione]] [12,17] diminished compound A uptake by 85%, while non-labeled compound A reduced it by 88%. It indicates that compound B competitively inhibited transport of the compound A. A certain degree of the decrease of compound A uptake was observed when membrane vesicles were incubated with camptothecin analogues, camptothecin and topotecan (52% and 28% of reduction, respectively). Strikingly, mitoxantrone failed to inhibit compound A transport although the cells showed 40-fold resistance to this compound. Therefore, these results suggested that there may be other binding sites for mitoxantrone or other transport mechanisms to excrete mitoxantrone other than that of indolocarbazole compounds.
Fig. 5. Nucleotide specificity of compound A transport by membrane vesicles: MCF7/T#8 membrane vesicles (20 lg protein/reaction) were incubated with 30 lM [14 C]compound A in the presence of the indicated nucleotide (4 mM) without an ATP regenerating system. Results are plotted as percentages of control value obtained with 4 mM ATP (hatched bar). Data are expressed as means SEM of values of three separate experiments.
673
Fig. 6. Inhibitions of compound A uptake by other cytotoxic drugs: MCF7/T#8 vesicle membranes (20 lg protein/reaction) were incubated with 0.88 lM [14 C]compound A (1:7 103 lCi/reaction). Non-labeled compounds were added to the reaction mixture at final concentrations of 28 lM. Results are plotted as percentages of control value obtained in the absence of the compound (hatched bar). Data are presented as mean values of two separate experiments. MX, mitoxantrone; TPT, topotecan; CPT, camptothecin; indolocarbazole, indolocarbazole compound B.
Discussion All the ABCG2-transfectants exhibited strong resistance to indolocarbazoles, implying that ABCG2 alone is sufficient for export of these agents. However, alteration of amino acid 482 seems to influence the resistance profiles between ABCG2 482T- and 482R-transfectants, supporting the early work. For further characterization of ABCG2-mediated transport, we investigated transport activities of indolocarbazoles using a plasma membrane vesicle system and showed that ABCG2 confers resistance to them by excretion in an energydependent manner. In the membrane vesicle assay, indolocarbazole compound A was a good substrate of ABCG2 and the maximum rate of ATP-dependent uptake of compound A was in a similar range to that of SN-38 [14]. The uptake of compound A was osmotically sensitive and energy-dependent. ATP hydrolysis supported the transport and the value for ATP affinity was comparable to that measured for ABCG2-ATPase activities in ABCG2 overexpressing Sf9 membranes [18]. However, compound A was accumulated more effectively under hydrolysis of GTP and UTP (Fig. 5). This result is intriguing because nucleoside triphosphates other than ATP usually cannot energize transport as capably as ATP [16,19], though it is not uncommon for ABC transporters to utilize various ribonucleotides. Biswas [20] lately showed that the first nucleotide-binding domain of the human retinal ABC transporter has higher affinity to CTP and GTP than to ATP for hydrolysis and indicated that a certain ABC transporter can produce energy for transport from any available ribonucleotide depending on cellular conditions. Hence, our
674
R. Nakagawa et al. / Biochemical and Biophysical Research Communications 299 (2002) 669–675
data suggested that ABCG2 is also functioning as a general nucleotidase. We examined the level at which cytotoxic drugs could inhibit [14 C]indolocarbazole compound A transport (Fig. 6). Twenty-eight percent decrease of compound A uptake was observed when membrane vesicles were incubated with topotecan. Since resistance to some of camptothecins, such as topotecan and SN-38, is considered to be attributable to ABCG2 overexpression and actually SN-38 was shown to be excreted by ABCG2 in membrane vesicle studies [14], it is likely that topotecan is directly transported by ABCG2. This clone is moderately resistant to topotecan, and therefore, it is probable that the drug inhibited compound A uptake in a certain degree by the similar mechanism of indolocarbazole transport. Unexpectedly, camptothecin inhibited compound A uptake by 52% though the intact cells are sensitive to this drug. Taking into account structural similarity between camptothecin and other camptothecins which are substrates of ABCG2, camptothecin itself may be able to bind to this transporter in a mutually competitive site with compound A, which resulted in its inhibition of transport of labeled-compound A. Of our interest, mitoxantrone did not inhibit [14 C]compound A transport in the vesicle assay system. This result suggested that unmodified mitoxantrone is not transported by ABCG2, or that other cytoplasmic factors are required for mitoxantrone export. We also cannot exclude the possibility that ABCG2 possesses multiple binding sites for substrates as shown in both Pgp and MRP1 [21,22]. Hence, these data suggested that other potential binding sites for mitoxantrone or different transport mechanisms to excrete mitoxantrone may exist other than that of indolocarbazole derivatives. In conclusion, this study confirmed that ABCG2 is responsible for indolocarbazole resistance and revealed that ABCG2 confers resistance by transporting indolocarbazole dependent on hydrolysis of ATP. Further investigations are needed for fully understanding transport mechanisms by ABCG2.
Acknowledgments We thank Kenji Tanaka, Hiroshi Hirai, Hidehito Kotani, and Mitsuaki Yoshida for helpful discussion.
References [1] T. Litman, T.E. Druley, W.D. Stein, S.E. Bates, From MDR to MXR: new understanding of multidrug resistance systems, their properties and clinical significance, Cell. Mol. Life Sci. 58 (2001) 931–959. [2] L.A. Doyle, W. Yang, L.V. Abruzzo, T. Krogmann, Y. Gao, A.K. Rishi, D.D. Ross, A multidrug resistance transporter from human MCF-7 breast cancer cells, Proc. Natl. Acad. Sci. USA 95 (1998) 15665–15670.
