Human Drug Eux Transporter ABCC5 Confers Sensitivity and Acquired Resistance to Pemetrexed in Breast Cancer

Aim Pemetrexed, a new generation antifolate drug, is approved for the treatment for locally advanced or metastatic breast cancer, but factors affecting the ecacy and resistance of it have yet to be fully explicit. ATP-binding cassette transporters have been reported as prognostic and adverse effects predictors of many xenobiotics. This study was designed to explore whether ABC transporters affect pemetrexed resistance and may contribute to treatment regimen optimization for breast cancer. Methods Firstly, the expression of ABC transporters family members was measured in cell lines, thereafter examined the potential role of ABC transporter in conferring resistance to pemetrexed in primary cancer cell lines isolated from 34 breast cancer patients, and then the role of ABCC5 in mediating transport of pemetrexed and apoptosis pathway in MCF-7 cell lines was assessed. Finally, the functions of ABCC5 on therapeutic effect of pemetrexed was evaluated in breast cancer bearing mice.


Introduction
Pemetrexed (MTA) is a novel multi-targeted antifolate for the treatment of non-small cell lung cancer and mesothelioma [1][2][3][4] through inhibiting thymidylate synthase, dihydrofolate reductase, and glycinamide ribonucleotide formyltransferase, which are folate-dependent enzymes involved in the de novo biosynthesis of thymidine and purine nucleotides [5,6]. Shaughnessy et al. proved that MTA had a good effect (overall response rate:8%, stable disease:36%, median survival:8 months) on metastatic breast cancer (BC) and well-tolerated in 80 patients in second-line treatment [7], and another study presented an approximately 30% response rate in advanced BC patients in rst-line, and 21% in second-line treatment, too [8].
The mechanism of sensitivity and/or acquired resistance of MTA is complex, mainly including decreased intracellular concentration and alteration of metabolism, etc [9].The cytotoxic activity of MTA is largely attributed to its concentration and retention time in the cells [10]. Thus, transmembrane transport of MTA is a critical determinant for its activity. In this context, the decreased expression of in ux transporters (i.e. folate receptor) and increased expression of e ux transporters could induce the cancer cells more resistant through reducing the intracellular drug exposure [11,12].
The ATP binding cassette transporter (ABC transporter), also named as multidrug resistance protein (MRP) [13], is capable of conferring resistance to nucleotide analogues such as 5-FU [14,15], 6thioguanine (6-TG) [16], 6-mercaptopurine (6-MP) [17], and 9-(2-phosphonylmethoxynyl)-adenine (PMEA) [18,19]. The expression of ABC transporters is regulated by PXR (pregnane X receptor), CAR (constitutive androstane receptor) and other transcriptional regulators, and the primary functions of them lie in e ux many endogenous and exogenous substrates [20]. ABC transporters comprise approximately of 50 members and are subdivided into seven groups (ABCA to ABCG) [21]. In particularly, the C branch of the ABC transporters (ABCC) superfamily dominated in multidrug resistance [22,23].The rst evidence about the hydrophilic antifolate transport by certain members of the ABCC family was put forward in 1997 by Masuda and his colleagues [24], who demonstrated that MTA was transported into bile only in the wild type rats but not in rats that has hereditary de ciency in ABCC2 function. After this study, many researchers have found more evidences of ABCC affecting the antifolate transport. Vlaming, et al. found that the absence of ABCC2 and/or ABCG2 in mice increased the oral availability of MTA [25,26]. Wielinga, et al. identi ed an important exporter, ABCC5, which is involved in the transport of MTA and folic acid in HEK 293 cells [28].
Previous studies have proven that the expression of ABC transporter in human cancer could regulate the e cacy of chemotherapy, but the contribution of ABC transporters to drug-resistance, especially to novel antifolate drug has not yet been fully elucidated. This study was designed to explore whether ABC transporters in uence the MTA sensitivity and resistance in BC.

