Organic cation transporter 3 mediates cisplatin and copper cross-resistance in hepatoma cells.

Platinum-based drugs are first-line compounds in the treatment of many solid cancers. Major obstacles are tumors that become resistant and toxic side effects, both largely due to the expression of transporters that mediate the cellular processing of platinum. In this study, we addressed the establishment of cisplatin resistance in the absence of copper transporter ATP7B that has been previously found to be overexpressed in various resistant cells. Cisplatin sensitivity, induction of apoptosis, drug accumulation, and transporter gene expression were determined in hepatoma cell lines. Knockout or overexpression of copper transporter ATP7B did not affect cisplatin sensitivity. Cisplatin resistant cells showed a stably reduced cisplatin accumulation and a downregulation of organic cation transporter 3 (OCT3). In contrast, OCT3 overexpression could reverse resistance. Reduced MT1 expression was detected in the resistant cell line, however transient and highly dependent on the presence of cisplatin. Cross-resistance to copper was also associated with OCT3 downregulation. Our results suggest that a decreased level of OCT3 expression results in resistance to cisplatin and copper. OCT3 may represent a novel target for improved prognosis and anticancer therapy, including HCC.

INTRODUCTION

For decades, cisplatin (Cp), a platinum-based anticancer agent, is widely used in the clinic for therapy of many solid tumors, including lung, ovary, testes, prostate, breast, colorectal, and liver [1, 2]. Cisplatin has also been suggested to reduce the recurrence rate after curative surgical resection of hepatocellular carcinoma (HCC), one of the most common cancers in the world [3]. Once within the cell, the activated platinum reacts with purine bases of DNA, mostly at N7 positions of guanosine and adenosine. Cross-links between two adjacent guanines as well as 1,2-d(GpG) interstrand cross-links represent major lesions observed in the DNA after Cp treatment. Besides this genotoxicity, Cp results in a significant damage of cytoplasm targets, prominently proteins. About half of the patients respond well to platinum-based therapy, however intrinsic or acquired resistance to Cp is a major obstacle [4]. The mechanisms of resistance have been classified into several functional and hierarchical categories: (i) pre-target, (ii) on-target, (iii) post-target, and (iv) off-target. Pre-target resistance covers the modification of target binding in the cytoplasma or to DNA, e.g. via decreased influx and/or increased efflux of the drug, whereas on-target resistance reduces the molecular damage, e.g. by aggravated DNA repair. Post-target resistance includes an adaptive response of common signaling pathways, e.g. for a reduced rate of regulated cell death, whereas modifications not directly associated to the chemotherapeutic agent are summarized as off-targets [4]. Most resistance described in the literature falls into the category of pre-targets resulting in reduced platinum accumulation in tumor cells [1].

The cellular uptake of anticancer drugs, their tissue distribution, toxicity, and the efficacy to erase tumor cells largely depend on the expression and activity of transporters [5, 6]. This group of carriers include the multidrug and toxin extrusion transporters (MATEs; SLC47A), organic cation/carnitine transporters (OCTNs; SLC22A4 and 5), and organic cation transporters (OCTs; SLC22A1-3). For OCTs, three different forms, OCT1, OCT2 and OCT3, have been characterized [6]. Most of the transporters display a high tissue and subcellular-specific expression translating into tumor-specific efficacy and cell type-specific side effects. Cisplatin is an excellent OCT2 substrate; however, it is a poor substrate of either MATE1 or MATE2-K [7, 8]. Of note, the processing pathway of platinum drugs is shared by the intracellular routes of metals, in particular copper, as exemplified by the family of copper transporters (CTRs; SLC31A). CTR1, the principal copper importer of mammalian cells, was reported to mediate the majority of platinum intake [9, 10]. Metallothionein 1 (MT1) located in the cytoplasm was proposed to bind cp resulting in a non-toxic storage [11]. In addition, increased expression of copper transporting P-type adenosine triphosphatase ATP7A and ATP7B have been reported to be involved in platinum drug transport [12, 13]. In line, overexpression of ATP7B has been suggested to be of importance for establishment of Cp resistance in various tumors derived e.g. from ovary, oral squamous cell carcinoma and prostate [14-17]. Given the significant overlap between platinum and copper pathways, combination therapy involving platinum and copper chelator has been recently proposed [18].

