Screening to Identify Multidrug Resistance-Associated Protein Inhibitors with Neuroblastoma-Selective Cytotoxicity

Abstract

The aim of the present study is to discover multidrug resistance-associated protein (MRP) inhibitors with neuroblastoma-selective cytotoxicity by means of fluorescence assay with a membrane-permeable fluorescent dye, Fluo-8 AM, based on our observation that gene expression of Mrp3 in neuroblastoma Neuro2a cells was remarkably higher than that in primary cultured cortical neurons, as determined by real-time PCR. Neuro2a cells showed minimal fluorescence upon incubation with Fluo-8 AM. However, blocking of Mrp3 efflux function by small interfering RNA (siRNA) transfection or inhibition with probenecid resulted in significant dye accumulation, observed as an increase of fluorescence. Interestingly, Mrp3 siRNA or probenecid treatment also resulted in increased cytotoxicity, as evidenced by decreased 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT)-reducing activity of Neuro2a, with a concomitant increase in release of lactate dehydrogenase. On the other hand, primary cultured neurons exhibited higher fluorescence intensity after incubation with Fluo-8 AM regardless of addition of probenecid. Also, probenecid only minimally affected MTT-reducing activity. Thus, probenecid showed selective cytotoxicity towards Neuro2a cells. Based on these findings, we screened a series of established therapeutic agents for ability to induce Fluo-8 accumulation in Neuro2a cells. Several uricosuric and nonsteroidal anti-inflammatory drugs were identified, and these drugs were confirmed to decrease MTT-reducing activity selectively in Neuro2a. There was a negative linear correlation between Fluo-8 accumulation and cytotoxicity of these agents. Although the compounds identified here are insufficiently potent for practical application, further screening to discover higher-affinity MRP3 inhibitors using larger chemical libraries may uncover drug candidates with potent neuroblastoma-selective cytotoxicity.

Neuroblastoma is one of the most common extracranial tumors in childhood,¹⁾ together with leukemia, brain cancer and malignant lymphoma.²⁾ The prognosis is mostly good, even at the advanced stage, if neuroblastoma is found before one year of age, whereas the prognosis is generally worse in older children.^3–5) Surgery, radiation therapy, chemotherapy and/or immunotherapy are used in the treatment of high-risk neuroblastoma,⁶⁾ but improvement of the cure rates is still desirable. In addition, chemotherapy for neuroblastoma can cause neurotoxicity,⁷⁾ which may require discontinuation of treatment.^8,9) Therefore, it is desirable to develop therapeutic agents that exhibit neuroblastoma-selective cytotoxicity via a mechanism different from those of existing drugs, with minimal neurotoxicity.

Some transporters are highly expressed in cancer cells compared to normal cells, and inhibitors of these transporters have recently been highlighted as possible anticancer drugs. For example, L-type amino acid transporter (LAT) 1 supplies essential amino acids that cannot be synthesized by intracellular metabolism in cancer cells.¹⁰⁾ LAT1 is highly expressed in various cancer cells, including hepatocellular carcinoma, breast cancer and prostatic carcinoma, whereas its expression is minimal in most normal cells.¹¹⁾ Thus, LAT1-selective inhibitors have been developed as anticancer drugs with minimal cytotoxicity to normal cells.¹²⁾ Glucose transporter (GLUT) is also highly expressed in cancer cells compared to normal cells,¹³⁾ and could be another target for cancer chemotherapy; indeed, anticancer activity of GLUT inhibitors has been reported.^14,15) Oligopeptide transporter (PEPT) 2 is highly expressed in chronic myelogenous leukemia stem cells, and pharmacological inhibition of this transporter resulted in a reduction of cellular activity.¹⁶⁾ Transporters generally regulate intracellular-to-extracellular concentration ratios of various nutrients and/or metabolites, and some of them play fundamental roles in cell homeostasis. Thus, they could be promising targets for cancer treatment, and their selective expression in cancer cells may mean that undesired cytotoxicity to normal cells would be minimal.

Transporters are categorized into two superfamilies, ATP-binding cassette (ABC) and solute carrier (SLC) transporters, which include LAT, GLUT and PEPT. ABC transporters actively exclude substrates from the intracellular space, whereas SLC transporters are generally involved in both influx and efflux transport of their substrates. Inhibition of some SLC transporters has been reported to have anticancer effects.^12,14,15,17) On the other hand, most work on ABC transporters has been focused on ABC transporters that mainly recognize xenobiotics and mediate elimination of anticancer drugs used in chemotherapy. We were interested in the possibility that inhibitors of ABC xenobiotic transporters might show selective cytotoxicity towards neuroblastoma, with minimal neurotoxicity.

In the present study, we used the mouse neuroblastoma-derived cell line Neuro2a cells and mouse primary cultured cortical neurons as models of neuroblastoma cells and neurons, respectively. Neuro2a are one of the most widely used neuroblastoma models.^18,19) On the other hand, primary cultured cortical neurons have been used to compare cytotoxicity in neuroblastoma cells and neurons.²⁰⁾ We first evaluated expression of various transporters in the two cell lines, and identified multidrug resistance-associated protein 3 (MRP3/ABCC3) as showing the greatest difference in expression level between the two cell lines among the ABC xenobiotic transporters examined. We therefore considered that inhibitors of MRP3 would be good candidates as neuroblastoma-selective cytotoxic agents. We established a fluorescence screening assay to discover MRP inhibitors among clinically used drugs, and identified several uricosuric and nonsteroidal anti-inflammatory drugs (NSAIDs). We confirmed that these agents showed neuroblastoma-selective cytotoxicity.