[3] R. Allikmets, L.M. Schriml, A. Hutchinson, V. Romano-Spica, M. Dean, A human placenta-specific ATP-binding cassette gene (ABCP) on chromosome 4q22 that is involved in multidrug resistance, Cancer Res. 58 (1998) 5337–5339. [4] K. Miyake, L. Mickley, T. Litman, Z. Zhan, R. Robey, B. Cristensen, M. Brangi, L. Greenberger, M. Dean, T. Fojo, S.E. Bates, Molecular cloning of cDNAs which are highly overexpressed in mitoxantrone-resistant cells: demonstration of homology to ABC transport genes, Cancer Res. 59 (1999) 8–13. [5] J. Klucken, C. Buchler, E. Orso, W.E. Kaminski, M. PorschOzcurumez, G. Liebisch, M. Kapinsky, W. Diederich, W. Drobnik, M. Dean, R. Allikmets, G. Schmitz, ABCG1 (ABC8), the human homolog of the Drosophila white gene, is a regulator of macrophage cholesterol and phospholipid transport, Proc. Natl. Acad. Sci. USA 97 (2000) 817–822. [6] J.W. Jonker, J.W. Smit, R.F. Brinkhuis, M. Maliepaard, J.H. Beijnen, J.H. Schellens, A.H. Schinkel, Role of breast cancer resistance protein in the bioavailability and fetal penetration of topotecan, J. Natl. Cancer Inst. 92 (2000) 1651–1656. [7] M. Maliepaard, G.L. Scheffer, I.F. Faneyte, M.A. van Gastelen, A.C. Pijnenborg, A.H. Schinkel, M.J. van de Vijver, R.J. Scheper, J.H.M. Schellens, Subcellular localization and distribution of the breast cancer resistance protein transporter in normal human tissues, Cancer Res. 61 (2001) 3458–3464. [8] G.L. Scheffer, M. Maliepaard, A.C.L.M. Pijnenborg, M.A. van Gastelen, M.C. de Jong, A.B. Schroeijers, D.M. van der Kolk, J.D. Allen, D.D. Ross, P. van der Valk, W.S. Dalton, J.H.M. Schellens, R.J. Scheper, Breast cancer resistance protein is localized at the plasma membrane in mitoxantrone- and topotecan-resistant cell lines, Cancer Res. 60 (2000) 2589– 2593. [9] E. Rocchi, A. Khodjakov, E.L. Volk, C.H. Yang, T. Litman, S.E. Bates, E. Schneider, The product of the ABC half-transporter gene ABCG2 (BCRP/MXR/ABCP) is expressed in the plasma membrane, Biochem. Biophys. Res. Commun. 271 (2000) 42–46. [10] M. Maliepaard, M.A. van Gastelen, L.A. de Jong, D. Pluim, R.C.A.M. van Waardenburg, M.C. Ruevekamp-Helmers, B.G.J. Floot, J.H.M. Schellens, Overexpression of the BCRP/MXR/ ABCP gene in a topotecan-selected ovarian tumor cell line, Cancer Res. 59 (1999) 4559–4563. [11] M. Brangi, T. Litman, M. Ciotti, K. Nishiyama, G. Kohlhagen, C. Takimoto, R. Robey, Y. Pommier, T. Fojo, S.E. Bates, Camptothecin resistance: role of the ATP-binding cassette (ABC), mitoxantrone-resistance half-transporter (MXR), and potential for glucuronidation in MXR-expressing cells, Cancer Res. 59 (1999) 5938–5946. [12] H. Komatani, H. Kotani, Y. Hara, R. Nakagawa, M. Matsumoto, H. Arakawa, S. Nishimura, Identification of breast cancer resistant protein/mitoxantrone resistance/placenta-specific, ATPbinding cassette transporter as a transporter of NB-506 and J107088, topoisomerase I inhibitors with an indolocarbazole structure, Cancer Res. 61 (2001) 2827–2832. [13] Y. Honjo, C.A. Hrycyna, Q.W. Yan, W.Y. Medina-Perez, R.W. Robey, A. van de Laar, T. Litman, M. Dean, S.E. Bates, Acquired mutations in the MXR/BCRP/ABCP gene alter substrate specificity in MXR/BCRP/ABCP-overexpressing cells, Cancer Res. 61 (2001) 6635–6639. [14] K. Nakatomi, M. Yoshikawa, M. Oka, Y. Ikegami, S. Hayasaka, K. Sano, K. Shiozawa, S. Kawabata, H. Soda, T. Ishikawa, S. Tanabe, S. Kohno, Transport of 7-ethyl-10-hydroxycamptothecin (SN-38) by breast cancer resistance protein ABCG2 in human lung cancer cells, Biochem. Biophys. Res. Commun. 288 (2001) 827–832. [15] T. Ishikawa, C.D. Wright, H. Ishizuka, GS-X pump is functionally overexpressed in cis-diamminedichloroplatinum (II)-resistant human leukemia HL-60 cells and down-regulated by cell differentiation, J. Biol. Chem. 269 (1994) 29085–29093.