In terms of carrier expression that mediate intracellular Cp transport, hepatocytes are of high interest, since transport of copper is one of the primary function of this well characterized cell type [19]. In this report, we took advantage of a human hepatoma cell line that carries a knockout (KO) of ATP7B [20]. We re-addressed the role of ATP7B in hepatocytes [21] and characterized transporter expression after establishment of Cp resistant hepatoma cells. Our findings significantly increase the knowledge on the molecular role of individual transporters for the efficacy of platinum therapy, particularly ATP7B, MT1 and OCT3, and shed new light on the copper cross-activity that may have significant clinical impact for the development of improved platinum-based therapy.

RESULTS

ATP7B does not affect cisplatin sensitivity in hepatoma cells

Copper transporter ATP7B was reported to be involved in the establishment of cisplatin (Cp) resistance in various cells [14-16]. In order to re-evaluate the impact of ATP7B for establishment of Cp resistance, a recently generated human hepatoma HepG2 cell line lacking functional ATP7B (KO) was compared to parental cell line expressing high levels of ATP7B. The ATP7B KO cell line was shown to lack expression of ATP7B protein following zinc-finger induced mutagenesis of ATP7B [20]. No significant differences with regard to cell survival following Cp exposure were observed between the two cell lines (Figure 1A). Variation of the time period of Cp exposure (5 minutes to 72 h) or Cp concentration (up to 200 µM) did not result in a different Cp sensitivity (data not shown). In order to assess any differences in the accumulation of the drug, intracellular Cp concentrations were determined in parental and ATP7B KO cells (Figure 1B). The soluble cellular fraction of both cell lines displayed almost identical levels of Cp suggesting that Cp uptake/storage was not altered by the KO of ATP7B. As ATP7B overexpression was implicated to confer resistance [14], the question was addressed whether retroviral vectors overexpressing ATP7B can confer improved Cp resistance in hepatoma cell lines. However, overexpression of ATP7B in HepG2 and Huh-7 cells did not result in an increased Cp resistance (Supplementary Figure 1). In contrast, both transduced cell lines displayed an increased resistance to copper suggesting that overexpression gives rise to functional ATP7B.

Figure 1

ATP7B expression does not affect cisplatin sensitivity in hepatoma cells

(A) Cell viability was determined by MTT assay relative to untreated cells (100%). Mean/SE are given (n = 5). (B) Intracellular cisplatin level was determined by TXRF in the soluble cellular fractions of the cells. Cells were incubated with cisplatin for 4 h. Mean/SE are given (n = 3).

Hepatoma cells lacking ATP7B can achieve cisplatin resistance

Having shown that ATP7B expression does not modulate Cp sensitivity and accumulation in hepatoma cells, the question was addressed which other genes may result in an adaptation to toxic Cp concentrations. First, the survival of ATP7B KO cells was determined following long-term Cp exposure. Exposure to 1.0 µM and 5.0 µM Cp resulted in cell death after 7–21 days, while 0.1 µM Cp did not disturb cell proliferation for more than 23 days (Supplementary Table 1). To adapt the cells to toxic Cp concentrations, the cisplatin concentration was stepwise increased by 0.1 µM at a weekly basis. Using this protocol over a time period of several months, a Cp resistant cell line (CpR) was established that showed cell proliferation despite being continuously grown in high Cp concentrations. Cp concentrations of up to 4 µM were well tolerated. CpR cells could be grown in the presence of high Cp for more than a year without evident changes in cell morphology (Figure 2A). The morphology of CpR cells was similar to parental cell line ATP7B KO and HepG2 cells (Supplementary Figure 2). The cumulative growth of CpR cells indicated similar proliferation rates as compared to untreated ATP7B KO cells (Figure 2B). Annexin V staining was used to characterize the induction of apoptosis in CpR cells. Experiments were carried out using 10 µM Cp for 72 h, since extensive necrosis was observed at higher Cp concentrations (data not shown). Induction of apoptosis was significantly reduced in the CpR cells as compared to ATP7B KO cells (Figure 2C). We next assessed the intracellular Cp concentration in the nuclear and soluble fractions of CpR cells (Figure 2D). While the nuclear fractions showed no differences of Cp accumulation, a significantly decreased level was observed in the soluble fraction of CpR cells as compared to ATP7B KO cells, corroborating that Cp is targeting cytoplasmic compartments [4, 22].