MATERIALS AND METHODS

Materials

Dulbecco’s modified Eagle’s medium (DMEM), poly-L-lysine and cytosine arabinoside were purchased from Sigma-Aldrich (St. Louis, MO, U.S.A.). Fetal bovine serum (FBS) was obtained from Biowest (Nuaillé, France). ISOGEN was provided by Nippon Gene (Tokyo, Japan). THUNDERBIRD SYBR qPCR Mix was supplied by TOYOBO (Osaka, Japan). Neuro2a cells were purchased from ATC C (Manassas, VA, U.S.A.). Lipofectamine RNAiMAX, Opti-MEM, small interfering RNA (siRNA) targeting the mouse mrp3 gene (siMrp3) and non-targeting (negative control) siRNA were obtained from Invitrogen (San Diego, CA, U.S.A.). Fluo-8 AM was provided by AAT Bioquest (Sunnyvale, CA, U.S.A.). All other chemicals and reagents were of the highest purity available and were purchased from commercial sources.

Animals

Pregnant ICR mice were purchased from Sankyo Labo Service Co. (Toyama, Japan). The mice were housed under pathogen-free conditions at controlled temperature (21–25°C) with a 12 h light/dark cycle. The lights remained on from 8 : 00 to 20 : 00, and food and water were available ad libitum. The animals were cared for in strict compliance with the guidelines outlined in the National Institutes of Health Guide for the Care and Use of Laboratory Animals. All animal procedures were approved by the Kanazawa University Animal Care Committee.

Cell Cultures

Primary cortical neuronal cultures were prepared according to the method of Nakamichi et al.²¹⁾ In brief, cerebral cortices from 15-d-old embryonic ICR mice were dissected and incubated with 0.25% trypsin in phosphate-buffered saline (PBS) containing 0.5% glucose at 37°C for 20 min. Cells were mechanically dissociated in culture medium by using a 1000 µL pipette tip and plated at a density of 1.5×10⁵ cells/cm² on plastic dishes coated with 7.5 µg/mL poly-L-lysine. Cortical neurons were cultured in DMEM supplemented with 5% FBS, 100 U/mL penicillin, 100 µg/mL streptomycin, 28 mM glucose, 2 mM glutamine, 5 mM N-(2-hydroxyethyl)-piperazine-N′-2-ethanesulfonic acid (HEPES), 25 µg/mL apo-transferrin, 250 ng/mL insulin, 0.5 pM β-estradiol, 1.5 nM triiodothyronine, 10 nM progesterone, 4 ng/mL sodium selenite and 50 µM putrescine for the initial 4 h at 37°C in a humidified 5% CO₂ incubator. Further culture was performed in DMEM supplemented with 100 U/mL penicillin, 100 µg/mL streptomycin, 28 mM glucose, 2 mM glutamine, 5 mM HEPES, 50 µg/mL apo-transferrin, 500 ng/mL insulin, 1 pM β-estradiol, 3 nM triiodothyronine, 20 nM progesterone, 8 ng/mL sodium selenite and 100 µM putrescine up to 3 d in vitro (DIV). During 2 to 3 DIV, cells were treated with 5 µM cytosine arabinoside for 12 h to reduce proliferating cells. Under these culture conditions, more than 99% of cells were immunoreactive with antibody against microtubule-associated protein 2.²¹⁾

Mouse neuroblastoma-derived cell line Neuro2a cells were plated on at 1.0×10⁵ cells/mL and cultured in DMEM supplemented with 10% FBS for 3 DIV. Cultures were maintained in a humidified atmosphere of 5% CO₂–95% air at 37°C. When knockdown of Mrp3 was performed, Neuro2a cells were transiently transfected with siRNA siMrp3 or negative control siRNA by using lipofectamine RNAiMAX in Opti-MEM according to the manufacturer’s instructions. Culture medium was replaced with DMEM supplemented with 10% FBS at 24 h after the transfection, and culture was continued for an additional 2 DIV.

Quantitative Reverse Transcription (RT)-PCR

Total RNA was extracted from cultured cells according to the standard ISOGEN procedure. cDNA was synthesized with oligo(dT)_12–18 primer, deoxynucleotide triphosphate mix, RT buffer and MultiScribe™ Reverse Transcriptase, and amplified on a Mx3005P (Agilent Technologies, Santa Clara, CA, U.S.A.) in a reaction mixture containing cDNA with relevant sense and antisense primers (Table 1), and THUNDERBIRD SYBR qPCR Mix. PCR reactions were initiated by template denaturation at 95°C for 15 min, followed by 40 cycles of amplification (denaturation at 95°C for 10 s, and primer annealing and extension at 60°C for 30 s). Relative quantification of expression levels of the target genes was determined by the delta-delta Ct method using transcripts of acidic ribosomal phosphoprotein P0 (36B4) as the internal standard.

Table 1. Primers Used for Real-Time PCR

Genes	Sense primers (5′→3′)	Antisense primers (3′→5′)
Abcc1	GCCCCAGTATTACTGGTCA	AATTCCCGAGGATTGGACTC
Abcc2	CAAATCCAATTCTCTACCTATGCAC	CCAGAACATGGACAGGAACC
Abcc3	GCTGAGGGTGGGGATAATCT	AGCAGTGGCCTCGTCTAAAA
Abcc4	GAGCACACGGACGAGGAG	TTGCAAGGCACACTAACTGTC
Abcc5	CTGTACCAGGGCAACACAGA	CGTCCAGAATTTGGAGATGAG
Abcc6	CATCTTGCCAGGAATCAACA	AGACAGGACAGCACCAGCTC
Abcb1a	CATTTGGCAAAGCTGGAGAG	TGCTCCGGTGGTGTTTTTAG
Abcb1b	AGTGGACCCAACAGTACTCTGAT	CAGGCACCAAAGTGAAACCT
Abcg2	TGGTTTGGACTCAAGCACAG	GGAATACCGAGGCTGATGAA
36B4	ACTGGTCTAGGACCCGAGAAG	TCCCACCTTGTCTCCAGTCT

