A-803467, a tetrodotoxin-resistant sodium channel blocker, modulates ABCG2-mediated MDR in vitro and in vivo.

ATP-binding cassette subfamily G member 2 (ABCG2) is a member of the ABC transporter superfamily proteins, which has been implicated in the development of multidrug resistance (MDR) in cancer, apart from its physiological role to remove toxic substances out of the cells. The diverse range of substrates of ABCG2 includes many antineoplastic agents such as topotecan, doxorubicin and mitoxantrone. ABCG2 expression has been reported to be significantly increased in some solid tumors and hematologic malignancies, correlated to poor clinical outcomes. In addition, ABCG2 expression is a distinguishing feature of cancer stem cells, whereby this membrane transporter facilitates resistance to the chemotherapeutic drugs. To enhance the chemosensitivity of cancer cells, attention has been focused on MDR modulators. In this study, we investigated the effect of a tetrodotoxin-resistant sodium channel blocker, A-803467 on ABCG2-overexpressing drug selected and transfected cell lines. We found that at non-toxic concentrations, A-803467 could significantly increase the cellular sensitivity to ABCG2 substrates in drug-resistant cells overexpressing either wild-type or mutant ABCG2. Mechanistic studies demonstrated that A-803467 (7.5 μM) significantly increased the intracellular accumulation of [(3)H]-mitoxantrone by inhibiting the transport activity of ABCG2, without altering its expression levels. In addition, A-803467 stimulated the ATPase activity in membranes overexpressed with ABCG2. In a murine model system, combination treatment of A-803467 (35 mg/kg) and topotecan (3 mg/kg) significantly inhibited the tumor growth in mice xenografted with ABCG2-overexpressing cancer cells. Our findings indicate that a combination of A-803467 and ABCG2 substrates may potentially be a novel therapeutic treatment in ABCG2-positive drug resistant cancers.


INTRODUCTION
Multidrug resistance (MDR) is defined as the resistance of cancer cells to antineoplastic agents that have distinct structures and mechanisms of action [1]. Particularly, cancer cells are said to have an MDR phenotype when the cells show cross-resistance to drugs that are structurally and functionally unrelated to the original drug which the cells are resistant [2]. Once MDR is acquired, the efficacy of chemotherapeutic drugs decreases. MDR is the most significant hindrance to successful chemotherapy, and is a main cause of cancer metastasis and relapse [3]. Chemotherapy resistance can be primary, which is initially determined by a refractory response to the pharmacotherapy, or secondary/acquired, which is developed during treatment progression [3].
The potential mechanisms of MDR include pharmacokinetic alterations, tumor micro-environmental changes, or cancer cell-specific factors that occur at different levels due to cellular alterations, which include increased drug efflux or decreased drug uptake, drug inactivation, drug target modification or apoptosis evasion [4]. The key effectors of drug resistance are multidrug transporters which can be elevated or hyper-activated during the genesis of drug resistant cancers. The ATPbinding cassette (ABC) multidrug transporters such as ABCB1 (MDR1/P-glycoprotein), ABCC1 (MRP1) and ABCG2 (BCRP/MXR) are considered to be accountable for the majority of drug efflux in human cancers [5]. The ABCG2 transporter is a 72-kDa half transporter, which was identified from a doxorubicin-selected MCF-7 human breast cancer cell line [6], human placenta [7], and a colon cancer cell line (S1-M1-80) [8]. ABCG2 is specifically localized at the apical surface of enterocytes, the luminal surface of liver canaliculi, the luminal surface of the proximal convoluted tubule of the kidneys, the blood-brain barrier (BBB), blood-testis barrier (BTB), blood-placental and blood-retinal barriers. Because of its localization on the secretory surface of the major organs involved in drug transport, ABCG2 alters the ADME (absorption, distribution, metabolism and elimination) of its substrate drugs. ABCG2 can transport large, hydrophobic, positively and negatively charged molecules, including cytotoxic compounds such as mitoxantrone (MX), topotecan, flavopiridol and methotrexate [9]. Although the contribution of ABCG2 in clinical MDR has not been completely investigated, some studies have described the association between ABCG2 expression and poor chemotherapeutic response. Increased ABCG2 gene expression has also been related to poor response to chemotherapy in childhood acute myeloid leukemia (AML) and relapsed AML [10,11]. ABCG2 expression has also been reported in leukemia, especially in pediatric AML [12]. In addition, increased ABCG2 mRNA has been reported in irinotecan treated hepatic metastases compared to irinotecan-naive metastases [13]. ABCG2 expression has been reported in various solid tumors, such as those present in the digestive tract, endometrium and melanoma [14].
Recently, ABCG2 has been recognized as a molecular marker for the side population (SP) cells, these are putative cancer stem cell CSC population. SP cells are identified using dual wavelength flow cytometry combined with Hoechst 33342 dye efflux [15]. For human Non-Small Cell Lung Cancer (NSCLC) cell lines, excluding 0.03 -6.1% of the tumor cells which were SP cells [16], the presence of a Hoechst dye 33342 showed elevated expression of ABCG2, an increased tumorigenicity in mice resistant to various chemotherapeutic agents [17]. Moreover, Yoh et al. found that positive immunostaining for ABCG2 appears to be a predictor of shorter survival in patients with advanced NSCLC [18]. Until now, several ABCG2 inhibitors with diverse chemical structures have been found or developed, but none of them have been tested clinically due to concerns of toxicity, safety or the pharmacokinetic uncertainty of the compounds [19].
A-803467 is a potent and selective Na v 1.8 sodium channel blocker, which has shown significant anti-nociception in animal models of neuropathic and inflammatory pain [20]. Previously, ion channel inhibitors such as verapamil and quinidine have shown to reverse ABC transporter mediated MDR [21]. We, and others, have further reported several natural drugs, marine drugs, semi-synthetic and synthetic compounds which could reverse ABCG2-mediated MDR [22][23][24][25]. Therefore, here we determine A-803467 as a therapeutic compound to enhance the chemosensitivity of conventional anticancer drugs through interaction with the ABCG2 transporter.