Materials And Methods
Cell lines MCF-7 and MCF-7-ADR (MCF-7-adriamycin-resistant breast cancer cells) cell lines were obtained from the Culture Collection Company (ATCC-LGC Promochem, Teddington, UK). Cells were routinely grown in DMEM (Invitrogen) medium supplemented with 10% FBS (Invitrogen) and 100 units of penicillin/streptomycin per mL (Invitrogen), at 37 ℃ and 5% CO 2 under humidifying conditions. The medium was refreshed every two orthree days regularly until the cell reaches 80-90% con uence, and the cells were transferred to next experiment or make stock solutions. The mycoplasma contamination was tested every month.
Reagents MTA (reference substance solution), which was used in quantitative analysis in a RP-HPLC experiment, was purchased from Sigma-Aldrich (St. Louis, MO). MTA used for cell incubation, was purchased from Eli Lilly company. ABCC5 Human shRNA was bought from OriGENE company (Locus ID 10057, Product ID TL315024).

Generation of ABCC5 and GFP adenovirus
For recombinant adenovirus construction, the ABCC5-gene cDNA (Ad-ABCC5) and the green uorescence protein gene cDNA (Ad-GFP; control) were cloned by PCR and inserted into pHBAD-EF1-MCS-3 ag-CMV-EGFP vector (supplemental Fig.1). The pDC315-ABCC5 and pBHGloxE1,3Cre were co-transfected into HEK293 cells using Lipo-FiterTM transfection reagent (QIAGEN) to generate the recombinant adenoviruses. The recombinant Ad-ABCC5 and Ad-GFP adenoviruses produced in the HEK293 cells were puri ed and the virus titer was measured by plaque assay. The stock solutions of Ad-ABCC5 and Ad-GFP were 1×10 11 plaque formation unit (PFU)/ml, respectively.
Collection of tumor tissue specimens from patients and primary breast cancer cell isolation A total of 34 patients with con rmed primary BC (2 cm or larger), who consecutively underwent neoadjuvant chemotherapy containing anthracyclines at the Breast Tumor Department, Xinhua Hospital, Shanghai Jiaotong University School of Medicine, were enrolled into this study from January 2014 to December 2015. Tumor specimens were obtained by surgical excision before MTA chemotherapy. Informed consents were signed and retrieved from all patients, following a protocol approved by the Ethics Committee of Xinhua Hospital, Shanghai Jiaotong University School of Medicine.
To isolate the BC cells, at least two consecutive frozen sections were prepared for each para n embedded tumor tissue sample, and one of the section was then subjected to the hematoxylin-eosin staining to con rm the presence of cancer cell, and the adjacent one was transferred to cancer cell isolation as described in the previous publication [29]. Brie y, the blood, fat and bro connective tissue were removed from the tumor tissue pieces, and then the residual was cut into 1-2 mm pieces for enzymatic disaggregation. The small tissue pieces were incubated with 2.5% crude trypsin for 30 minutes at 37°C and with collagenase (0.15%) overnight. Cells released after enzymatic treatments were further tested for cell viability and were cultured to perform subsequent experiments.