Fluorescence Assay for Mrp3 Transport Activity

Fluorescence assay using Fluo-8, a Ca²⁺-sensitive fluorescent dye, was performed to measure Mrp3 transport activity according to the method of Nakamichi et al. with minor modifications.²¹⁾ Cultured cells were washed with recording medium containing 129 mM NaCl, 4 mM KCl, 1 mM MgCl₂, 2 mM CaCl₂, 4.2 mM glucose and 10 mM HEPES (pH 7.4) once and incubated at 37°C for 1 h in recording medium containing 0.02% Pluronic F-127 and 2 µM Fluo-8 AM, which is a membrane-permeable form of Fluo-8, in either the presence or absence of MRP inhibitors. Cultures were then washed with recording medium twice, followed by settlement for 20 min in the recording medium before observation with a confocal laser scanning microscope or measurement with a fluorescence microplate reader. The calcium ionophore A23187 at 20 µM was added 5 min before the observation or measurement in order to obtain maximal fluorescence intensity. Dye-loaded cells were monitored for fluorescence visualized with a confocal laser-scanning microscope (LSM710; Carl Zeiss, Jena, Germany) equipped with an argon laser. Fluorescence images were acquired using an excitation wavelength of 488 nm. The parameters of illumination and detection were digitally controlled to ensure that the same settings were maintained throughout the experiments. Fluorescence images were quantified as the mean of the whole screen using ImageJ software. The fluorescence intensity was also measured with a fluorescence microplate reader (MTP-880Lab; Corona Electric, Hitachinaka, Japan) at excitation and emission wavelengths of 490 and 530 nm, respectively.

3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium Bromide (MTT) Assay

Mitochondrial activity as an index of cell survival was determined by using the MTT assay according to the method of Nakamichi et al. with minor modifications.²¹⁾ In brief, cultured cells were washed once with PBS and incubated with MTT solution (0.5 mg/mL in PBS) for 1 h at 37°C. Subsequently, solubilizing solution (0.04 M HCl in isopropanol) at equivalent volume to the MTT solution was added, and the mixture was well shaken for 10 min to dissolve the formazan. The absorbance of the dissolved suspension at 570 nm was measured with a microplate reader.

Lactate Dehydrogenase (LDH) Assay

LDH released into the medium, used as an index of cytotoxicity, was determined by LDH assay according to the method of Abe et al. with minor modifications.²²⁾ In brief, the culture medium was centrifuged at 16400×g for 5 min at 4°C. After the centrifugation, 50 µL aliquots of the supernatant were collected in a 96-well dish and incubated with an equal volume of reaction mixture (2.5 mg/mL L-lactate lithium, 2.5 mg/mL β-nicotinamide adenine dinucleotide, 600 µM MTT and 100 µM 1-methoxyphenazine methosulfate diluted with 200 mM Tris–HCl, pH 8.2) for 15 min at 37°C. The reaction was stopped by adding 100 µL of stop buffer (50% N,N-dimethylformamide (DMF)/20% sodium dodecyl sulfate (SDS), pH 4.7). The absorbance was measured with a microplate reader at a test wavelength of 570 nm and a reference wavelength of 655 nm.

Statistical Analysis

All experiments were performed at least three times. Data are expressed as the mean±standard error of the mean (S.E.M.). The statistical significance of differences was determined by means of Student’s t-test. Correlation was assessed with nonparametric Spearman’s rank correlation test. p<0.05 was regarded as denoting a significant difference.

RESULTS

Expression Profiles of ABC Transporters in Neuroblastoma Cells and Neurons

In order to identify ABC xenobiotic transporter(s) expressed at higher levels in neuroblastoma cells compared to neurons, we first examined the mRNA expression levels of various ABC xenobiotic transporters in mouse primary cultured cortical neurons and mouse neuroblastoma-derived cell line Neuro2a cells. Expression levels of mrp1, mrp3, mdr1a and mdr1b in Neuro2a cells were significantly higher than those in primary cultured neurons (Fig. 1). In particular, expression levels of mrp3 and mdr1b in Neuro2a cells were about 70 and 50 times higher than in primary cultured neurons, respectively (Fig. 1). The expression levels of mrp4, mrp5 and bcrp in Neuro2a cells were significantly lower than those in primary cultured neurons, while expression levels of mrp2 and mrp6 were under the quantitative detection limit (<60 copies/µg total RNA) in both cells (Fig. 1). Thus, we focused on mrp3 as a target ABC xenobiotic transporter for obtaining neuroblastoma-selective toxicity.

Fig. 1. Expression Profiles of ABC Xenobiotic Transporters in Neuro2a Cells and Primary Cultured Cortical Neurons

Total RNA was extracted from Neuro2a cells (black columns) and cortical neurons (white columns) for quantitative RT-PCR analysis. Each value was normalized by the expression level of 36B4 mRNA and expressed as a relative value to the expression of Mrp1 in cortical neurons. Each value represents the mean±S.E.M. (n=3–6). * p<0.05, Significant difference from the corresponding mRNA level in cortical neurons. N.D., not detectable, means that samples were not amplified until the end of the PCR reactions.

Knockdown of Mrp3 in Neuro2a Cells

To examine whether inhibition of Mrp3 induces cellular toxicity in neuroblastoma, we performed knockdown of Mrp3 with siRNA in Neuro2a cells. The mRNA expression of mrp3 was markedly reduced at 48 h after transfection of siMrp3 in Neuro2a cells, compared to non-specific siRNA-transfected cells (Fig. 2A). To confirm that the knockdown of Mrp3 inhibits its transport function, we next assessed accumulation of fluorescent dye Fluo-8. The number of cells with marked fluorescence in the siMRP3-treated group was larger than in the negative control group (Fig. 2B). Quantification of the fluorescence intensity of the cells with Image software showed that fluorescence in the siMrp3-treated group was much higher than that in the negative control group (Fig. 2C). Then, to establish whether knockdown of Mrp3 induces cytotoxicity in Neuro2a cells, we performed MTT and LDH assays of siMrp3-transfected cells. Knockdown of Mrp3 significantly decreased MTT-reducing activity of Neuro2a cells (Fig. 2D), with a concomitant increase of LDH release into the culture medium (Fig. 2E). Thus, inhibition of Mrp3 suppresses efflux of Fluo-8 and induces cytotoxicity in Neuro2a cells.