Figure 2

Human hepatoma cells lacking ATP7B can adapt to proliferation in toxic cisplatin

(A) Light microscopic image of CpR cells after continuous growth in toxic cisplatin concentration for more than a year. Cell morphology was identical to parental cells. Scale bar, 50 µm. (B) Cumulative growth of CpR cells in the presence (+Cp) and absence (-Cp) of cisplatin. Growth of parental ATP7B KO cells is depicted for comparison. Mean/SE are given (n = 3). (C) Determination of apoptosis by Annexin V staining in CpR cells and ATP7B KO cells. Mean/SE are given (n = 3). (D) Intracellular cisplatin concentration was determined in nuclear and soluble fractions of CpR and ATP7B KO cells. Mean/SE are given (n = 3). *P < 0.05.

Organic cation transporter 3 is stably downregulated in cisplatin resistant hepatoma cells lacking ATP7B

In order to address the molecular determinants governing the reduced Cp accumulation in CpR cells, a RT-qPCR analysis was performed. A set of 16 genes was chosen from the literature that covers some of the major transporters involved in transport of metals and/or copper homeostasis. The selected genes were implicated to be involved in metal transport (DMT1, MATE1 and 2, MDR1, MRP1 and 2, OCT1-3, OCTN1 and 2) and/or copper homeostasis (ATP7A, CP, CTR1 and 2, GST, MT1) (Supplementary Table 2). RT-qPCR analysis showed that the expression level of most genes was within a threshold of ±2 fold that was chosen to indicate similar expression as compared to ATP7B KO cells (Figure 3A). However, the expression of organic cation transporter 3 (OCT3) was significantly downregulated (5.17 ± 2 fold), while metallothionein 1 (MT1) was significantly upregulated (8.91 ± 4 fold). For further validation, we assessed basal expression of OCT3 and MT1 in hepatoma cell line Huh-7 and primary human hepatocytes. Basal OCT3 expression was almost identical in the samples, while MT1 expression was highly elevated in human hepatocytes relative to the two hepatoma cell lines (not shown). Next, we determined whether the gene expression of OCT3 and MT1 in CpR cells is permanently affected even after drug removal. CpR cells were regrown in basal media for at least one month (weaning) giving rise to CpRw cells. Analysis of the cell viability in CpRw cells revealed an almost identical high level of resistance as compared to CpR cells suggesting that the acquired resistance is a stable modification (Figure 3B). Gene expression analysis showed that MT1 expression was normalized in CpRw cells to a level observed in the non-resistant ATP7B KO cells (Figure 3C). However, OCT3 gene expression remained downregulated in CpRw cells suggesting that expression is stably compromised.

Figure 3

Cisplatin resistance is a stable modification of hepatoma cells

(A) mRNA expression of CpR cells relative to ATP7B KO cells was calculated using the 2–ΔΔ method. Dotted line at a fold change of ±2 indicates significance. Mean/SE are given (n = 3). (B) Cell viability as determined by MTT assay relative to untreated cells (100%). Mean/SE are given (n = 5). *P < 0.05. (C) mRNA expression of CpRw cells is presented as fold change relative to ATP7B KO cells. Mean/SE are given (n = 3).

To further assess the adaptive role of MT1 and OCT3 gene expression for the establishment of Cp resistance, ATP7B KO cells were exposed to Cp for a short time period (72 h) and subjected to RT-qPCR analysis. MT1 was significantly upregulated when toxic Cp concentrations of 1.0 µM and 5.0 µM were used indicating that induction of MT1 represents a transient adaptive response to toxic Cp levels (Supplementary Figure 3A). In contrast, expression of OCT3 was not affected corroborating that OCT3 downregulation has been acquired during long-term Cp exposure of CpR cells. Of note, knockdown of MT1 by siRNA revealed no significant difference of Cp resistance suggesting that MT1 has a minor role for the establishment of Cp resistance (Supplementary Figure 3B and 3C). Also, downregulation of copper and cisplatin transporter CTR1 did not affect Cp resistance (Supplementary Figure 4).

Overexpression of OCT3 can reverse cisplatin resistance

In order to further define the role of OCT3 for establishment of Cp resistance, the question was addressed whether overexpression of OCT3 in CpR cells can revert resistance. CpR cells were stably transfected with a vector expressing OCT3 resulting in CpROCT3 cells. OCT3 mRNA was increased in CpROCT3 cells as compared to CpR cells by a factor of ≈20 fold (Figure 4A). Using Western blot analysis, OCT3 protein expression was found to be increased to a lesser extent (≈2–3 fold) in the stably transfected cells (Figure 4B; Supplementary Figure 5). To assess the function of OCT3, transport of 4-(4-Dimethylamino)styryl)-N-Methylpyridinium Iodide (ASP+), a well-known substrate of OCT3 [23], was determined. OCT3 transport was increased in stably transfected CpROCT3 cells (Figure 4C). Importantly, overexpression of OCT3 resulted in almost same levels of Cp sensitivity as observed in ATP7B KO cells (Figure 4D) suggesting that OCT3 is sufficient to revert Cp resistance of CpR cells.