R. Nakagawa et al. / Biochemical and Biophysical Research Communications 299 (2002) 669–675 [16] T. Ishikawa, ATP/Mg2þ -dependent cardiac transport system for glutathione S-conjugates. A study using rat heart sarcolemma vesicles, J. Biol. Chem. 264 (1989) 17343–17348. [17] T. Yoshinari, M. Ohkubo, K. Fukasawa, S. Egashira, Y. Hara, M. Matsumoto, K. Nakai, H. Arakawa, H. Morishima, S. Nishimura, Mode of action of a new indolocarbazole anticancer agent, J-107088, targeting topoisomerase I, Cancer Res. 59 (1999) 4271–4275. [18] C. Ozvegy, T. Litman, G. Szakacs, Z. Nagy, S. Bates, A. Varadi, B. Sarkadi, Functional characterization of the human multidrug transporter, ABCG2, expressed in insect cells, Biochem. Biophys. Res. Commun. 285 (2001) 111–117. [19] L. Bishop, R. Agbayani Jr., S.V. Ambudkar, P.C. Maloney, G.F. Ames, Reconstitution of a bacterial periplasmic permease in
675
proteoliposomes and demonstration of ATP hydrolysis concomitant with transport, Proc. Natl. Acad. Sci. USA 86 (1989) 6953– 6957. [20] E.E. Biswas, Nucleotide binding domain 1 of the human retinal ABC transporter functions as a general ribonucleotidase, Biochemistry 40 (2001) 8181–8187. [21] D.W. Loe, R.G. Deeley, S.P.C. Cole, Characterization of vincristine transport by the M(r) 190,000 multidrug resistance protein (MRP): evidence for cotransport with reduced glutathione, Cancer Res. 58 (1998) 5130–5136. [22] S. Dey, M. Ramachandra, I. Pastan, M.M. Gottesman, S.V. Ambudkar, Evidence for two nonidentical drug-interaction sites in the human P-glycoprotein, Proc. Natl. Acad. Sci. USA 94 (1997) 10594–10599.
ABCG2 confers resistance to indolocarbazole compounds by ATP-dependent transportq Rinako Nakagawa, Yoshikazu Hara, Hiroharu Arakawa, Susumu Nishimura, and Hideya Komatani* Banyu Tsukuba Research Institute in Collaboration with Merck Research Laboratories, 3 Okubo, Ibaraki 300-2611, Japan Received 1 November 2002
Abstract The ABC half-transporter, ABCG2, is known to confer resistance to chemotherapeutic agents including indolocarbazole derivatives. MCF7 cells were introduced by either wild type ABCG2 (ABCG2-482R) or mutant ABCG2 (-482T), whose amino acid at position 482 is substituted to threonine from arginine, and their cross-resistance pattern was analyzed. Although this amino acid substitution seems to affect cross-resistance patterns, both 482T- and 482R-transfectants showed strong resistance to indolocarbazoles, confirming that ABCG2 confers resistance to them. For further characterization of ABCG2-mediated transport, we investigated indolocarbazole compound A (Fig. 1) excretion in cell-free system. Compound A was actively transported in membrane vesicles prepared from one of the 482T-transfectants and its uptake was supported by hydrolysis of various nucleoside triphosphates. This transport was inhibited completely by the other indolocarbazole compound, but not by mitoxantrone, implying that the binding site of mitoxantrone or the transport mechanisms for mitoxantrone is different from those of indolocarbazoles. These results showed that ABCG2 confers resistance to indolocarbazoles by transporting them in an energy-dependent manner. Ó 2002 Elsevier Science (USA). All rights reserved. Keywords: ABC transporter; ABCG2; MXR; BCRP; Drug resistance; Indolocarbazole; Vesicle transport
The development of multidrug resistance in tumor cells is frequently associated with overexpression of the ABC transporter family. Well-characterized ABC transporter members, such as P-gp (ABCB1) and MRP1 (ABCC1), have been shown to be involved in excretion of certain anticancer drugs by active transport [1]. The second member of the ATP-binding cassette transporter G subfamily (ABCG2), which is also denoted as BCRP/ABCP/MXR, is also capable of conferring drug resistance. Unlike other multidrug ABC transporters, ABCG2 is a half-transporter consisting of six putative transmembrane domains and a single
q
Abbreviations: ABC, ATP-binding cassette; BCRP, breast cancer resistance protein; MXR, mitoxantrone resistance-associated protein; ABCP, placenta specific ABC transporter; P-gp, P-glycoprotein; MRP, multidrug resistance-associated protein; SN-38, 7-ethyl-10hydroxycamptothecin; ATPcS, adenosine 50 -O-(3-thiotriphosphate). * Corresponding author. Fax: +81-298-77-2027. E-mail address: [email protected] (H. Komatani).
ATP-binding domain [2–4]. The ABCG subfamily contains the Drosophila white/brown/scarlet genes, which are known to transport eye pigment, and the human homologue of the Drosophila white gene, ABCG1, which was recently proved to be involved with lipophilic xenobiotics [5]. The physiological role of ABCG2 is not yet understood. However, high expression of ABCG2 was observed in the placenta, colon, and small intestine, suggesting protective functions on the fetus and gastrointestinal tract [6,7]. Extensive studies using drug-selected cell lines and cells with enforced ABCG2 expression demonstrated that this protein is localized in the plasma membrane [8,9] and confers high levels of resistance to multiple chemotherapeutic drugs, including mitoxantrone [2], anthracyclines [2], the camptothecins (i.e., topotecan and active form of irinotecan, SN-38) [10,11], and indolocarbazole analogues [12]. Originally this gene was cloned by different groups with a different amino acid at codon 482 [2–4]. We recently proposed that the
0006-291X/02/$ - see front matter Ó 2002 Elsevier Science (USA). All rights reserved. PII: S 0 0 0 6 - 2 9 1 X ( 0 2 ) 0 2 7 1 2 - 2
670
R. Nakagawa et al. / Biochemical and Biophysical Research Communications 299 (2002) 669–675
gene carrying arginine at this position is the naturally expressed form, whereas the genes holding threonine or glycine are mutated forms [12]. It has also been proved that this amino acid substitution alters the cross-resistance pattern to rhodamine 123 and adriamycin [13]. To date, Nakatomi et al. [14] reported that ABCG2 functions as an ATP-dependent efflux pump for SN-38 using plasma membrane vesicles obtained from SN-38 selected cells. However, ABCG2-mediated export systems have not been characterized in detail and are still largely unknown. In the present study, we examined drug resistance phenotype of ABCG2-482T and -482R introduced MCF7 cells. Although the amino acid substitution seems to affect cross-resistance patterns, both types of cells are highly resistant to indolocarbazoles, confirming that ABCG2 confers resistance to indolocarbazole compounds. In order to characterize mechanisms of resistance to indolocarbazoles, we investigated transport activities of ABCG2 in a plasma membrane vesicle from one of the transfectants. Here we provide the direct evidence about ATP-dependent export of the indolocarbazole, by which ABCG2 confers great resistance to the compound.