A-803467 significantly increases the cytotoxicity of anticancer drugs which are substrates of ABCG2, but not of ABCB1 and ABCC10
Cytotoxicity of A-803467 treatment alone on ABCG2-overexpressing cell lines was investigated and found to be nontoxic with IC 50 values greater than 10 μM (Supplementary Figure S1). Accordingly, reversal concentrations of 2.5 and 7.5 μM, at which no significant cytotoxicity was detected for A-803467 alone, were chosen for further experiments. HEK293 cells transfected with wild-type (HEK293/R482) and mutant (HEK293/ R482G and HEK293/R482T) ABCG2 (Supplementary Figure S2) showed significant resistance to MX and topotecan compared to HEK293/pcDNA3.1 ( Table 1). The test compound A-803467 at 7.5 μM significantly increased the cytotoxicity of MX and topotecan in ABCG2-transfected cell lines (Table 1). In addition, the reversal effect of A-803467 on ABCG2-mediated MDR was comparable to the effect produced by 5 μM of FTC, a known ABCG2 inhibitor. However, A-803467 did not sensitize ABCG2-transfected cells to cisplatin, a non-substrate of ABCG2 (Table 1). Furthermore, the reversal effect of A-803467 was also analyzed in parental H460, and drug selected ABCG2 overexpressing H460/ MX20 cells. We found similar results where A-803467 significantly increased the cytotoxicity of MX and topotecan in ABCG2 overexpressing H460/MX20 cells (Table 2). However, A-803467 did not sensitize the parental H460 cells to MX and topotecan (Table 2). Independently, we also analyzed the effect of A-803467 on ABCB1-and ABCC10-mediated MDR. We found www.impactjournals.com/oncotarget that A-803467 did not affect the ABCB1-and ABCC10mediated MDR in ABCB1 overexpressing HEK293/ ABCB1 cells and ABCC10 overexpressing HEK293/ ABCC10 cells, respectively (Table 3). Together these results indicate that A-803467 selectively and significantly reverses the ABCG2-mediated MDR.

A-803467 significantly augments the intracellular accumulation of [ 3 H]-MX in cells overexpressing ABCG2
To investigate the reversal mechanism, we studied the effect of A-803467 on the intracellular accumulation of [ 3 H]-MX in ABCG2 overexpressing cells. HEK293/ pcDNA3.1 and ABCG2-transfected cells were incubated with [ 3 H]-MX, a known substrate of ABCG2, with or without A-803467 at different concentrations for 2 h. A-803467 at 7.5 μM significantly enhanced the intracellular [ 3 H]-MX accumulation in ABCG2-transfected cells. However, A-803467 did not significantly impact the intracellular accumulation in HEK293/pcDNA3.1 cells (Fig. 1A). These results suggest that the increased intracellular levels of [ 3 H]-MX in ABCG2 overexpressing cells is due to the inhibitory effect of A-803467 on the drug efflux function of the ABCG2 transporter.