Fig. 2. Effects of Mrp3 Knockdown on Transport Activity and Viability of Neuro2a Cells

Neuro2a cells were transiently transfected with siMrp3 (black columns) or negative control siRNA (white columns) for 24 h, and cultured for an additional period of 48 h. (A) Total RNA was extracted, followed by measurement of Mrp3 expression level by means of quantitative RT-PCR. Each value is normalized by the expression level of 36B4 mRNA and expressed as a relative value to the control obtained in cells transfected with negative control siRNA. Each value represents the mean±S.E.M. (n=4). (B) Neuro2a cells were preincubated with 2 µM Fluo-8 AM at 37°C for 1 h and subsequently exposed to 20 µM A23187 for 5 min to achieve maximal fluorescence intensity. The experiment was repeated in 20 wells of 4 independent preparations, and typical fluorescence images are shown. (C) Fluorescence images obtained in the experiment of panel (B) were quantified using ImageJ software and normalized by those in control siRNA. Scale bar: 200 µm. For determination of cytotoxicity, mitochondrial activity (D) and LDH released into the medium (E) were measured by means of MTT and LDH assays, respectively. Each value represents the mean±S.E.M. (n=4). * p<0.05, Significant difference from the control value obtained in cells transfected with negative control siRNA.

Selective Toxicity of Mrp3 Inhibition in Neuro2a Cells with Minimal Neurotoxicity

We next compared the functional expression of Mrp3 between neuroblastoma and neurons in order to obtain insight into the selectivity of cytotoxicity provoked by Mrp3 knockdown. Fluorescence intensity of Fluo-8 after preincubation of Neuro2a cells and primary cultured neurons with Fluo-8 AM was measured in the presence or absence of the Mrp3 inhibitor probenecid. In Neuro2a cells, fluorescence intensity was minimal in the absence of probenecid (Fig. 3A), but exposure to probenecid markedly increased the number of cells showing high fluorescence (Fig. 3B). On the other hand, many cells with high fluorescence intensity were observed in primary cultured neurons, regardless of the addition of probenecid (Figs. 3C, D). Quantitative analysis was performed for these data, and the result was shown in Fig. 3E, revealing that fluorescence intensity in Neuro2a cells was remarkably increased by the addition of probenecid whereas fluorescence intensity in cultured neurons was scarcely affected by probenecid (Fig. 3E). These results suggest that functional activity of Mrp3 is much greater in Neuro2a cells than in primary cultured neurons.

Fig. 3. Functional Expression of Mrp3 in Neuro2a Cells, but Not in Primary Cultured Cortical Neurons

Neuro2a cells (A, B) or cortical neurons (C, D) were preincubated with 2 µM Fluo-8 AM at 37°C for 1 h in either the presence (B, D) or absence (A, C) of 2.5 mM probenecid, and subsequently exposed to 20 µM A23187 for 5 min. The experiment was repeated in independent preparations, and typical fluorescence images are shown. Scale bar: 200 µm. (E) Fluorescence images obtained in the experiment of panel (A–D) were quantified using ImageJ software. Each value represents the mean±S.E.M. (n=4). * p<0.05, Significant difference from the control value obtained in the absence of probenecid.

Next, we attempted to establish whether inhibition of Mrp3 function using Mrp inhibitor drugs results in selective cytotoxicity to neuroblastoma cells. MTT-reducing activity was examined after exposure of Neuro2a cells and primary cultured neurons to various concentrations of Mrp inhibitor, probenecid or sulfinpyrazone, for 8 h in the absence of Fluo-8 (Figs. 4A, B). The addition of probenecid (Fig. 4A) or sulfinpyrazone (Fig. 4B) at a concentration of 1000 µM or more decreased the MTT-reducing activity in Neuro2a cells, but had a minimal effect in primary cultured neurons, suggesting that Mrp inhibitors are more cytotoxic to neuroblastoma cells than to neurons. On the other hand, the addition of probenecid (Fig. 4C) or sulfinpyrazone (Fig. 4D) in the presence of Fluo-8 for 1 h increased the fluorescence intensity of Fluo-8 in Neuro2a cells in a dose-dependent manner, and the fluorescence intensity at concentrations inducing selective cytotoxicity (ca. 1000 µM) was about 10–20 times higher than that in the absence of the inhibitor (Figs. 4C, D). These results suggest that inhibition of Mrp3 may be selectively cytotoxic to neuroblastoma cells, with minimal neurotoxicity, and the fluorescence intensity of Fluo-8 accumulated in Neuro2a cells may be a surrogate parameter of the cytotoxicity to neuroblastoma.

Fig. 4. Selective Cytotoxic Effect in Neuro2a Cells and Inhibitory Effect on Fluo-8 Efflux of Probenecid and Sulfinpyrazone

(A, B) Neuro2a cells (●) and cortical neurons (□) were exposed to probenecid (A) or sulfinpyrazone (B) for 8 h, and mitochondrial activity was measured by means of MTT assay and normalized by that in their absence. Each value represents the mean±S.E.M. (n=4). * p<0.05, Significant decrease from the corresponding control value obtained in non-exposure cells. (C, D) Neuro2a cells were preincubated with 2 µM Fluo-8 AM at 37°C for 1 h in the absence or presence of probenecid (C) or sulfinpyrazone (D), and subsequently exposed to 20 µM A23187 for 5 min. Fluorescence intensity was measured with a microplate reader and normalized by that obtained in the control. Each value represents the mean±S.E.M. (n=4).

Fluorescence Screening of Mrp3-Inhibitory Activity to Find Drugs Showing Selective Cytotoxicity in Neuro2a Cells

In order to find drugs with selective toxicity towards Neuro2a cells among clinically used drugs, we performed Fluo-8 accumulation assay in the presence of 75 candidate agents at various concentrations. All the candidate agents except MK571 are clinically used therapeutic agents, and include organic acids that would potentially be recognized by Mrps (Table 2). The fluorescence intensity of Fluo-8 in the presence of each drug was evaluated as a percentage of that in the presence of 2.5 mM probenecid (Table 2). Diclofenac sodium, flubiprofen, mefenamic acid, sulfinpyrazone and etodolac at 1 mM, and benzbromarone at 50 µM increased the fluorescence intensity to a level comparable to that in the presence of 2.5 mM probenecid (Table 2). In addition to these drugs, many uricosuric drugs and NSAIDs increased the fluorescence intensity to over 50% of that in the presence of 2.5 mM probenecid (Table 2).