Materials and methods Reagents. Indolocarbazole compound A (Fig. 1), indolocarbazole compound B, and topotecan were synthesized in our institute as previously described [12]. 14 C-labeled compound A was from Daiichi Pure Chemical (Tokyo, Japan). Mitoxantrone, camptothecin, ADP, ATPcS, and creatine phosphokinase were purchased from Sigma Chemical (St. Louis, MO). ATP, GTP, CTP, and UTP were obtained from Amersham Pharmacia Biotech (Piscataway, NJ). Creatine phosphate was from Wako Pure Chemicals (Osaka, Japan). ABCG2 transfectants. MCF7 human breast carcinoma cells were transfected with pcDNA3.1/V5-His-TOPO containing full-length ABCG2 cDNA (482T variant) [2] or pcDNA3.1/V5-His-TOPO using
Effectene Transfection Reagent (Qiagen, Valencia, CA), selected by culture with geneticin (G418, 0.9 mg/ml), and isolated with cloning rings. Clones with high-level expression of ABCG2 were selected by Northern blots. All cell lines were maintained in RPMI 1640 medium supplemented with 10% FCS, 100 lg/ml gentamicin, and selected concentrations of G418. Northern blotting. Total RNA was extracted from ABCG2-transfectants using an RNeasy Kit (Qiagen). Samples of 8 lg total RNA were separated on 1% agarose gel and transferred to nylon membranes. The membranes were probed with a 32 P-labeled 269-bp fragment of human ABCG2 cDNA (161–429 of AF103796). Cytotoxicity assay. Samples of 1500 cells/well were plated one day before adding a drug and grown for a further 3 days with various concentrations of the drugs. The sensitivity of MCF7 transfectants to chemotherapeutic agents was determined by sulforhodamine-B cytotoxicity assay. DMSO-treated cells in control wells were assigned a value of 100% and IC50 values were defined as the dose required to inhibit growth of cells to 50% of that of the controls. Membrane vesicle preparation. Plasma membrane vesicles were prepared as described previously [15]. Briefly, cells (3:0 108 ) were harvested by scraping and washed once in ice-cold phosphate-buffered saline (PBS). The cell pellet was homogenized in buffer containing 0.5 mM Na-phosphate (pH 7.0), 0.1 mM EDTA, and protease inhibitors (Roche, Mannheim, Germany) using a Potter–Elvehjem homogenizer and then the cell lysate was centrifuged at 100,000g for 30 min. The crude membrane fraction was layered over 38% sucrose solution and centrifuged at 100,000g for 30 min. The interface was collected, suspended in 0.25 M sucrose containing 10 mM Tris–HCl (pH 7.4), and centrifuged at 100,000g for 20 min. The membrane pellet was resuspended in transport buffer [0.25 M sucrose containing 10 mM Tris– HCl (pH 7.4)], frozen in liquid N2 and stored at )80 °C until use. Western blotting. Samples of 5.5 lg cellular membrane proteins were loaded onto 7.5% polyacrylamide gel and subjected to electrophoresis. Proteins were transferred to a PVDF membrane (Millipore, Bedford, MA). The membrane was blocked with Tris-buffered saline (pH 7.6) containing 0.1% (v/v) Tween 20 and 2% Block Ace (Dainippon Seiyaku, Osaka, Japan). The membranes were probed for 1.5 h with 1:700 dilution of a polyclonal anti-ABCG2 antibody, which we raised against a 15-mer peptide of ABCG2:NREEDFKATEIIEPS, followed by a secondary horseradish peroxidase-conjugated donkey anti-rabbit antibody (Amersham) (1:10,000 dilution) for 1 h. Detection was enhanced by chemiluminescence detection (Amersham) and immunoreactive bands were visualized. Vesicle transport studies. Thawed membranes were diluted in the transport buffer: 0.25 M sucrose, 10 mM Tris–HCl (pH 7.4), and passed 20-times through a 27-gauge needle using a syringe for vesicle formation. Protein concentration was measured by Bradford assay. Standard transport assays were performed at 37 °C in a 50 ll volume of mixture consisting of 20 lg vesicle protein, 30 lM [14 C]compound A (3:4 103 lCi/reaction), 10 mM MgCl2 , 4 mM ATP, and an ATP regenerating system: 10 mM creatine phosphate and 100 lg/ml creatine kinase in transport buffer. The mixtures were incubated at 37 °C for 5 min. The reactions were terminated by adding 1 ml ice-cold transport buffer and then the mixtures were filtered through glass fiber (type GF/C) filters (Millipore) using a rapid filtration device with light suction and rinsed four times with 1 ml ice-cold transport buffer. The filters were then dissolved and radioactivity was determined. Data on ATP-dependent transport were calculated by subtracting values in the absence of ATP from those in the presence of ATP [16].