Animal studies
Twenty-four female Balc/b nude mice (5 weeks, 18 g) were purchased from the Shanghai Super B&K Laboratory Animal Corp. Ltd. (Shanghai, China), and all the mice were raised in speci c pathogen-free environment with free access to food and water. All animal studies were approved by the Research Ethics Committee at Xinhua hospital, a liated to Shanghai Jiaotong University, school of Medicine.
On day 0, every mouse was injected with 1×10 7 MCF-7 cells subcutaneously into the right armpit. When tumors were approximately 100 mm 3 in size on day 30, adenoviruses containing ABCC5 (5×10 11 PFU) were injected into the tumors of 12 mice to over-express ABCC5 in the tumor cells, and the vehicle was applied to the other 12 control mice. The expression of ABCC5 was checked by diffused green uorescence. When the volume of tumors were approximately 150 mm 3 on day 35, 6 of the ABCC5 overexpression mice and 6 mice in the control group were treated with intravenous injections (via the tail vein) of MTA (20 mg/kg, saline) once a day, and the same dose of vehicle was administrated to the others from day 35 to 46. The tumor volumes (V) were measured using a caliper once a day (V = width 2 × length/2). Mice were killed at the end of 7-week, and tumor volume and weight were measured. The animal experiments' design has been shown in supplement Fig.2.
Measuring the MTA in MCF-7 cell using RP-HPLC MTA concentrations in the MCF-7 cell were determinated based on a developed HPLC method. The chromatographic separation and quanti cation were performed on an RP-column (ZORBAX Eclipse XDB-C8, 250 mm×4.6 mm, 5 μm; Agilent) with the column temperature maintained at 25 °C, and the MTA was detected with a DAD detector at a wavelength of 240 nm. The mobile phase was composed of water plus 0.02 M phosphate buffer (, pH 4.0)/acetonitrile (86:14, V:V) and delivered at a ow rate of 1 mL/min. The sample pretreatment was completed using a ultra ltration method (0.22 μm). All experiments were completed on an Agilent 1260 HPLC system. A calibration curve was constructed at a range of 80-625 ng/mL for the MTA measurement. The injection volume was 20 μL, and all analysis was performed in triplicate.
Cell preparation for cellular uptake analysis MCF-7 cells were seeded at 2×10 5 /well into six-well at-bottom tissue-culture plates in triplicate. After 24 hours, the cells were infected with ABCC5, GFP and sh-ABCC5-RNA expressing adenovirus, separately. The cells were incubated with adenoviral particles for another 24 hours and then refreshed with medium containing 50 μM MTA. At the particular time points (0, 0.5, 1, 2, 4, 24 hours), the cells were washed three times with cold PBS (0.1 M, pH 7.4) and then resuspended in 0.2 mL of RPMI-1640 medium and homogenized. After centrifugation at 13000 ×g for 10 min, the supernatant was harvested and stored at -80°C for the detection of MTA by the RP-HPLC method as mentioned above.
Cell viability assay Cell viability assay was performed using CCK-8 kit (Dojindo Laboratories) according to the manufacturer's instructions. Brie y, cells were plated in 96-well plates at a density of 2000 cells/well, and then treated with different concentrations of MTA for 72 h. Then CCK-8 solution diluted with DMEM/F12 with 10% FBS at 1/10 ratio was added to each well and incubated for 2 hours at 37°C. Finally, the absorbance at 450 nm was measured using a microplate reader. The (%) cell viability was calculated using the formulae; (OD treatment group-ODBlank)/(OD control group-OD Blank) × 100. The IC 50 value was determined using GraphPad Prism software. All experiments were performed in triplicate, and the presented data represent the mean of three biological repeats.

Western blot
Protein extracts were separated through 5% to 12% SDSPAGE, and then transferred to nitrocellulose membranes, probing with mouse monoclonal antibodies against ABCC5 (cell signaling) or β-actin (Proteintech), and then followed by incubation with either IRDye 700 or 800 secondary antibodies and visualized using the Odyssey Infrared Imaging System software (Li-Cor, Lincoln, NE).

RNA Isolation and Real-Time RT-PCR
Total RNA was extracted from the cells using RNeasy mini kit (QIAGEN) and qRT-PCR were performed on the cDNAs generated from 250 ng of total RNA by using HotStart-IT® SYBR® Green qPCR Master Mix along with UDG (2X) by a user Friendly TM kit (USB Corporation). ABC transporters subfamily primers were designed and synthesized by Sangon Biotech (Shanghai) Co., Ltd. (supplemental Tab.1).

Immuno uorescence microscopy
Cells were initially seeded onto coverslips, and then harvested and washed three times with PBS. Cells were xed with 4% paraformaldehyde for 30 min at room temperature and blocked with 1% (w/v) bovine serum albumin, 0.1% Triton X-100, and 0.05% Tween-20 overnight at 4°C to avoid nonspeci c staining. Next, the cells were incubated with goat polyclonal antibody, anti-ABCC5 (PA5-18965, Pierce Biotechnology) for overnight. Subsequently, the secondary antibody (TRITC-rabbit-anti-goat, 1:100) was added and the cells were incubated for 1 hour in dark room. DAPI staining was used to visualize the cell nuclei. The images were captured by Leica DMI300B inverted uorescent microscope.

Data analysis
The results were presented as mean ± SD. The graphics and calculations were nished using Microsoft excel software (Microsoft Corp) or Prism 5.0 software (GraphPad Software Inc). The IC 50 values were calculated by nonlinear regression from a sigmoidal dose-response curve (variable slope, bottom value 0) using Prism software. Pearson's correlation test was used to analyze the correlation between the expression of target gene and the IC 50 values. The p value<0.05 was considered statistically signi cant using unpaired t test analysis unless stated otherwise.