Figure 4

OCT3 mediates cisplatin resistance in hepatoma cells

(A) mRNA expression in CpROCT3 cells and CpR cells. Mean/SE are given (n = 3). (B) Densitometric analysis of OCT3 protein expression as determined by Western blot. ATP7B KO cells were set to 100%. Mean/SE are given (n = 3). (C) OCT3 function was assessed by ASP+ uptake rates. Mean/SE are given (n = 3). (D) Cell viability was determined by MTT assay relative to untreated cells (100%). Mean/SE are given (n = 5). *P < 0.05.

OCT3 confers copper cross-resistance in hepatoma cells

Cisplatin resistant cells were shown to also resist toxic copper by a phenomenon described as cross-resistance [24]. We addressed the question of cross-resistance in CpR cells. Following copper exposure, CpR cells displayed a significant higher cell viability as compared to ATP7B KO cells (Figure 5A). OCT3 overexpression increased copper induced toxicity to the level observed in ATP7B KO cells. In contrast, OCT3 downregulation in ATP7B KO cells via siRNA (Supplementary Figure 6) was followed by an increased copper resistance relative to vehicle treated cells (Figure 5B). To more directly assess the role of OCT3 for the transport of copper, the intracellular copper concentration was determined. A significantly reduced intracellular copper concentration was found after OCT3 siRNA knockdown in ATP7B KO cells (Figure 5C) suggesting that, in addition to mediate the uptake of Cp, OCT3 may also support the uptake of copper.

Figure 5

OCT3 confers cross-resistance

(A) Cell viability was determined by MTT assay relative to untreated cells (100%). Cells were exposed to 0.75 mM copper for 48 h. HepG2 cells expressing ATP7B are shown as control. Mean/SE are given (n = 3). (B) ATP7B KO cells were treated with siRNA directed against OCT3 and cell viability was assessed by MTT assay. Mean/SE are given (n = 3). (C) Intracellular copper concentration as determined by AAS. ATP7B KO cells served for comparison (100%). Mean/SE are given (n = 3). *P < 0.05.

DISCUSSION

Growing evidence suggests that the efficacy of platinum-based chemotherapy is closely related to the activity of transporters that mediate the intracellular processing of the drugs both at the level of the targeted tumor cell for efficient induction of cytotoxicity as well as in non-tumor tissues for the reduction of unwanted side effects. As primary hepatocytes do not show long-lasting proliferation in tissue culture, we used most frequently employed human hepatoma cell lines. Our results shed new light on several major transporters and intracellular carriers that have been implicated to be key determinants of the efficacy of Cp treatment. (i) ATP7B does not have an impact on Cp-induced toxicity, at least for hepatoma cells which represent the primary physiological expression site of this transporter. Of note, the cells used in this study allowed an analysis of Cp sensitivity in the absence of the second human P-type ATPAse as hepatocytes do not express ATP7A [12]. (ii) MT1, ubiquitously expressed in mammalian cells, was shown here to be highly induced by cisplatin, however induction can be transient and normalizes after withdrawal of the drug. (iii) Expression level of OCT3 seems to be directly linked to cellular Cp accumulation; conversely, its downregulation results in an increased Cp resistance. Interestingly, OCT3 expression is relatively low in human hepatocytes and hepatoma cell lines as compared to OCT1, while OCT2, the previously reported cisplatin transporter of kidney, is missing in these cells [25]. OCT3 rather than OCT1 may therefore represent the major target for establishment of Cp resistance when OCT2 expression is missing. (iv) OCT3 also contributed to intracellular copper accumulation further corroborating the clinical impact of the two intermingled transport networks for efficient platinum-based therapy.