Results Fig. 1. Structures of indolocarbazole compound A: 6-N-formylamino12,13-dihydro-1,11-dihydroxy-13-(b-D -glucopyranosil)5H-indolo[2,3-a] pyrrolo [3,4-c]carbazole-5,7(6H)-dione.
We introduced an expression vector containing fulllength human ABCG2 cDNA (482R and 482T variants)
R. Nakagawa et al. / Biochemical and Biophysical Research Communications 299 (2002) 669–675
671
Fig. 2. (A) Northern blot of MCF7 cells transfected with vector, ABCG2-482R- or ABCG2-482T-variant (8 lg total RNA/lane). (B) Western blot with polyclonal anti-ABCG2 antibody of membranes (5.5 lg protein/lane) from MCF7/T#8 or MCF7/pc#1. (C) Time courses of indolocarbazole compound A uptake in membranes from MCF7/T#8 and MCF7/pc#1 cells: membrane vesicles (20 lg protein/reaction) from T#8 (j,) and pc#1 (d,s) were incubated with 30 lM [14 C]compound A (3:4 103 lCi/reaction) in transport buffer [0.25 M sucrose, 10 mM Tris–HCl (pH 7.4)] with an ATP-regenerating system (10 mM creatine phosphate and 100 lg/ml creatine kinase). Closed symbols represent uptake in the presence of 4 mM ATP; open symbols represent uptake in the absence of ATP. Values are means SEM in three separate experiments.
into MCF7, which was reported to be highly resistant to mitoxantrone upon introduction of 482T [2]. ABCG2 mRNA expression was analyzed by Northern blots in ABCG2- or vector-transfected cells (Fig. 2A). All the ABCG2-introduced clones expressed prominent amount of ABCG2 mRNA with a similar level. The mRNA of ABCG2 was not detected in a vector-transrfected cell (MCF7/pc#1). Table 1 shows the relative resistance value to various anti-cancer drugs. ABCG2-transfected cells displayed pronounced resistance to indolocarbazole compound A (26–298-fold resistance) as PC-13 cells transfected with ABCG2-482R did in an earlier study [12], confirming that ABCG2 is responsible for indolocarbazole resistance. MCF7 cells expressing 482R (MCF7/R#7 and R#59) showed no resistance to adriamycin, but showed mild resistance to mitoxantrone and topotecan (7.6–8.8-fold, 5.1–6.6-fold, respectively) and the strongest resistance to indolocarbazole compound A (106–298-fold), thereby displaying a rather indolocar-
bazole-specific resistance phenotype. On the other hand, MCF7 clones expressing 482T (MCF7/T#8 and T#79) showed modest resistance to adriamycin and topotecan (3.8–5.5-fold, 2.7–4.7-fold, respectively), strong resistance to mitoxantrone (16–40-fold) and strongest resistance, but in less magnitude than 482R-transfectants, to compound A (26–124-fold). Though there are differences in resistance magnitude between the clones with similar transcript level of 482R or 482T, these results are concordant with the previous finding [13] that alteration of amino acid 482 affects the cross-resistance patterns of ABCG2. The differences of resistance might be due to what expression level of the protein or stability of the protein is different between the clones. Transfection of both 482R- and 482T-transfectants was found to have only a modest effect on resistance to topotecan and not to affect sensitivity to camptothecin. The MCF7/T#8 was chosen out of the transfectants to investigate the mechanism of drug transport associated
Table 1 Cross-resistance profiles of ABCG2-transfected cells Drug
MCF7/pc#1 IC50 (lM)
MCF7/T#8 RRa
MCF7/T#79 RRa
MCF7/R#7 RRa
MCF7/R#59 RRa
Compound A Adriamycin Camptothecin Topotecan Mitoxantrone
0.84 0.28 0.024 0.0064 0.012 0.0018 0.018 0.0038 0.0031 0.00071
124 5.5 1.3 4.7 40
26 3.8 1.1 2.7 16
298 1.4 1.4 6.6 8.8
106 1.4 1.8 5.1 7.6
Note. IC50 values are shown as means SEM of more than five separate experiments. a RR, relative resistance was calculated by dividing the IC50 values of the MCF7 clones by those of MCF7/pc#1 cells.
672
R. Nakagawa et al. / Biochemical and Biophysical Research Communications 299 (2002) 669–675
with resistance to indolocarbazole. Western blot analysis of membrane protein from this clone showed that the cells expressed a large amount of ABCG2 protein (Fig. 2B). The plasma membrane localization of ABCG2 in MCF7/T#8 cells was confirmed by immunocytochemical analysis with a monoclonal antibody (data not shown). The time course and ATP-dependence of compound A accumulation by vesicles prepared from MCF7/T#8 cells or MCF7/pc#1 cells are shown in Fig. 2C. When membrane vesicles were incubated with 30 lM [14 C]compound A in transport buffer containing 4 mM ATP, considerable uptake of compound A into MCF7/ T#8 vesicles was observed. The rate of ATP-dependent uptake into MCF7/T#8 vesicles was almost constant for
Fig. 3. Effect of osmolarity on compound A uptake by membrane vesicles: MCF7/T#8 membrane vesicles (20 lg protein/reaction) were preincubated for 30 min in transport buffer containing sucrose (0.25, 0.35, 0.5, 0.8, and 1 M). Data are averages for duplicate determinations.