Table 2. Screening of Mrp3 Inhibitors by Evaluation of Fluo-8 Accumulation in Neuro2a Cells

		Candidate agents
Fluorescence intensitya) (% of 2.5 mM probenecid)	>100%b)	Diclofenac sodium*, Flurbiprofen*, Mefenamic acid*, Sulfinpyrazone*
	75–100%b)	Benzbromarone*, Etodolac*
	50–75%b)	Cyclosporin A, Deoxycholic acid, Fenofibrate, Glibenclamide, MK571, Phenylbutazone*, Prednisolone, Quinidine, Ketoprofen*, Sulindac*
	0–50%b)	Acemetacin*, Acetylsalicylic acid*, Amantadine hydrochloride, Ampiroxicam*, Amoxycillin, Antipyrine, Benzylpenicillin potassium, Bumetanide, Caffein, Cefazolin sodium, Cefoperazone sodium, Cefalexin, Cefsulodin, Cimetidine, Cefmetazole sodium, Clarithromycin, Cyclophosphamide, Dipyridamole, Erythromycin, Famotidine, Fenbufen*, Furosemide, Glutathione, Hypoxanthine, Ibuprofen*, Imipramine hydrochloride, Indometacin*, Itraconazole, Ketoconazole, L-carnitine, Nabumetone*, Naproxen*, Nifedipine, Oxaprozine*, Phenytoin, P-aminohippuric acid, Piroxicam*, Quercetin, Quinine, Ranitidine, Rifampicin, Salicylate*, Salicylic acid*, Sodium cholate, Sodium hippurate, Sodium valproate, Streptomycin sulfate, Sulfasalazine, Taurin, Tartaric acid, Taurocholic acid sodium, Testosterone, Tetracycline, Theophylline, Thyroxine, Torsemide, Tolmetin*, Vancomycin, Verapamil

a) Determined with a microplate reader and normalized by the value in the presence of 2.5 mM probenecid. b) Concentration of most drugs was set to be 10–1000 µM, except for MK571, nabumetone, fenbufen, ampiroxicam, oxaprozin, and acemetacin (1–100 µM), salicylic acid and acetylsalicylic acid (100–2500 µM), and benzbromarone (5–50 µM). The maximum effect of each compound is indicated in the table. *: Uricosuric drugs or NSAIDs.

We next focused on the uricosuric drugs and NSAIDs. Neuro2a cells were exposed to them at various concentrations for 8 h in the absence of Fluo-8, and the MTT-reducing activity was measured. The MTT-reducing activity was negatively correlated with fluorescence intensity observed in Fluo-8 accumulation assay, with a correlation coefficient of −0.544 (p<0.05; Fig. 5), suggesting an association between Mrp inhibition by these agents and the cytotoxic effects in Neuro2a cells. In particular, 2.5 mM probenecid, 1 mM sulfinpyrazone, sulindac, etodolac, phenylbutazone, flurbiprofen, mefenamic acid and dicrofenac sodium, and 50 µM benzbromarone greatly increased fluorescence intensity and decreased MTT-reducing activity (Fig. 5). Thus, these drugs are candidate agents to exhibit neuroblastoma-selective cytotoxicity.

Fig. 5. Correlation between Fluorescence Intensity of Fluo-8 and MTT-Reducing Activity in Neuro2a Cells

Neuro2a cells were preincubated with 2 µM Fluo-8 AM at 37°C for 1 h in the presence of various concentrations of uricosuric drugs and NSAIDs, and subsequently exposed to 20 µM A23187 for 5 min. Fluorescence intensity was measured with a microplate reader and shown in the abscissa. On the other hand, mitochondrial activity was measured by MTT assay after exposure to the same concentration of the same drugs for 8 h in the absence of Fluo-8-AM and shown in ordinate. The drugs that showed over 50% of the fluorescence intensity observed in the presence of 2.5 mM probenecid are indicated.

Selective Cytotoxicity of Clinically Used Drugs to Neuroblastoma

To establish whether the candidate drugs exhibit neuroblastoma-selective cytotoxicity, primary cultured neurons were also exposed to various concentrations of the NSAIDs selected in the screening assay for 8 h, and the MTT-reducing activity was compared to that in Neuro2a cells (Fig. 6). Exposure to 1000 µM sulindac (Fig. 6A), 100 µM mefenamic acid (Fig. 6B), 100 and 1000 µM flurbiprofen (Fig. 6C) and 1000 µM phenylbutazone (Fig. 6E) significantly decreased the MTT-reducing activity of Neuro2a cells without causing a significant decrease in that of primary cultured neurons. MTT-reducing activity after exposure of Neuro2 cells to 1000 µM mefenamic acid (Fig. 6B) was lower than that of cultured neurons, but this drug and etodolac at 1000 and 2500 µM decreased the MTT-reducing activity of not only Neuro2a, but also cultured neurons (Figs. 6B, D). The MTT-reducing activity of primary cultured neurons was significantly increased at some concentrations of sulindac and etodolac (Figs. 6A, D).

Fig. 6. Cytotoxicity in Neuro2a Cells and Primary Cultured Cortical Neurons of Mrp3 Inhibitor Drugs

Neuro2a cells and cortical neurons were exposed to sulindac (A), mefenamic acid (B), flurbiprofen (C), etodolac (D) and phenylbutazone (E) for 8 h, and mitochondrial activity was measured by means of MTT assay and normalized by that in their absence. Each value represents the mean±S.E.M. (n=4). * p<0.05, Significant difference from each control value obtained in the absence of inhibitors.