ATP7B expression is mostly limited to hepatocytes [19]. However, induction of ATP7B has been suggested to be associated to resistance [12, 13]. In order to explore an establishment of Cp resistance in the absence of ATP7B, we used a previously generated hepatoma KO cell line [20]. Expression of ATP7B in primary rat hepatocytes was associated with increased Cp sensitivity as compared to Atp7b KO cells [21]. The previous finding indicates that ATP7B may have a different role in hepatocytes as compared to other cells, including ovarian carcinoma cells and primary MNK fibroblasts, where overexpression of the transporter was shown to rescue from Cp toxicity [14, 16, 26, 27]. However, primary hepatocytes rapidly undergo apoptosis and expression of ATP7B in non-hepatic cells may not resemble the physiological function of the transporter. Our results that were derived in isogenic sublines of a human hepatoma cell line, either after native expression, gene knockout or overexpression, suggest that ATP7B does not affect Cp sensitivity. Our measurements presented almost identical intracellular levels of Cp regardless whether ATP7B was expressed or not, suggesting that the transporter does not actively expel significant amounts of Cp from the cell endorsing the previous proposal of an ATP7B-linked detoxification mechanism other than export [21]. However, this conclusion is challenged by other reports where ATP7B expression in non-hepatic cells resulted in decreased intracellular Cp levels [14]. In line, confocal microscopic analysis of ovarian cell lines expressing ATP7B suggested Cp transport into vesicles [15]. Our observations therefore do not exclude that elevated expression of ATP7B in cells other than hepatocytes may have clinical significance for the course of platinum drug therapy [28-30]. However, such mechanisms may involve molecular events unrelated to export.

MT1 has been proposed to be capable of inactivating cisplatin [11]. MT1 but not GST, a second major thiol-rich protein of cells, was identified here to be highly induced by Cp. Importantly, MT1 upregulation by Cp is transient and its expression seems to be related to the immediate presence of the drug. Our results suggest that MT1 induction may be a primary defense mechanism of the cell to cope with exposition to a metal-based drug.

Molecular characterization of our cells indicates that OCT3 represents the dominant transporter for the establishment of Cp resistance. Whereas it has been previously demonstrated that expression of OCT1 and OCT2 was followed by intracellular accumulation of platinum drugs [31-33], analysis of OCT3 gave ambiguous results [34-36]. Studies involving resistant cell lines derived from cervical cancer and oral squamous cell carcinoma slowed downregulation of OCT3, corroborating our results that this transporter is a major contributor for the generation of resistance after platinum drug treatment [37, 38]. Notably, OCT3 overexpression could almost fully reverse resistance in our cells suggesting that downregulation of OCT3 is a single molecular event for establishment of resistant cells. Other genes, e.g. DMT1 and CTR1, were only moderately (≤2 fold) modulated, however such gene expression might contribute to resistance. For CTR1, the well-known importer of Cp [9, 10], our data using knockdown experiments indicate that this transporter does not affect resistance. Our results do not provide strict evidence that Cp is directly transported by OCT3 across the cell membrane, since other substrates unrelated to the drug or secondary interacting molecules may critically influence the uptake [39]. In addition, we did not observe improved cell viability after OCT3 knockdown in parental cells (data not shown) corroborating similar findings of others [37, 40]. We therefore hypothesize that the role of downregulated OCT3 with regard to Cp resistance may either be indirectly or modulated by other transporters, e.g. CTR1.

Our study is limited since hepatoma cell lines like HepG2 differ from primary human hepatocytes, although we showed that the basal expression of OCT3 mRNA is similar. OCT3 may also have other roles in different cell types [37]. It should be noted that oxaliplatin, which was not assessed in this study, can also be used for treatment of HCC, however as shown by others, subtle differences between the uptake of both platinum compounds may exist [8, 35]. From the clinical point of view, it is interesting that OCT3 is widely expressed in tissues, including placenta, intestine, heart, lung, brain, kidney and liver, whereas OCT1 and OCT2 show a very limited tissue distribution and are most prominently expressed in the liver and the kidney, respectively [6]. Following oxaliplatin treatment, the mRNA level of OCT3 was found to be downregulated in several cancerous tissues, including uterus, ovary, and lung where platinum treatment is most effective, suggesting that OCT3 downregulation might be operative in patient tumor cells [34]. OCT1 and to a lesser extent OCT3 were found to be downregulated in HCC [41]. Given the broad OCT3 expression amongst tissues and tumors as well as the finding on OCT3-mediated Cp resistance presented here and elsewhere [37], further analysis of resistant tumor cells could be of significant value to assess OCT3 as a novel marker for prognosis of platinum drug therapy.