5 min and was at least 4-fold higher than that in MCF7/ pc#1 vesicles (Fig. 2C). Moreover, the rate of transport into MCF7/T#8 vesicles increased linearly with the amount of total membrane protein up to 50 lg (data not shown). In contrast, the level of accumulation of compound A in MCF7/T#8 vesicles without ATP was as low as that in MCF7/pc#1 with or without ATP. Thus, these data demonstrated that ABCG2-mediated transport of indolocarbazole in membrane vesicles was noticeably stimulated by ATP. To confirm that uptake of compound A by vesicles reflects substrate movement across the membrane rather than random binding, we examined the effect of buffer osmolarity on uptake of compound A. Fig. 3 shows obvious osmotic sensitivity in the rate of compound A uptake by MCF7/T#8 vesicles. This result clearly indicated that the drug is actually transported into the intravesicular space. The relationship between [14 C]compound A uptake in MCF7/T#8 and various concentrations of [14 C]compound A (0.88–30 lM) and ATP (0.06–4 mM) are shown in Figs. 4A and B, respectively. Kinetic analysis revealed that the transport of compound A is consistent with the Michaelis–Menten model. Rates of ATP-dependent uptake reached a steady-state level of over 10 lM compound A (Fig. 4A). The kinetic parameters for compound A transport were calculated from double-reciprocal plots: Km ¼ 4:58 lM, Vmax ¼ 1:31 nmol mg1 min1 (Fig. 4A inset). The kinetic values for ATP were also determined from double-reciprocal plots and were Km ¼ 270 lM and Vmax ¼ 1:85 nmol mg1 min1 (Fig. 4B inset). It is probable that ATP serves as an energy source for ABCG2-mediated indolocarbazole uptake. To determine the nucleotide specificity of compound A transport and to ascertain whether ATP hydrolysis is necessary,
Fig. 4. (A) Effect of compound A concentration on compound A uptake by membrane vesicles: MCF7/T#8 vesicle membranes (20 lg protein/reaction) incubated with different concentrations of [14 C]compound A (0.88–30 lM) in the presence of 4 mM ATP. The inset shows a double-reciprocal plot (1/V vs. 1/[S]) of the transport activity. (B) Effect of ATP concentration on compound A uptake by membrane vesicles: ATP-dependent uptake of [14 C]compound A measured at various concentrations of ATP (0.06–4 mM) in the presence of 30 lM [14 C]compound A. Data up to 1.2 mM ATP are plotted. All data are expressed as means SEM of values in three separate experiments. The inset shows a double-reciprocal plot (1/V vs. 1/[S]) of the transport activity.
R. Nakagawa et al. / Biochemical and Biophysical Research Communications 299 (2002) 669–675
we measured compound A uptake in MCF7/T#8 membrane vesicles in the presence of four nucleoside triphosphates, ADP and a non-hydrolyzable ATP analogue, ATPcS. As shown in Fig. 5, ABCG2 displayed broad nucleotide specificity and higher transport in the presence of GTP and UTP than of ATP (148% and 131%, respectively). CTP was also substitutable for ATP. Nucleotide-dependent uptake was not visible when vesicles were incubated with ADP and ATPcS, indicating that hydrolysis of c-phosphate is required for ABCG2-mediated transport activity. Finally, we examined whether other cytotoxic drugs to which ABCG2 is or is not responsible for resistance could inhibit [14 C]compound A transport (Fig. 6). MCF7/T#8 membrane vesicles were incubated with uniform concentration (30-fold molar excess) of unlabeled drugs to which intact cells are highly resistant (mitoxantrone), modestly resistant (topotecan), and sensitive (camptothecin) (Table 1). The non-labeled other indolocarbazole compound, compound B, [6-N(1-hydroxymethyla-2-hydroxyl) ethylamino-12,13-dihydro-13-(b-D -glucopyranosyl)-5H-indolo[2,3-a]-pyrrolo [3,4-c]-carbazole-5,7(6H)-dione]] [12,17] diminished compound A uptake by 85%, while non-labeled compound A reduced it by 88%. It indicates that compound B competitively inhibited transport of the compound A. A certain degree of the decrease of compound A uptake was observed when membrane vesicles were incubated with camptothecin analogues, camptothecin and topotecan (52% and 28% of reduction, respectively). Strikingly, mitoxantrone failed to inhibit compound A transport although the cells showed 40-fold resistance to this compound. Therefore, these results suggested that there may be other binding sites for mitoxantrone or other transport mechanisms to excrete mitoxantrone other than that of indolocarbazole compounds.
Fig. 5. Nucleotide specificity of compound A transport by membrane vesicles: MCF7/T#8 membrane vesicles (20 lg protein/reaction) were incubated with 30 lM [14 C]compound A in the presence of the indicated nucleotide (4 mM) without an ATP regenerating system. Results are plotted as percentages of control value obtained with 4 mM ATP (hatched bar). Data are expressed as means SEM of values of three separate experiments.
673
Fig. 6. Inhibitions of compound A uptake by other cytotoxic drugs: MCF7/T#8 vesicle membranes (20 lg protein/reaction) were incubated with 0.88 lM [14 C]compound A (1:7 103 lCi/reaction). Non-labeled compounds were added to the reaction mixture at final concentrations of 28 lM. Results are plotted as percentages of control value obtained in the absence of the compound (hatched bar). Data are presented as mean values of two separate experiments. MX, mitoxantrone; TPT, topotecan; CPT, camptothecin; indolocarbazole, indolocarbazole compound B.
