- Email: [email protected]
Delivery of doxorubicin across the blood–brain barrier by ondansetron pretreatment: a study in vitro and in vivo
Cancer Letters 353 (2014) 242–247
Contents lists available at ScienceDirect
Cancer Letters j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / c a n l e t
Original Articles
Delivery of doxorubicin across the blood–brain barrier by ondansetron pretreatment: a study in vitro and in vivo Iacopo Sardi a,b,*, Ornella Fantappiè c, Giancarlo la Marca d, Maria Grazia Giovannini e, Anna Lisa Iorio a,b, Martina da Ros a,b, Sabrina Malvagia d, Stefania Cardellicchio a,b, Laura Giunti a,b, Maurizio de Martino a,b, Roberto Mazzanti c a
Neuro-oncology Unit, Department of Paediatric Medicine, Anna Meyer Children’s University Hospital, Florence, Italy Department of Health Sciences, University of Florence, Florence, Italy c Second Medical Oncology Unit, Azienda Ospedaliero-Universitaria Careggi, Florence University, Florence, Italy d Newborn Screening, Biochemistry and Pharmacology Laboratory, Paediatric Neurology Unit and Laboratories, Department of Neuroscience, Meyer Children’s Hospital, Florence, Italy e Department of Health Sciences, Section of Pharmacology and Clinical Oncology, University of Florence, Florence, Italy b
A R T I C L E
I N F O
Article history: Received 31 May 2014 Accepted 14 July 2014 Keywords: Blood–brain barrier Doxorubicin Ondansetron Multidrug resistance Brain tumors
A B S T R A C T
Doxorubicin (Dox) has got a limited efficacy in the treatment of central nervous system tumors because of its poor penetration through blood–brain barrier mediated by MDR efflux transporters. We investigated the possibility that ondansetron (Ond) enhances Dox cytotoxicity in cell lines interfering with P-glycoprotein and increases Dox concentration in rat brain tissues. The MDR phenotype was studied using human hepatocellular carcinoma cell line PLC/PRF/5 (P5 and P1(0.5) clones), two subclones of NIH 3T3 cells (PSI-2 and PN1A) and two glioblastoma cell lines (A172, U87MG). Rats were pretreated with Ond before injection of Dox. Quantitative analysis of Dox was performed by mass spectrometry. Our in vitro experiments demonstrated that Ond at 10 μg/ml is not toxic to all cell lines. However, Ond reverses the MDR phenotype in P1(0.5) and PN1A cell lines. In addition, we showed that pretreatment with Ond increases Dox concentration in rat brain tissues, without increasing acute heart and renal toxicity. © 2014 Elsevier Ireland Ltd. All rights reserved.
Introduction Prerequisite for the efficacy of antineoplastic agents is that they reach efficacious concentration inside the cancerous cells. In tumors of the central nervous system (CNS) the achievement of therapeutic concentration of chemical agents is hampered by the presence of the P-glycoprotein (P-gp) efflux transporter localized on the blood– brain barrier (BBB), a physiological barrier that protects the brain from toxic insults [1,2]. Consequently, cure rate of CNS tumors by chemotherapy is not possible because of limited diffusion and accumulation of anticancer agents within the nervous parenchyma. Among the variety of anticancer drugs, anthracyclines show a potent effect in inhibiting tumor cell growth in several cancer cell lines and are now used to treat many type of neoplasms.
* Corresponding author. Address: Neuro-oncology Unit, Department of Paediatric Medicine, Meyer Children’s Hospital, Viale G. Pieraccini 24, 50139 Florence, Italy. Tel.: +39 055 5662631; fax: +39 055 5662746. E-mail address: [email protected] (I. Sardi). http://dx.doi.org/10.1016/j.canlet.2014.07.018 0304-3835/© 2014 Elsevier Ireland Ltd. All rights reserved.
Doxorubicin (Dox) is a highly effective inhibitor of growth of glioma cells in vitro and is toxic to glioblastoma cell lines in vivo [3,4]. Unfortunately, Dox has a poor penetration into the brain when intravenously administered due to the lack of drug penetration through the BBB. It has lately emerged that the mechanism of BBB-mediated drug resistance is complicated by the cooperation of P-gp (ABCB1) and breast cancer resistance protein (BCRP, ABCG2). They very efficiently remove several molecules from the CNS, thus limiting their entry into the brain [5]. These two “gatekeeper” efflux pumps work in synergy on the BBB and are usually present on the plasma membrane of many brain tumors. Inhibition of P-gp or BCRP can be compensated by the respective other transporter [6]. Interestingly, P-gp and BCRP have broad substrate specificity and interact with a range of chemically assorted molecules, including chemotherapeutic agents, such as anthracyclines [5,7] and many other drugs as ondansetron (Ond). Thus, it is conceivable that Ond and other chemical agents could inhibit P-gp and BCRP localized on the BBB, neurons and glial cells increasing the access of doxorubicin to the brain. Interestingly, we have recently verified in an animal model that morphine pretreatment increases the level of doxorubicin into the brain [8,9].
I. Sardi et al./Cancer Letters 353 (2014) 242–247
Searching for substrates for P-gp, we observed that drugs such as benzquinamide and Ond were substrates for P-gp and could interfere with anticancer efflux from cancer cells that overexpress the MDR phenotype [10]. This fact is interesting as both benzquinamide and Ond are antiemetics and especially Ond is now commonly used to control cancer chemotherapy induced emesis. This study therefore was performed to assess whether Ond enhances Dox cytotoxicity in glioblastoma cell lines interfering with P-gp and possesses the ability to increase Dox concentration in brain tissue without increasing heart and kidney toxicity.