One major finding of the study is the observed Cp and copper cross-resistance both linked to OCT3 downregulation. A copper resistant human hepatoma Huh-7 cell subline was shown before to have increased cell survival after Cp exposure [42]. Whereas some of the main copper transporters, including CTR1, ATP7A and ATP7B, have been described to mediate Cp and copper cross-resistance [24, 26, 43], our data add to the candidates mediating cross-resistance. The implication of OCT3 mediating copper cross-resistance is twofold. Firstly, downregulation of OCT3 in parental cells resulted in increased copper resistance, whereas OCT3 overexpression in Cp resistant cells was followed by a higher sensitivity towards copper. Secondly, analysis of the cellular copper accumulation suggested a significantly reduced copper uptake after knockdown of OCT3. To our knowledge this is the first report on copper being a substrate of OCTs. Typically, organic cations with one or two positive charges represent a substrate of OCTs [6] implying that the copper cation fits into this category. OCT3 may therefore contribute to the activity of the universally expressed CTR1 transporter that accounts for a significant import of copper [44]. Similarity, the impact of CTR1 for the import of platinum was recently challenged suggesting that other transporters, likely OCT3, might add to the uptake of the drug [9, 10, 45]. Since the transport of platinum drugs is partly overlapping with copper homeostasis, an improved understanding of the molecular mechanisms seems mandatory for the identification of networks that could improve the bioavailability of the drug. Of note, established clinical procedures of copper chelation have been recently proposed to overcome Cp resistance via enhanced CTR1 expression [46]. Induction of OCT3 expression, possibly via concomitant copper chelation therapy might therefore also represent an alternative approach to improve current therapy involving Cp for treatment of HCC [47, 48]. Thus, our data on OCT3-mediated cross-resistance further substantiate that more approaches to improve the efficacy of platinum drug therapy by modulation of copper homeostasis could result in a clinical benefit. In addition, it can be speculated that assessments of OCT3 expression in patients may be of value to monitor the efficacy of anticancer therapy and to advance current prognosis [40].

MATERIALS AND METHODS

Cell lines

The cell lines HepG2 and Huh-7 were obtained from American Type Culture Collection (ATCC). The ATP7B knockout derivate of HepG2 (KO) were obtained as described [20]. HepG2 cells were cultured in RPMI (Lonza). Huh-7 cells were cultured in DMEM High Glucose (GE Healthcare). All media contained 10% fetal bovine serum (FBS) supplemented with 100 U/mL penicillin/streptomycin (Hyclone, Logan). Cell lines were maintained in 5% CO2 at 37°C in a humidified chamber.

Growth in cisplatin containing medium

Cisplatin solutions (TEVA Pharma) were prepared from a 3 mM stock in standard cell culture medium. Cells were cultured in the presence of 0.1 µM Cp for one week. The cell growth was monitored by trypan blue staining (Sigma-Aldrich). Passaging of cells was performed twice a week at 70–90% confluency. Images were obtained with an Olympus CKX41-X10 microscope using cellSens Standard 1.11 imaging software.

Transfection

Stable ATP7B overexpressing cell lines were generated using retroviral vectors harboring wild type ATP7B. Wildtype ATP7B cDNA was cloned into pGCsamEN retroviral vector encoding blasticidin resistance (kind gift of O. Wildner, Ruhr Universität Bochum). Cells expressing empty vector pGCsamEN were used as controls. Cells were selected in media containing 6 µg/mL blasticidin (Invitrogen). Stable OCT3 overexpressing cell line was generated using pcDNA3.1_hOCT3 (kind gift of Prof. Koepsell, University Würzburg, Germany). Cells expressing OCT3 were selected with 1.5 µg/ml Geneticin (Gibco). Knockdown experiments were performed using 25 pmol siRNA directed against OCT3 (Ambion; #109426), MT1 (Ambion; #215639), and antisense oligonucleotide (ASO) directed against CTR1 (kindly provided by Ionis Pharmaceuticals, Inc.). Reverse transfection of siRNA was carried out following the protocol of the manufacturer using 4 µl Lipofectamine RNAiMax (Invitrogen). All transfection experiments were confirmed via RT-qPCR after 24 h post transfection.

Cumulative cell growth

5 × 105 cells were seeded in a 6 well plate (Cellstar) and cultured in cell culture medium with and without Cp addition. The cell number was determined every three days using trypan blue and 5×105 cells were reseeded. The cumulative cell number was calculated. Each cell count was independently repeated at least three times.