A-803467 decreases the efflux of [ 3 H]-MX in cells overexpressing ABCG2
We accomplished a time course [

A-803467 does not alter the total expression or translocation of ABCG2
To analyze the effect of A-803467 on the ABCG2 expression, we incubated H460/MX20 cells with A-803467 (7.5 μM) for 24, 48, and 72 h and found that there was no significant change in the expression level of ABCG2 upon A-803467 treatment ( Fig. 2A). To analyze if A-803467 causes a translocation of ABCG2 from the plasma membrane to the cytoplasm, contributing to a reduction of functional ABCG2, we performed an immunofluorescence analysis with H460 and ABCG2 overexpressing H460/MX20 cells. The results showed that the membrane expression and location of ABCG2, in H460/MX20 cells, was not altered after treatment with A-803467 (7.5 μM) for 72 h (Fig. 2B).

A-803467 stimulates the ATPase activity of ABCG2
Several reversal agents have been reported as an inhibitor and/or substrate of ABC transporters [33][34][35][36]. To determine interaction of A-803467 with ABCG2 ATPase, we performed an ATPase assay using membranes of High Five insect cells overexpressing ABCG2 with different concentrations of A-803467. A-803467 stimulated the ATPase activity of ABCG2 in a concentration dependent manner, with a maximal stimulation of 2.13-fold greater than the basal activity (Fig. 3A). The inset of Fig. 3A reveals the concentration of A-803467 required to obtain 50% stimulation is 0.718 μM. Similarly, we assessed the effect of ABCG2 known substrates, topotecan and MX, on the ATPase activity of ABCG2. We measured ABCG2mediated ATP hydrolysis in the presence of topotecan and MX at various concentrations from 0 to 10 μM. Interestingly, topotecan and MX stimulated the ATPase activity of ABCG2 in a concentration dependent manner, with a maximal stimulation of 1.81-fold and 2.04-fold greater than the basal activity, respectively ( Fig. 3B and 3C). In addition, the concentration of topotecan and MX required to obtain 50% stimulation were 2.60 μM and 2.20 μM, respectively. These results suggest A-803467 interacts at the drug-substratebinding site and stimulates the ATPase activity of ABCG2.

A-803467 potentiates the anticancer activity of topotecan in ABCG2-overexpressing tumor xenograft model
Parental H460 cells and drug resistant ABCG2 overexpressing H460/MX20 cells were implanted into athymic nude mice to create xenograft tumor models to analyze the efficacy of A-803467 to circumvent resistance to topotecan in vivo. A 35 mg/kg oral dose of A-803467 was chosen based on our preliminary study (data not shown) and showed no noticeable toxicity in the male NCR nude mice. Topotecan alone at 3 mg/kg i.p. dose demonstrated significant growth retardation in the parental H460 xenografts as well as ABCG2 overexpressing H460/ MX20 xenografts (Fig. 5A-5D). However, the H460 tumor xenografts exhibited a more dramatic reduction when compared to H460/MX20 xenografts due to lack of ABCG2 expression in H460 tumors results in increased concentration of topotecan when compared to H460/ MX20 tumors. The tumor growth rate of the xenograft mice implanted with ABCG2 overexpressing cells was significantly reduced in the A-803467-topotecan combination group as compared to the vehicle, A-803467 alone, and topotecan alone groups (Fig. 5A). Not only was the H460/MX20 tumor growth minimized, but also the size and weight of the tumors were significantly reduced in the combination treatment group (Fig. 5B and  6A). It should be mentioned that A-803467 alone had no significant effect on the growth rate of H460 ( Fig. 5C and 6B) and H460/MX20 (Fig. 5A and Fig. 6A) xenografts. Furthermore, topotecan with or without A-803467 did not produce any apparent toxicity or weight loss (Fig. 6C). Overall, A-803467 did not present any increased toxicity in the mice, yet improved the efficacy of topotecan in the ABCG2 overexpressing H460/MX20 resistant xenograft model.