Discussion All the ABCG2-transfectants exhibited strong resistance to indolocarbazoles, implying that ABCG2 alone is sufficient for export of these agents. However, alteration of amino acid 482 seems to influence the resistance profiles between ABCG2 482T- and 482R-transfectants, supporting the early work. For further characterization of ABCG2-mediated transport, we investigated transport activities of indolocarbazoles using a plasma membrane vesicle system and showed that ABCG2 confers resistance to them by excretion in an energydependent manner. In the membrane vesicle assay, indolocarbazole compound A was a good substrate of ABCG2 and the maximum rate of ATP-dependent uptake of compound A was in a similar range to that of SN-38 [14]. The uptake of compound A was osmotically sensitive and energy-dependent. ATP hydrolysis supported the transport and the value for ATP affinity was comparable to that measured for ABCG2-ATPase activities in ABCG2 overexpressing Sf9 membranes [18]. However, compound A was accumulated more effectively under hydrolysis of GTP and UTP (Fig. 5). This result is intriguing because nucleoside triphosphates other than ATP usually cannot energize transport as capably as ATP [16,19], though it is not uncommon for ABC transporters to utilize various ribonucleotides. Biswas [20] lately showed that the first nucleotide-binding domain of the human retinal ABC transporter has higher affinity to CTP and GTP than to ATP for hydrolysis and indicated that a certain ABC transporter can produce energy for transport from any available ribonucleotide depending on cellular conditions. Hence, our
674
R. Nakagawa et al. / Biochemical and Biophysical Research Communications 299 (2002) 669–675
data suggested that ABCG2 is also functioning as a general nucleotidase. We examined the level at which cytotoxic drugs could inhibit [14 C]indolocarbazole compound A transport (Fig. 6). Twenty-eight percent decrease of compound A uptake was observed when membrane vesicles were incubated with topotecan. Since resistance to some of camptothecins, such as topotecan and SN-38, is considered to be attributable to ABCG2 overexpression and actually SN-38 was shown to be excreted by ABCG2 in membrane vesicle studies [14], it is likely that topotecan is directly transported by ABCG2. This clone is moderately resistant to topotecan, and therefore, it is probable that the drug inhibited compound A uptake in a certain degree by the similar mechanism of indolocarbazole transport. Unexpectedly, camptothecin inhibited compound A uptake by 52% though the intact cells are sensitive to this drug. Taking into account structural similarity between camptothecin and other camptothecins which are substrates of ABCG2, camptothecin itself may be able to bind to this transporter in a mutually competitive site with compound A, which resulted in its inhibition of transport of labeled-compound A. Of our interest, mitoxantrone did not inhibit [14 C]compound A transport in the vesicle assay system. This result suggested that unmodified mitoxantrone is not transported by ABCG2, or that other cytoplasmic factors are required for mitoxantrone export. We also cannot exclude the possibility that ABCG2 possesses multiple binding sites for substrates as shown in both Pgp and MRP1 [21,22]. Hence, these data suggested that other potential binding sites for mitoxantrone or different transport mechanisms to excrete mitoxantrone may exist other than that of indolocarbazole derivatives. In conclusion, this study confirmed that ABCG2 is responsible for indolocarbazole resistance and revealed that ABCG2 confers resistance by transporting indolocarbazole dependent on hydrolysis of ATP. Further investigations are needed for fully understanding transport mechanisms by ABCG2.
Acknowledgments We thank Kenji Tanaka, Hiroshi Hirai, Hidehito Kotani, and Mitsuaki Yoshida for helpful discussion.
References [1] T. Litman, T.E. Druley, W.D. Stein, S.E. Bates, From MDR to MXR: new understanding of multidrug resistance systems, their properties and clinical significance, Cell. Mol. Life Sci. 58 (2001) 931–959. [2] L.A. Doyle, W. Yang, L.V. Abruzzo, T. Krogmann, Y. Gao, A.K. Rishi, D.D. Ross, A multidrug resistance transporter from human MCF-7 breast cancer cells, Proc. Natl. Acad. Sci. USA 95 (1998) 15665–15670.