Cytotoxicity assay

104 cells were seeded in triplicates in a 96 well plate (Falcon) and cultivated in 100 µl RPMI media lacking phenol red (Lonza). The next day, different Cp or copper concentrations were added. For Cp and copper, incubation times of 72 h and 48 h were used, respectively. Cell viability was determined by addition of MTT (3-[4, 5-dimethylthiazolyl-2]-2, 5-diphenyltetrazolium bromide; Sigma) that was added at 1mg/ml for 2 h. Absorbance at 560 nm was measured after addition of sodium dodecyl sulfate (SDS; Roth) and dimethyl sulfoxide (DMSO; Roth). Viability was determined as a percentage of untreated control cells (100%).

Apoptosis

106 cells were incubated with 10 µM Cp at 37°C for 72 h. Supernatants and cells were collected and subjected to Annexin-V and propidium iodide staining (Roche Annexin-V-FLUOS kit) and analyzed using flow cytometry (Epics XL.MCL, Beckman Coulter).

Cisplatin accumulation

106 cells were incubated in standard cell culture medium with Cp at 37°C for 4 h. After incubation, cells were washed with ice cold PBS and resuspended in Lysis Buffer (15 mM NaCl, 0.5% Triton X-100). Cells were centrifuged at 14,000 rpm for 5 min at 4°C. Supernatant (soluble fraction) and the cell pellet (nuclear fraction) were collected. The platinum concentration was determined by means of total reflection X-ray analysis (TXRF, S2 Picofox, Bruker Nano). Quantification was performed by the Bruker Spectra software (version 6.1.5.0) and based on the known concentration of the internal arsenic standard solution (Fluka Chemie GmbH).

Copper determination

5 × 105 cells were seeded in 12 wells and incubated with 0.01 mM copper prepared from a 10 mM CuCl2 stock solution (Sigma) in standard cell culture medium for 4 h. After two washings using PBS, cells were lysed to determine total protein by Bradford assay (BIO RAD). Samples were dried at 65°C for 24 h dissolved in 300 µl 65% nitric acid (Suprapure; Merck, Germany). Analysis was performed by atomic absorption spectroscopy (Shimadzu AA-6300).

Real-time quantitative PCR

PCR analysis was performed as described previously [49]. RNA derived from primary human hepatocytes after whole liver perfusion and cryopreservation was used as control. Total RNA was isolated by RNeasy kit (Qiagen). 1 µg of RNA was transcribed using SuperScript II (Invitrogen) according to the instructions of the manufacturer. For quantitative real time PCR (qPCR) the RT product was incubated with SYBR Green PCR Core Plus (Eurogentec, Belgium) and 150 nM of primers. PCR was analyzed on the ABI Prism 7900 HT Sequence Detection System (PE Applied Biosystems). Each sample was tested in three independent experiments using all primers in the same experiment. Ct values were normalized to the expression of the house-keeping gene (ΔΔct method) and fold change was calculated using the 2ΔΔCt-method.

OCT3 protein determination

For lysis of cells, RIPA buffer (60 mM tris-HCl, 150 mM NaCl, 2% Na-deoxycholate, 2% Triton X-10, 0.2% SDS, and 15 mM EDTA) and protease inhibitors (Roche, Basel, Switzerland; Complete Mini, EDTA-free) were used. 50 µg protein lysate was loaded on a 12.5% SDS gel. Polyclonal anti-rabbit OCT3 antibody (1:1,000; Abcam; ab124826) was added overnight. β-Actin was assessed for protein loading control (1:1,000; sc-47778 HRP, Santa Cruz Biotechnology, Santa Cruz, CA, USA).

OCT3 functional assay

106 cells were seeded in 96 wells and grown to 80–90% confluency. Cell culture medium was removed and HCO3-free Ringer-like solution (pH 7.4) was added. Cellular uptake was performed as described before [50] using 5 µM 4-(4-Dimethylamino)styryl)-N-Methylpyridinium Iodide (ASP+). A fluorescence plate reader (Infinity M200; Tecan, Crailsheim, Germany) with a monochromator system (excitation at 450 nm and emission at 590 nm) was used for analysis.

Statistical analysis

Statistical analysis was performed by Kruskal-Wallis 1-way ANOVA and Wilcoxon Mann-Whitney-test using SPSS 22.0 software. Data are given as mean ± standard error of the mean (SE). A p < 0.05 value was used to indicate significance.