DISCUSSION
ABCG2 plays an important role in development of drug resistance in clinical medicine. There is a strong correlation between ABCG2 overexpression and development of drug resistance in several cancer cells, including NSCLC, colon carcinoma, hepatocellular carcinoma (HCC) and breast cancer [37]. In the past four years, there has been rising evidence for overexpression of ABCG2 in hematologic malignancies and solid tumors; in these studies, ABCG2 overexpression significantly correlated with decreased patient survival [38]. The ABCG2 transporter is present in certain populations of cancer stem cells and normal primitive stem cells, increasing the likelihood of overexpression and thus resistance to various anticancer drugs [39][40][41]. Sensitizing these cells to anticancer drugs with the help of ABC transport reversal agents can effectively eradicate the tumor population, leading to better clinical outcomes for patients.
A wide variety of compounds that can inhibit ABCG2 have been comprehensively studied [42]. ABCG2 inhibitors include, fumitremorgin C (FTC), the FTC analogue Ko143, the acridone carboxamide derivative GF120918, anti-HIV protease inhibitors nelfinavir and ritonavir, the dietary flavonoids chrysin and biochanin A, the tyrosine kinase inhibitors gefitinib and imatinib [43], and some herbal extracts [44]. ABCG2 inhibitors such as Ko143, GF120918 and gefitinib are highly potent with their IC 50 values in nano molar range. FTC and Ko143 are highly selective in inhibiting the ABCG2 transporter, whereas rest of the compounds seems to be general inhibitors of ABC transporters [42]. The selective ABCG2 inhibitors such as FTC and Ko 143 are effective only in vitro [45]. Several ABCG2 inhibitors have been identified but none of them are in clinical use due to toxicity and pharmacokinetic uncertainty. Hence there is still an ongoing search for a safer and specific inhibitor of the ABCG2 transporter.
In order to identify novel inhibitors of ABCG2, a cell based assay using MTT in ABCG2 overexpressing H460/MX20 cells was used to screen libraries of compounds. Similar approaches have been carried out by many researchers to obtain the inhibitors specific to ABCG2 [46]. In our study, we screened several tyrosine kinase inhibitors, synthetic small molecule inhibitors [47,51] and ion channel inhibitors for activity in ABCG2 overexpressed cells and found that A-803467 was effective in inhibiting ABCG2 mediated drug resistance at micro molar concentrations.
One of the major findings of this study was that A-803467 significantly increased the sensitivity of ABCG2 overexpressing H460/MX20 cells to ABCG2 substrates such as topotecan and MX (Table 2). In addition, A-803467 did not enhance the cytotoxic effect of cisplatin, a drug that is not a substrate for the ABCG2 transporter, further demonstrating the specificity of A-803467. Moreover, previous studies have found that the 482 nd position in ABCG2 is a hot spot for mutation, Arg482 to Gly482 or Thr482 mutant variants of ABCG2 have shown to be significant in substrate specificity as well as the potency of ABCG2 antagonist [26,52]. Robey et al. further reported that the activity of the ABCG2 transporter varies in these mutant cell lines. For example, novobiocin only antagonizes wild-type ABCG2 but does not show any effect in mutant variants. However, FTC has been shown to inhibit both wild-type as well as mutant ABCG2 [26]. This study has revealed that similar to FTC, A-803467 significantly enhances the chemosensitivity of ABCG2 substrates in both the cells with wild-type Arg482 and mutant Gly482 or Thr482 of ABCG2 (Table 1). These results clearly demonstrate the A-803467 activity in the aforementioned mutants of ABCG2. Furthermore, A-803467 could not reverse ABCB1-and ABCC10mediated drug resistance in cells overexpressing ABCB1 and ABCC10 transporters (Table 3); thus implying that the reversal effect of A-803467 is ABCG2 specific.
In order to find the possible mechanism of action of A-803467, we investigated its effect on the ABCG2 expression. In this study, A-803467 at 7.5 μM did not significantly alter the expression of the ABCG2 protein in H460/MX20 cells ( Fig. 2A). In addition, A-803467 did not translocate the ABCG2 protein from the cell membrane after 72 h of treatment (Fig. 2B). This clearly demonstrates that reversal of MDR by A-803467 is unlikely due to its decreasing ABCG2 protein expression or a translocation and most likely due to the interaction with ABCG2. Further functional analysis was performed by measuring the intracellular accumulation of [ 3 H]-MX in wild-type HEK293/R482, mutant HEK293/R482T, and mutant HEK293/R482G cells (Fig. 1A). In addition, we also investigated the effect of A-803467 on the efflux of [ 3 H]-MX in wild-type HEK293/R482 cells (Fig. 1B). A-803467 at 7.5 μM produced a significant increase in accumulation of MX by inhibiting the efflux function in the forenamed cell lines but not in parental HEK293/ pcDNA3.1 cells.