Results
The expression of ABC transporters was increased in MCF-7ADR cell lines The expression of ABC subfamily was measured by qPCR in resistance cell lines MCF-7-ADR and control cells, and 18S expression was utilized as an internal standard to normalize all the data. As shown in Figure 1, almost all the ABC transporters were up-regulated in MCF-7-ADR cells compared with control cells. The expressions of ABCC1 (up to 1.50 times), ABCC2 (1.46 times), ABCC4 (4.31 times), and ABCC5 (5.21 times) were signi cantly increased, and among them, the change of ABCC5 was most obvious, which imply an critical role of it in the drug-resistance of MTA ABCC5 expression in BC tissues correlates with MTA induced cell toxicity To investigate the correlation between ABCC expression and cellular sensitivity (IC 50 ) to MTA in patients, we enrolled 34 BC patients and tested the IC 50 of primary BC cells isolated from their tissue samples. In addition, we performed the mRNA expression analysis of 11 ABC transporter family genes in the BC cells from the samples (supplemental Fig.3). A pearson's correlation analysis was performed to evaluate the association between ABCC expression and BC cell viability. The ABCC2 expresses highest in all ABC transporter family but no association was found with MTA sensitivity (R=0.07, P=0.71), and only ABCC5, as shown in Figure 2, presented a signi cant correlation with the IC 50 of MTA(R=0.741, p<0.001).The ABCC5 may dominate the ABC transporter-mediated MTA drug resistance in BC.

Effect of ABCC5 on e ux of MTA
To validate the functional signi cance of ABCC5 in drug e ux, we transducted MCF-7 cells with ad-ABCC5 or Ad-GFP. As shown in Fig. 3A the transduction of MCF-7 cells with ad-ABCC5 resulted in approximately 10-fold increase of ABCC5 protein expression. The immuno uorescence microscopy showed that ABCC5 was expressed on the cell membrane and an obvious increase was presented of ABCC5 levels in the ad-ABCC5 group compared to the control ( Figure 3B). The ABCC5 siRNA was used to knockdown ABCC5 expression, and the result showed ABCC5 mRNA was totally suppressed ( Figure 3C).
Next, we interrogated if alterations in ABCC5 expression in MCF-7 could in uence the e ux of MTA and thereby alter its intracellular concentration. The MCF-7, ABCC5-overexpressing MCF-7 or ABCC5knockdown MCF-7 cells were treated with 50 uM MTA for 0, 0.5, 1, 2, 3, 4 or 24 hours at 37°C, and then the cells were collected and processed to extract the intracellular MTA, and the supernatant was analyzed for the intracellular MTA concentrations. The intracellular MTA concentration was much lower in the ABCC5overexpression cells, and higher in ABCC5-knockdown cells, compared with the control cells (Figure 4). Our results prove a negative correlation between ABCC5 expression and the intracellular concentration of MTA in MCF-7.
Over-expression of ABCC5 weaken the cytotoxicity of MTA The cytotoxic effect of MTA was evaluated in ABCC5-slienced MCF-7, ABCC5-overexpressing MCF-7 and MCF-7 cells using the CCK-8 assay. As shown in Figure 5A, the IC 50 of MTA was signi cantly declined after the expression of ABCC5 was silenced (IC 50 =0.06±0.01 and 0.11±0.06 μg/ml for ABCC5 silenced and normal expression MCF-7 cell lines, respectively, P=0.02), and the IC 50 increased to 0.2±0.05 μg/ml if the expression of ABCC5 was upregulated(P=0.003, compared to MCF-7 group). Compared with control cells, there was an obvious right shift of the dose-response curve in ABCC5-overexpression cells, and an inverse shift in ABCC5-knockdown MCF-7 cells. To further investigate apoptosis induced by MTA, the expressions of the cleaved caspase-3 and bax were measured using western blot. As shown in Fig. 5B, the expressions of the cleaved caspase-3 and bax were down-regulated in ABCC5-overexpression cells, which represent a declined cell apoptosis.
Tumor growth affected by ABCC5 in mice treated by MTA The therapeutic effects of MTA were evaluated in mice bearing MCF-7+GFP or MCF7+ABCC5 tumor xenografts. As shown in Fig. 6A GFP group vs ABCC5 mice treated with MTA: P<0.01). Additionally, the tumor size of ABCC5overexpression group was signi cantly higher than that of GFP groups when treated with MTA (p<0.01). These results showed the over-expression of ABCC5 could reduce the cytotoxic effect of MTA in vivo. (Fig. 6B, Fig 6C).