[3] R. Allikmets, L.M. Schriml, A. Hutchinson, V. Romano-Spica, M. Dean, A human placenta-specific ATP-binding cassette gene (ABCP) on chromosome 4q22 that is involved in multidrug resistance, Cancer Res. 58 (1998) 5337–5339. [4] K. Miyake, L. Mickley, T. Litman, Z. Zhan, R. Robey, B. Cristensen, M. Brangi, L. Greenberger, M. Dean, T. Fojo, S.E. Bates, Molecular cloning of cDNAs which are highly overexpressed in mitoxantrone-resistant cells: demonstration of homology to ABC transport genes, Cancer Res. 59 (1999) 8–13. [5] J. Klucken, C. Buchler, E. Orso, W.E. Kaminski, M. PorschOzcurumez, G. Liebisch, M. Kapinsky, W. Diederich, W. Drobnik, M. Dean, R. Allikmets, G. Schmitz, ABCG1 (ABC8), the human homolog of the Drosophila white gene, is a regulator of macrophage cholesterol and phospholipid transport, Proc. Natl. Acad. Sci. USA 97 (2000) 817–822. [6] J.W. Jonker, J.W. Smit, R.F. Brinkhuis, M. Maliepaard, J.H. Beijnen, J.H. Schellens, A.H. Schinkel, Role of breast cancer resistance protein in the bioavailability and fetal penetration of topotecan, J. Natl. Cancer Inst. 92 (2000) 1651–1656. [7] M. Maliepaard, G.L. Scheffer, I.F. Faneyte, M.A. van Gastelen, A.C. Pijnenborg, A.H. Schinkel, M.J. van de Vijver, R.J. Scheper, J.H.M. Schellens, Subcellular localization and distribution of the breast cancer resistance protein transporter in normal human tissues, Cancer Res. 61 (2001) 3458–3464. [8] G.L. Scheffer, M. Maliepaard, A.C.L.M. Pijnenborg, M.A. van Gastelen, M.C. de Jong, A.B. Schroeijers, D.M. van der Kolk, J.D. Allen, D.D. Ross, P. van der Valk, W.S. Dalton, J.H.M. Schellens, R.J. Scheper, Breast cancer resistance protein is localized at the plasma membrane in mitoxantrone- and topotecan-resistant cell lines, Cancer Res. 60 (2000) 2589– 2593. [9] E. Rocchi, A. Khodjakov, E.L. Volk, C.H. Yang, T. Litman, S.E. Bates, E. Schneider, The product of the ABC half-transporter gene ABCG2 (BCRP/MXR/ABCP) is expressed in the plasma membrane, Biochem. Biophys. Res. Commun. 271 (2000) 42–46. [10] M. Maliepaard, M.A. van Gastelen, L.A. de Jong, D. Pluim, R.C.A.M. van Waardenburg, M.C. Ruevekamp-Helmers, B.G.J. Floot, J.H.M. Schellens, Overexpression of the BCRP/MXR/ ABCP gene in a topotecan-selected ovarian tumor cell line, Cancer Res. 59 (1999) 4559–4563. [11] M. Brangi, T. Litman, M. Ciotti, K. Nishiyama, G. Kohlhagen, C. Takimoto, R. Robey, Y. Pommier, T. Fojo, S.E. Bates, Camptothecin resistance: role of the ATP-binding cassette (ABC), mitoxantrone-resistance half-transporter (MXR), and potential for glucuronidation in MXR-expressing cells, Cancer Res. 59 (1999) 5938–5946. [12] H. Komatani, H. Kotani, Y. Hara, R. Nakagawa, M. Matsumoto, H. Arakawa, S. Nishimura, Identification of breast cancer resistant protein/mitoxantrone resistance/placenta-specific, ATPbinding cassette transporter as a transporter of NB-506 and J107088, topoisomerase I inhibitors with an indolocarbazole structure, Cancer Res. 61 (2001) 2827–2832. [13] Y. Honjo, C.A. Hrycyna, Q.W. Yan, W.Y. Medina-Perez, R.W. Robey, A. van de Laar, T. Litman, M. Dean, S.E. Bates, Acquired mutations in the MXR/BCRP/ABCP gene alter substrate specificity in MXR/BCRP/ABCP-overexpressing cells, Cancer Res. 61 (2001) 6635–6639. [14] K. Nakatomi, M. Yoshikawa, M. Oka, Y. Ikegami, S. Hayasaka, K. Sano, K. Shiozawa, S. Kawabata, H. Soda, T. Ishikawa, S. Tanabe, S. Kohno, Transport of 7-ethyl-10-hydroxycamptothecin (SN-38) by breast cancer resistance protein ABCG2 in human lung cancer cells, Biochem. Biophys. Res. Commun. 288 (2001) 827–832. [15] T. Ishikawa, C.D. Wright, H. Ishizuka, GS-X pump is functionally overexpressed in cis-diamminedichloroplatinum (II)-resistant human leukemia HL-60 cells and down-regulated by cell differentiation, J. Biol. Chem. 269 (1994) 29085–29093.
R. Nakagawa et al. / Biochemical and Biophysical Research Communications 299 (2002) 669–675 [16] T. Ishikawa, ATP/Mg2þ -dependent cardiac transport system for glutathione S-conjugates. A study using rat heart sarcolemma vesicles, J. Biol. Chem. 264 (1989) 17343–17348. [17] T. Yoshinari, M. Ohkubo, K. Fukasawa, S. Egashira, Y. Hara, M. Matsumoto, K. Nakai, H. Arakawa, H. Morishima, S. Nishimura, Mode of action of a new indolocarbazole anticancer agent, J-107088, targeting topoisomerase I, Cancer Res. 59 (1999) 4271–4275. [18] C. Ozvegy, T. Litman, G. Szakacs, Z. Nagy, S. Bates, A. Varadi, B. Sarkadi, Functional characterization of the human multidrug transporter, ABCG2, expressed in insect cells, Biochem. Biophys. Res. Commun. 285 (2001) 111–117. [19] L. Bishop, R. Agbayani Jr., S.V. Ambudkar, P.C. Maloney, G.F. Ames, Reconstitution of a bacterial periplasmic permease in
675
proteoliposomes and demonstration of ATP hydrolysis concomitant with transport, Proc. Natl. Acad. Sci. USA 86 (1989) 6953– 6957. [20] E.E. Biswas, Nucleotide binding domain 1 of the human retinal ABC transporter functions as a general ribonucleotidase, Biochemistry 40 (2001) 8181–8187. [21] D.W. Loe, R.G. Deeley, S.P.C. Cole, Characterization of vincristine transport by the M(r) 190,000 multidrug resistance protein (MRP): evidence for cotransport with reduced glutathione, Cancer Res. 58 (1998) 5130–5136. [22] S. Dey, M. Ramachandra, I. Pastan, M.M. Gottesman, S.V. Ambudkar, Evidence for two nonidentical drug-interaction sites in the human P-glycoprotein, Proc. Natl. Acad. Sci. USA 94 (1997) 10594–10599.