DISCUSSION

We found here that siRNA knockdown of Mrp3 resulted in selective cytotoxicity of neuroblastoma Neuro2a cells, compared to primary cultured cortical neurons (Fig. 2). Further, inhibition of Mrp3 function by established therapeutic agents, such as probenecid, sulfinpyrazone, sulindac, mefenamic acid, flurbiprofen and phenylbutazone, at appropriate concentrations resulted in selective cytotoxicity to Neuro2a cells with minimal toxicity to primary cultured neurons (Figs. 4, 6). Our results indicate that this ABC transporter may be a promising target molecule for neuroblastoma treatment. Indeed, MRP1, 3 and 4 are highly expressed in rodent and human neuroblastoma and are involved in efflux of anticancer drugs.^23,24) High expression of MRP1 or MRP4 was reported to be associated with a poor prognosis of neuroblastoma.^17,25) Henderson et al. also proposed that MRPs may be involved in exacerbation of neuroblastoma via unknown mechanisms in humans, and administration of the MRP1-selective inhibitor reversan (a lead compound of pyrazolopyrimidines, a prominent structural class of potent MRP1 inhibitors) suppressed growth of implanted neuroblastoma in mice.¹⁷⁾ MRP3 mediates efflux transport of various endogenous compounds and metabolites, including leukotriene C4, glycocholic acid and estradiol glucuronide,²⁶⁾ from intracellular to extracellular space. Therefore, its inhibition by siRNA or inhibitor drugs may result in accumulation of various compounds in the intracellular space, and this in turn may result in cytotoxicity as shown in the present study. Although we demonstrated using mouse neuroblastoma that inhibition of Mrp3 by uricosuric drugs and NSAIDs induced cytotoxicity in the present study, MRP3 is also expressed in human neuroblastoma, and its transport activity can be similarly inhibited by uricosuric drugs and NSAIDs.^24,25,27) Mouse and human MRP3 share more than 80% homology on a genetic level, and have several common substrates and inhibitors.^27,28) Thus, MRPs are candidate targets for neuroblastoma treatment. However, further studies are needed to clarify the mechanism underlying the selective toxicity of MRP3 inhibition to neuroblastoma.

In this work, we established a fluorescence assay using Fluo-8 for screening Mrp3 inhibitors among candidate agents. Fluo-8 was simply used as a probe for evaluation of Mrp3 function, but not expected to induce cytotoxicity by its accumulation. Fluorescent dyes Fura-2 and Fluo-3 are transported by MRPs, but it is not yet established whether MRPs transport Fluo-8. We chose Fluo-8 for the screening system to detect Mrp3 inhibitors on the basis that it is a structural analog of Fluo-3 and its fluorescence intensity is about 4 times that of Fluo-3. Its suitability was confirmed by the observations that addition of the MRPs inhibitor probenecid or knockdown of Mrp3 with siRNA greatly increased the fluorescence intensity of Fluo-8 in Neuro2a cells preincubated with Fluo-8 AM (Figs. 2, 3). These results show that Mrp3 is a major efflux transporter of Fluo-8 in Neuro2a cells. This is interesting, because Fluo-8 is a cationic fluorescent dye, whereas Mrp3 preferentially transports organic acids.

Among inhibitors identified with the fluorescence assay, uricosuric drugs and NSAIDs induced a marked increase of fluorescence in Neuro2a cells, and we found a negative correlation (r=−0.544, p<0.05) between the fluorescence intensity of Fluo-8-loaded Neuro2a cells and MTT-reducing activity (Fig. 5). This result seems reasonable, because inhibition of Mrp function would lead to accumulation of cytotoxic products in these cells. Further screening for more potent inhibitors of Mrp3 might identify candidate therapeutic agents with even more selective cytotoxicity towards neuroblastoma. The strategy of high-throughput fluorescence screening using a transporter-specific fluorescent substrate might also be applicable to diseases other than neuroblastoma or cancer, if ABC xenobiotic transporters are possible molecular targets in those cases.

Probenecid is widely used in basic cytohistological experiments with the fluorescent dye Fura-2 to prevent leakage of the dye from the cells. This is consistent with the present observation that probenecid increases accumulation of Fluo-8 in Neuro2a cells (Fig. 3). In the present study, we found that several uricosuric drugs and NSAIDs inhibited efflux of Fluo-8 as potently as did 2.5 mM probenecid (Table 2). It has already been reported that these drugs inhibit MRP2 or MRP4.^29,30) Also, NSAIDs such as salicylate, piroxicam, ibuprofen, naproxen, sulindac, tolmetin, etodolac, dicrofenac, indomethacin, ketoprofen, phenylbutazone and celecoxib inhibit MRP1, MRP2 and/or MRP4.^30,31) MRP2 and MRP3 share 48% homology at the amino acid level,³²⁾ and have several common inhibitors.³³⁾ Thus, it is not unexpected that uricosuric drugs and NSAIDs also inhibit Mrp3 (Figs. 4, 6, Table 2). There is no report of side effects of uricosuric drugs on the nervous system, and this is consistent with the present finding that Mrp3 inhibition by uricosuric drugs may selectively suppress growth of neuroblastoma with minimal neuronal cytotoxicity. On the other hand, the pharmacological target of NSAIDs is cyclooxygenase (COX), and cytotoxicity of NSAIDs in Neuro2a cells may be partly caused by inhibition of this enzyme. However, NSAIDs at higher concentrations showed cytotoxicity to not only Neuro2a cells, but also primary cultured neurons (Fig. 6), in accordance with a previous report on neurotoxicity to hippocampal neurons,³⁴⁾ suggesting that inhibition of COX may induce nonspecific cytotoxicity. Thus, the present fluorescent screening system may also detect drugs with cytotoxicity via the mechanisms other than direct inhibition of Mrp3, but it would be beneficial as the first screening system to find candidate drugs which may possess selective cytotoxicity to neuroblastoma with minimal neurotoxicity.