Discussion
Chemotherapy is the one of the main options for the treatment of nonresectable BC, but the development of resistance to chemotherapy has become a critical problem in the clinical practice. Multidrug resistance can be in uenced by several factors, and the modulation of the expression and function of drug resistance proteins is an important way among that by which drug resistance is achieved [31], and specially, ABC transporters seems to be critical for emergence of multidrug resistance in cancer [32]. In our study, we have identi ed that the expression of all multidrug resistance proteins increased in the MCF-7-ADR cell line, and particularly, ABCC5 gene showed highest increase in its expression. The drug resistance against MTA had also signi cantly increased in the MCF7-ADR cell lines. This was further con rmed by the ABCC5 overexpression in primary BC cell lines, in which the cells showed reduced sensitivity to MTA; decreased accumulation of MTA; enhanced e ux of MTA, and eventually led to the repression of cell apoptosis. Furthermore, the effect of ABCC5 on MTA resistance was a rmed in an animal study.
It has been reported that the expressions of many genes, for instance, P-glycoprotein (P-gp), multidrug resistance related protein (ABC transporter) [32], ABCG2 (breast cancer resistance protein, BCRP) and lung resistance related protein (LRP) were upregulated during multidrug resistance development in cell lines [33], which is largely in consistent with the results in our study.
There are many studies implicating the roles of ABC transporters in the e ux of folate and antifolate drugs [10,34]. ABCC5 has also been shown to be involved in the e ux of many different anticancer drugs, for instance, 6-MP, 6-TG, 5-FU and its metabolites [10-12, 24, 25, 29]. In our study, we observed that over expression of ABCC5 in MCF-7 cells resulted in increased resistance against high concentrations of MTA. This can be explained by the increased e ux of MTA out of the cells through ABCC5, which consequently led to the decrease in MTA accumulation inside the cells and thus reduced activity and enhanced resistance. In general, it has been observed that up-regulated ABCC5 exports the nucleoside analogs and increase the drug resistance in the range of 2-to 10-fold during in vitro assays.
Clinically, the role of ABC transporter proteins in the intrinsic or acquired resistance is not clearly resolved. Uemura, et al suggested that ABCC11 directly confers the resistance to MTA by enhancing e ux of the intracellular anti-cancer drug in lung cancer [30], which suggested that ABCC11 maybe one of the biomarkers for MTA in the treatment of lung cancer. Oguri et al. found that the paclitaxel could induce the expression of ABCC10 gene, and which then in turn increase paclitaxel resistance by enhancing the paclitaxel e ux [35]. These groups also identi ed that ABCC5 expression is signi cantly associated with the sensitivity of a panel of non-small-cell lung cancer cell lines to gemcitabine. Inhibition of the transporter activity by small molecule inhibitors or siRNA knockdown, markedly sensitized the cancer cells to gemcitabine [36].
Moreover, the expression of ABCC5 has presented signi cantly over-expressed in the non-responding group after neoadjuvant chemotherapy than in the responding group in BC [37]. Nambaru PK et al. reported that ABCC5 e ux the monophosphorylated metabolite of 5-FU when ABCC5 was upregulated in colorectal and BC, which contributed to the 5-FU drug resistance [10]. Other researchers have identi ed that ABCC5 is expressed and functionally active in pancreatic adenocarcinoma cell lines and contributes to their sensitivities to different drugs [10,11]. Our study observed a close relationship between ABCC5 expression and cellular sensitivity to MTA in BC cells (R=0.741) too, and this phenomenon was accompanied by signi cantly decreased accumulation and enhanced e ux of MTA in ABCC5 overexpressing cell lines.
In conclusion, our results showed that ABC transporter ABCC5 is expressed, functionally active, and positively correlates with MTA resistance in BC. Further studies are still required to quantitatively assess the relationship between ABCC5 expression and the MTA dosage, so as to evaluate the ABCC5 expression as a biomarker for dosage optimization of MTA.