To further understand the interaction of A-803467 with ABCG2, we performed an ATPase assay using ABCG2 overexpressed membranes. The majority of TKIs that interact with the ABC drug transporters stimulate ATP hydrolysis [53] and the fact that A-803467 stimulates the ATP hydrolysis of ABCG2 in a concentration dependent manner (Fig. 3A) indicates that it behaves similar to other known substrates (Fig. 3B and 3C) of ABCG2 transporter, such as MX and topotecan. These results further prove that A-803467 not only interacts directly with the ABCG2 transporter, but may also be a competitive inhibitor of the transporter.
To identify the molecular interaction of A-803467 with the ABCG2 transporter, docking simulation was performed at various sites of the human ABCG2 homology model. The crystal structure of human ABCG2 transporter is not completely elucidated. Comparing the docking scores shown in Table 4, the most favorable binding site was identified as site-1. Molecular docking of topotecan, a well-known ABCG2 substrate, at the same site of ABCG2 was performed. The docking score of topotecan (-5.57 kcal/mol) is much higher than that of A-803467 (-8.07 kcal/mol). The lower docking score indicates stronger interaction between A-803467 to ABCG2 (Fig. 4). Moreover, molecular structure of A-803467 also exhibited the pharmacophoric features such as hydrophobic groups, aromatic ring centers (phenyl ring and furan ring) and hydrogen bond acceptors that have been reported as essential for ABCG2 inhibition [54]. Overall, this molecular simulation will provide clues to optimize further derivatives of ABCG2 inhibitors.
To our knowledge, this is the first study that demonstrates the combined effect of A-803467 with anticancer drug topotecan in NCR nude mice implanted with ABCG2 overexpressing H460/MX20 cells. The dose and route of administration of A-803467 (35 mg/kg, p.o) was chosen based on our preliminary study (data not shown). A-803467 was administered on the 2 nd and 3 rd day before administration of topotecan to compensate for its high binding to plasma proteins. Jarvis et al., reported that A-803467 was highly bound (98.7%) to plasma proteins in rats [20]. The in vivo study results indicated that A-803467 in combination with topotecan, significantly decreased the tumor growth in mice implanted with ABCG2 overexpressing H460/MX20 cells (Fig 5A, 5B and 6A). A-803467 effectively restored the sensitivity of tumors overexpressing ABCG2 transporter to topotecan without having any significant effect on tumors lacking ABCG2 expression (Fig. 5C, 5D and 6B). Furthermore, A-803467 alone, or in combination with topotecan, did not produce significant observable toxicity or weight loss during the study period (Fig. 6C). The safety profile of A-803467 in humans is not investigated yet, but currently A-803467 is in pre-clinical testing stage for neuropathic and inflammatory pain [20]  B. Mean H460 tumor weight (n = 6). C. Changes in mean body weight before and after treatment for xenograft model. *: P < 0.05 versus the vehicle group; # : P < 0.05 versus the topotecan group. www.impactjournals.com/oncotarget ABCG2 is responsible for the "side population" (SP) phenotype, frequently used in the identification and isolation of cancer stem cells (CSCs). ABCG2 has also been suggested as a prognostic biomarker as well as a novel therapeutic target for the eradication of CSCs [40]. Presently, conventional chemotherapeutic anticancer agents target highly proliferative tumor cells. The CSCs survive such chemotherapy due to their high expression of ABCG2 transporter, mediating their chemoresistance, and ultimately leading to tumor relapse and metastasis. Completely eradicating cancer stem cells by overcoming the resistance to chemotherapy, mediated by ABCG2, would be a new targeted therapeutic strategy. This study suggests that combined therapy of A-803467 with ABCG2 substrates may provide a more effective way of sensitizing ABCG2-mediated MDR and possibly eliminating CSCsIn conclusion, A-803467, a tetrodotoxin resistant sodium channel blocker, effectively inhibits membrane ABCG2 function, without affecting its expression or cellular location and re-sensitizes the ABCG2 substrates in ABCG2 overexpressing cells. A-803467, even at micro molar concentrations, stimulates ATP hydrolysis of the ABCG2 transporter with strong binding interactions at the transmembrane site of the transporter. A-803467 significantly potentiates the antitumor efficacy of topotecan in tumor xenograft nude mice. Therefore, it is likely that A-803467, in combination with anticancer agents that are ABCG2 substrates, would be very useful in the treatment of certain drug resistant cancers.