In the present study, we searched for Mrp3 inhibitors among clinically used drugs. Such a drug-repositioning strategy can greatly reduce the cost of drug development. However, the cytotoxicity of the selected therapeutic drugs to Neuro2a cells was only observed at much higher concentrations than the clinically relevant ones. For example, the plasma unbound concentrations of probenecid, sulfinpyrazone, sulindac, mefenamic acid, flurbiprofen and phenylbutazone are 0.02–20 µM in the clinical situation (Table 3), whereas selective cytotoxicity in Neuro2a cells was only observed at concentrations over 100 or 1000 µM (Figs. 4, 6, Table 3). On the other hand, various types of chemical libraries are currently available, and they have been used to explore potent MRP inhibitors.^35,36) Thus, further screening of high-affinity MRP3 inhibitors using larger chemical libraries may lead to the discovery of better candidates for agents with highly neuroblastoma-selective cytotoxicity.

Table 3. Comparison of Drug Concentration Exerting Selective Cytotoxicity in Neuro2a Cells with the Clinically Relevant Concentration

Drugs	Concentration exerting selective cytotoxicity in Neuro2a cellsa)	Maximum unbound concentration in plasmab)
Probenecid	1000 µM	ca. 13 µMc)
Sulfinpyrazone	1000 µM	ca. 0.52 µMd)
Sulindac	1000 µM	ca. 0.70 µMd)
Mefenamic acid	100 µM	ca. 19 µMd)
Flurbiprofen	100 µM	ca. 0.023 µMd)
Phenylbutazone	1000 µM	ca. 0.050 µMd)

a) Obtained in the present study. b) Calculated according to the Interview Form of each drug. c) The product of average concentration after repeated administration of the highest dose available, and the plasma unbound fraction. d) The product of maximum concentration after single administration of the highest dose available, and the plasma unbound fraction.

Acknowledgments

The authors would like to thank Ms. Lica Ishida in Kanazawa University for technical assistance. This study was in part supported by Grants-in-Aid for Scientific Research to N.N. [No. 25460092] and Y.K. [No. 24390040] from the Ministry of Education, Culture, Sports, Science and Technology of Japan.

Conflict of Interest

The authors declare no conflict of interest.

REFERENCES

• 1) Brodeur GM. Neuroblastoma: biological insights into a clinical enigma. Nat. Rev. Cancer, 3, 203–216 (2003).
• 2) Smith MA, Seibel NL, Altekruse SF, Ries LA, Melbert DL, O’Leary M, Smith FO, Reaman GH. Outcomes for children and adolescents with cancer: challenges for the twenty-first century. J. Clin. Oncol., 28, 2625–2634 (2010).
• 3) London WB, Castleberry RP, Matthay KK, Look AT, Seeger RC, Shimada H, Thorner P, Brodeur G, Maris JM, Reynolds CP, Cohn SL. Evidence for an age cutoff greater than 365 days for neuroblastoma risk group stratification in the Children’s Oncology Group. J. Clin. Oncol., 23, 6459–6465 (2005).
• 4) Simon T, Hero B, Faldum A, Handgretinger R, Schrappe M, Niethammer D, Berthold F. Infants with stage 4 neuroblastoma: the impact of the chimeric anti-GD2-antibody ch14.18 consolidation therapy. Klin. Padiatr., 217, 147–152 (2005).
• 5) Isaacs H Jr. Fetal and neonatal neuroblastoma: retrospective review of 271 cases. Fetal Pediatr. Pathol., 26, 177–184 (2007).
• 6) Heczey A, Louis CU. Advances in chimeric antigen receptor immunotherapy for neuroblastoma. Discov. Med., 16, 287–294 (2013).
• 7) Mano K. Countermeasures to neurological adverse reactions of chemotherapy. Gan To Kagaku Ryoho, 30, 779–786 (2003).
• 8) Wiernik PH, Schwartz EL, Strauman JJ, Dutcher JP, Lipton RB, Paietta E. Phase I clinical and pharmacokinetic study of taxol. Cancer Res., 47, 2486–2493 (1987).
• 9) Vahdat L, Papadopoulos K, Lange D, Leuin S, Kaufman E, Donovan D, Frederick D, Bagiella E, Tiersten A, Nichols G, Garrett T, Savage D, Antman K, Hesdorffer CS, Balmaceda C. Reduction of paclitaxel-induced peripheral neuropathy with glutamine. Clin. Cancer Res., 7, 1192–1197 (2001).
• 10) Yanagida O, Kanai Y, Chairoungdua A, Kim DK, Segawa H, Nii T, Cha SH, Matsuo H, Fukushima J, Fukasawa Y, Tani Y, Taketani Y, Uchino H, Kim JY, Inatomi J, Okayasu I, Miyamoto K, Takeda E, Goya T, Endou H. Human L-type amino acid transporter 1 (LAT1): characterization of function and expression in tumor cell lines. Biochim. Biophys. Acta, 1514, 291–302 (2001).
• 11) Kim DK, Kanai Y, Choi HW, Tangtrongsup S, Chairoungdua A, Babu E, Tachampa K, Anzai N, Iribe Y, Endou H. Characterization of the system L amino acid transporter in T24 human bladder carcinoma cells. Biochim. Biophys. Acta, 1565, 112–122 (2002).
• 12) Fukumoto S, Hanazono K, Komatsu T, Ueno H, Kadosawa T, Iwano H, Uchide T. L-Type amino acid transporter 1 (LAT1): a new therapeutic target for canine mammary gland tumour. Vet. J., 198, 164–169 (2013).
• 13) Reske SN, Grillenberger KG, Glatting G, Port M, Hildebrandt M, Gansauge F, Beger HG. Overexpression of glucose transporter 1 and increased FDG uptake in pancreatic carcinoma. J. Nucl. Med., 38, 1344–1348 (1997).
• 14) Zhan T, Digel M, Küch EM, Stremmel W, Füllekrug J. Silybin and dehydrosilybin decrease glucose uptake by inhibiting GLUT proteins. J. Cell. Biochem., 112, 849–859 (2011).
• 15) Watanabe M, Abe N, Oshikiri Y, Stanbridge EJ, Kitagawa T. Selective growth inhibition by glycogen synthase kinase-3 inhibitors in tumorigenic HeLa hybrid cells is mediated through NF-κB-dependent GLUT3 expression. Oncogenesis, 1, e21 (2012).
• 16) Naka K, Jomen Y, Ishihara K, Kim J, Ishimoto T, Bae E, Mohney R, Stirdivant SM, Oshima H, Oshima M, Kim DW, Nakauchi H, Takihara Y, Kato Y, Ooshima A, Kim SJ. Dipeptide species regulate p38MAPK-Smad3 signalling to maintain chronic myelogenous leukaemia stem cells. Nat. Commun., 6, 8039 (2015).
• 17) Henderson MJ, Haber M, Porro A, Munoz MA, Iraci N, Xue C, Murray J, Flemming CL, Smith J, Fletcher JI, Gherardi S, Kwek CK, Russell AJ, Valli E, London WB, Buxton AB, Ashton LJ, Sartorelli AC, Cohn SL, Schwab M, Marshall GM, Perini G, Norris MD. ABCC multidrug transporters in childhood neuroblastoma: clinical and biological effects independent of cytotoxic drug efflux. J. Natl. Cancer Inst., 103, 1236–1251 (2011).
• 18) Chakrabarti L, Wang BD, Lee NH, Sandler AD. A mechanism linking Id2-TGFβ crosstalk to reversible adaptive plasticity in neuroblastoma. PLoS ONE, 8, e83521 (2013).
• 19) Sugantha Priya E, Selvakumar K, Bavithra S, Elumalai P, Arunkumar R, Raja Singh P, Brindha Mercy A, Arunakaran J. Anti-cancer activity of quercetin in neuroblastoma: an in vitro approach. Neurol. Sci., 35, 163–170 (2014).
• 20) Jakubowicz-Gil J, Rzeski W, Zdzisińska B, Piersiak T, Weiksza K, Glowniak K, Gawron A. Different sensitivity of neurons and neuroblastoma cells to quercetin treatment. Acta Neurobiol. Exp. (Wars), 68, 463–476 (2008).
• 21) Nakamichi N, Taguchi T, Hosotani H, Wakayama T, Shimizu T, Sugiura T, Iseki S, Kato Y. Functional expression of carnitine/organic cation transporter OCTN1 in mouse brain neurons: possible involvement in neuronal differentiation. Neurochem. Int., 61, 1121–1132 (2012).
• 22) Abe K, Matsuki N. Measurement of cellular 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) reduction activity and lactate dehydrogenase release using MTT. Neurosci. Res., 38, 325–329 (2000).
• 23) Matsuoka Y, Okazaki M, Kitamura Y, Taniguchi T. Developmental expression of P-glycoprotein (multidrug resistance gene product) in the rat brain. J. Neurobiol., 39, 383–392 (1999).
• 24) Yu DM, Huynh T, Truong AM, Haber M, Norris MD. ABC transporters and neuroblastoma. Adv. Cancer Res., 125, 139–170 (2015).
• 25) Haber M, Smith J, Bordow SB, Flemming C, Cohn SL, London WB, Marshall GM, Norris MD. Association of high-level MRP1 expression with poor clinical outcome in a large prospective study of primary neuroblastoma. J. Clin. Oncol., 24, 1546–1553 (2006).
• 26) Slot AJ, Molinski SV, Cole SP. Mammalian multidrug-resistance proteins (MRPs). Essays Biochem., 50, 179–207 (2011).
• 27) Zelcer N, Saeki T, Reid G, Beijnen JH, Borst P. Characterization of drug transport by the human multidrug resistance protein 3 (ABCC3). J. Biol. Chem., 276, 46400–46407 (2001).
• 28) Borst P, de Wolf C, van de Wetering K. Multidrug resistance-associated proteins 3, 4, and 5. Pflugers Arch., 453, 661–673 (2007).
• 29) Horikawa M, Kato Y, Sugiyama Y. Reduced gastrointestinal toxicity following inhibition of the biliary excretion of irinotecan and its metabolites by probenecid in rats. Pharm. Res., 19, 1345–1353 (2002).
• 30) El-Sheikh AA, van den Heuvel JJ, Koenderink JB, Russel FG. Interaction of nonsteroidal anti-inflammatory drugs with multidrug resistance protein (MRP) 2/ABCC2- and MRP4/ABCC4-mediated methotrexate transport. J. Pharmacol. Exp. Ther., 320, 229–235 (2007).
• 31) O’Connor R. The pharmacology of cancer resistance. Anticancer Res., 27 (3A), 1267–1272 (2007).
• 32) Keppler D. The roles of MRP2, MRP3, and OATP1B1 and OATP1B3 in conjugated hyperbilirubinemia. Drug Metab. Dispos., 42, 561–565 (2014).
• 33) Wang XB, Wang SS, Zhang QF, Liu M, Li HL, Liu Y, Wang JN, Zheng F, Guo LY, Xiang JZ. Inhibition of tetramethylpyrazine on P-gp, MRP2, MRP3 and MRP5 in multidrug resistant human hepatocellular carcinoma cells. Oncol. Rep., 23, 211–215 (2010).
• 34) Baik EJ, Kim EJ, Lee SH, Moon C. Cyclooxygenase-2 selective inhibitors aggravate kainic acid induced seizure and neuronal cell death in the hippocampus. Brain Res., 843, 118–129 (1999).
• 35) Duan Z, Choy E, Hornicek FJ. NSC23925, identified in a high-throughput cell-based screen, reverses multidrug resistance. PLoS ONE, 4, e7415 (2009).
• 36) Iram SH, Gruber SJ, Raguimova ON, Thomas DD, Robia SL. ATP-binding cassette transporter structure changes detected by intramolecular fluorescence energy transfer for high-throughput screening. Mol. Pharmacol., 88, 84–94 (2015).