Equipment
OPSYS microplate reader was purchased from Dynex Technologies (Chantilly, VA). Packard TRI-CARB1 1900CA liquid scintillation analyzer was purchased from Packard Instrument Company, Inc (Downers Grove, IL). Nikon eclipse TE2000-S fluorescence microscope was purchased from Nikon (Melville, NY).

Cell lines and cell culture
HEK293/pcDNA3.1, wild-type HEK293/R482, mutant HEK293/R482T and mutant HEK293/R482G cells were established by transfecting HEK293 cell with either the empty pcDNA3.1 vector or pcDNA3.1 vector containing a full-length ABCG2, with coding arginine (R), threonine (T), or glycine (G) at amino acid position 482, respectively, after selection with G418 and maintained in medium with 2 mg/ml of G418 [26]. HEK293/ABCB1 and HEK293/ABCC10 cell lines were generated by selection with G418 (2 mg/ml) after transfecting HEK293 cell with ABCB1 vector or ABCC10 vector, respectively [27]. The human lung cancer cell line H460, and its MX-selected derivative ABCG2-overexpressing cell line H460/MX20 were used in the study [28]. All cell lines were maintained in RPMI 1640 or DMEM medium, containing 10% fetal bovine serum and 1% penicillin/streptomycin and cultured in an incubator at 37°C with 5% CO 2 .

Cell viability assay
Cytotoxicity tests and reversal experiments were performed using the MTT colorimetric assay as described previously [29]. Cells were harvested and resuspended in a final concentration of 6 × 10 3 cells/well for HEK293/ pcDNA3.1, HEK/ABCB1, HEK/ABCC10, HEK293/R482, HEK293/R482G and HEK293/R482T cells, and 4 × 10 3 cells/well for H460 and H460/MX20 cells. Cells were seeded evenly into 96-well plates. To determine the cytotoxicity of A-803467, different concentrations of drug were added into the each well after 24 h of incubation. To determine the reversal capability of A-803467, various concentrations of chemotherapeutic drugs were added into designated wells after 2 h preincubation with A-803467, FTC, verapamil or cepharanthine. After 68 h of drug incubation, MTT reagent (4 mg/mL) was added. The plates were incubated for an additional 4 h, the supernatant was discarded and 100 μl of DMSO were added to dissolve the formazan crystals. Cell viability was measured at a wavelength of 570 nm. All the experiments were repeated at least 3 times, and the mean and standard deviation (SD) values were calculated.

[ 3 H]-MX accumulation and efflux assay
We examined the effect of A-803467 on the intracellular accumulation and efflux of [ 3 H]-MX in ABCG2-overexpressing cells as previous described [30]. Briefly, the cells (5 × 10 6 /cells) were resuspended and incubated in the RPMI 1640 medium in the presence or absence of A-803467 (7.5 μM) or FTC (5 μM) at 37°C for 2 h. Cells were then incubated with 0.01 μM [ 3 H]-MX containing medium for additional 2 h at 37°C, with or without A-803467 (7.5 μM) or FTC (5 μM), and subsequently washed twice with ice-cold PBS. For the accumulation assay, cells were lysed by the 10 mM lysis buffer (pH 7.4, containing 1% Triton X-100 and 0.2% SDS) and then placed in scintillation fluid. For the efflux assay, the suspended cells were then cultured in [ 3 H]-MX free medium, with or without A-803467 (7.5 μM) or FTC (5.0 μM) at 37°C. The aliquots of cells were harvested at the indicated times (0, 30, 60, and 120 min), and then washed with ice-cold PBS and transferred to respective scintillation vials. The radioactivity was measured using the Packard TRI-CARB1 190`A liquid scintillation analyzer.

Western blot analysis
Cell lysates were prepared as described previously [31]. Equal amounts of total cell lysates (30 μg protein) were resolved by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and electrophoretically transferred onto polyvinylidene fluoride (PVDF) membranes. After incubation in a blocking solution (5% milk) for 1 h at room temperature, the membranes were immunoblotted overnight with primary monoclonal antibodies against actin at 1:1000 dilution or ABCG2 at 1:500 dilution at 4°C, and were then further incubated for 2 h at room temperature with horseradish peroxide (HRP)-conjugated secondary antibody (1:1000 dilution). The protein-antibody complex was detected by enhanced chemiluminescence detection system (Amersham, NJ).

Immunofluorescence analysis
For immunofluorescence analysis, H460 and H460/MX20 cells were seeded in 24 well plates. Cells were incubated with or without A-803467 (7.5 μM) for 72 h. Thereafter, cells were washed with PBS and fixed with 4% paraformaldehyde for 15 min at room temperature and then rinsed with PBS three times, followed by permeabilization with 1% triton X-100 for 10 min at 4°C. Cells were again washed for three times www.impactjournals.com/oncotarget with PBS, and then blocked with 2 mg/ml of BSA for 1 h at 37°C. Fixed cells were incubated with monoclonal antibody against ABCG2 (BXP 21) (1:50) for 16 h at 4°C, followed by three washes with PBS. The cells were then further incubated with Alexa flour 488 goat antimouse IgG (1:60) for 1 h at 37°C. DAPI was used for nuclear counterstaining. Immunofluorescence images were taken with a Nikon fluorescence microscope.

ABCG2 ATPase assay
The Vi-sensitive ATPase activity of ABCG2 in the membrane vesicles of High Five insect cells was measured as previously described. The membrane vesicles (100 μg protein/ml) were incubated in ATPase assay buffer with or without 0.3 mM vanadate at 37°C for 5 min and then incubated with different concentrations of A-803467 ranging from 0 to 80 μM, topotecan, and MX (0 -30 μM), at 37°C for 3 min. The ATPase reaction was induced by the addition of 5 mM Mg-ATP, and the total volume was 0.1 mL. After incubation at 37°C for 20 min, the reactions were stopped by loading 0.1 mL of 5% SDS solution. The liberated inorganic phosphate (Pi) was measured as described previously [28]

Molecular modeling
A-803467 was prepared as ligands for docking simulation onto human ABCG2 homology model following the same protocols as previously described [32]. All grids of ABCG2 were prepared and generated as per our previous protocols [32]. Grid-1 generated using residue Arg482 as the centroid had the highest docking score; therefore, docking discussion was based on binding position of A-803467 at this site. Glide v6.0 (Schrödinger, LLC, New York, NY, 2013) docking protocol was followed with the default parameters. Top scoring conformation was used for graphical analysis. All computations were carried out on a Dell 490n dual processor with Linux OS (Ubuntu 12.04 LTS).

Animals
Athymic NCr (nu/nu) nude mice, weighing 18 to 22 g (Taconic Farms, NCRNU-M, Homozygous, Albino color), were used for the ABCG2 xenograft models. All animals were provided with sterilized water and rodent chow ad libitum and maintained with an alternating 12 h light/dark cycle. All the experiments were approved by the Institutional Animal Care & Use Committee (IACUC) of St. John's University, and were carried out in accordance with the guidelines from Animal Welfare Act and The U.S. Public Health Service.

Nude mouse MDR xenograft models
The ABCG2-overexpressing NSCLC cell H460/ MX20 xenograft mouse models were established as previously explained [41]. H460/MX20 cells (6 × 10 6 ) and H460 cells (4 × 10 6 ) were injected subcutaneously under the right and left armpit regions of the nude mice, respectively. We performed a pilot study using three different doses of A-803467 (17.5, 35 and 70 mg/kg) and we found that 35 mg/kg dose was effective in increasing the topotecan sensitivity in tumors without significantly increase toxicity, therefore 35 mg/kg dose was used throughout the following study.
The ratio of growth inhibition (IR) described previously [41] was estimated according to the formula given below.

Statistical analysis
All experiments were repeated at least three times and the differences were determined by using the oneway ANOVA followed by Newman-Keuls post hoc test for comparing multiple groups with one variable in the following experiments: cell viability assay, accumulation assay, quantification of immunoblotting and tumor weight measurement. Statistical analysis was performed by two way ANOVA followed by Bonferroni post hoc test for comparing multiple groups with more than one variable in the following experiments: efflux assay, tumor growth rate measurement and body weight measurement. Statistical analysis was performed by un-paired student t-test for comparing two groups in immunoblotting. Statistical significance was set at P < 0.05. Statistical analysis was performed using GraphPad Prism version 6.01 for Windows (GraphPad Software, La Jolla, CA).