243
Fig. 1. Western blot analysis for P-gp. The drug sensitive cell lines PSI-2, U87MG, A172 do not express P-gp. PN1A and P1(0.5) are P-gp positive cells. P5 has a lower expression of the protein than its MDR positive clone P1(0.5).
hemispheres, cerebellum, brainstem were collected, immediately frozen on dry ice and kept at −80 °C until assay.
Materials and methods Dox determination by mass spectrometry Cell lines and treatment in vitro Experiments were performed using different cell lines: human hepatocellular carcinoma drug sensitive cell line PLC/PRF/5 [11], clone (P5), and highly MDR positive P1(0.5) clone that was developed by prolonged serial exposures to increasing concentrations of Dox (Santa Cruz Biotechnology) starting from parental P5 cells as explained elsewhere [12]. P1(0.5) cell clone was cultured in DMEM containing 0.5 μg/ ml Dox and 10% fetal bovine serum (FBS), as previously reported [12]. Two subclones of NIH 3T3 cells were used; one was the parental drug sensitive clone (PSI-2) and the other was the MDR expressing clone (PN1A). These cells were cultured in DMEM supplemented with 10% FBS and 0.1 μg/ml Dox was added to the PN1A cell medium, as previously described [13]. Conferring the MDR phenotype to PN1A cell clone was obtained by transfection of PSI-2 cells with the pBA-mdr vector that contained mouse mdr1 cDNA as explained before [14]. Finally, we used U87MG and A172 glioblastoma cell lines, purchased from the American Type Culture Collection (ATCC) and cultured in EMEM or DMEM containing 10% FBS according to the manufacturer’s instructions (ATCC). Western blot analysis Preparation of total protein lysates and western blot analysis was performed as previously described [12] by using anti-P-gp (C219) (Gene Tex) and anti-β-actin (Sigma Aldrich) monoclonal antibody. Evaluation of toxic effect on cell survival Cells were plated in 60-mm Petri dishes (2 × 105 per dish) and cultured in serumcontained medium for 24 h. The next day the medium was removed and serumcontained medium with or without Dox (0.1 or 0.5 μg/ml) and Ond (10 or 30 μg/ ml) in combination or respectively, was added. After 3 days, the cells were washed twice with PBS, trypsinized, and counted by using the trypan blue exclusion method [15].
Dox was quantified in plasma and tissues by LC-MS/MS. An Applied BiosystemsSciex (Toronto, Canada) API 4000 bench-top TurboIonSpray-Triple-Quad Mass Spectrometer (MS/MS) was used for this study. The ion source operated under positive ion mode (5500 V). Declustering Potential (DP), Collision Exit Potential (CXP) and Collision Energy (CE) were automatically optimized for doxorubicin and 13C4-Dox. The resulting DP was 35 V. Optimal CE and CXP were found at 16 V and 10 V, respectively. From these experiments, the resulting most selective ion-pair transitions for the quantitative experiment (SRM) are 544.3 > 397.2 for doxorubicin and 548.3 > 401.2 for 13C4Dox. We have chosen one additional transition as qualifier: 544.3 > 379.2 for Dox and 548.3 > 383.2 for 13C4-Dox. Quantitative analysis was undertaken using an HPLC Series 1100 Agilent Technologies (Waldbronn, Germany). Liquid chromatography was performed using a Phenomenex Synergi 4 μm Polar-RP 80A 4 μm, 2 × 150 mm HPLC column (Phenomenex, Anzola Emilia, Italy) injecting 5 μl of the extracted sample. Column flow was 0.2 ml/min using an isocratic aqueous solution of 80% methanol containing 0.1% formic acid. The eluent from the column was directed to the TurboIonSpray probe without split ratio. LDH activity and tissue lipid peroxidation Lactate dehydrogenase (LDH) activity was measured in rat plasma using the Cytotoxicity Detection Kit (Roche Diagnostics GmbH, Mannheim, Germany) and absorbance was read at 490 nm. Tissue lipid peroxidation in terms of malondialdehyde (MDA) production (thiobarbituric acid-reactive species) was determined as previously described [16]. Statistics Data were analyzed using one-way ANOVA or Student’s t-test, as appropriate. Significance was set at P < 0.05.
Drugs and reagents
Results
Doxorubicin was purchased from Sandoz Spa, Varese, Italy; Ondansetron was (Zofran®) from GlaxoSmithKline Spa, Verona, Italy. The internal standard Doxorubicin-13C4 (98% D purity) was purchased from Medical Isotopes Inc. (Pelham, NY, USA). All other chemicals and solvents were of the highest purity grade commercially available.
In vitro experiments
Animals Male adult Wistar rats (260–280 g, Harlan Italia, Milano, Italy) were used. All animal manipulations were carried out according to the European Community guidelines for animal care (DL 116/92, application of the European Communities Council Directive 86/609/EEC). All efforts were made to minimize animal sufferings and to use only the number of animals necessary to produce reliable scientific data. Rats were randomly allocated to one of the following experimental groups: Group 1 (Ond + Dox): on day 1 rats were treated twice with Ond (2 mg/kg, i.p., at 10.00 a.m. and again at 5.00 p.m.); on day 2 at 10.00 a.m., the rats received a third dose of Ond; 1, 1.30 and 2 h later they received Dox (12 mg/kg, i.p.). Group 2 rats (Dox): rats were treated with Dox alone (12 mg/kg, i.p.). Group 3 rats (Control): naïve rats
Collection of brain tissue samples One hour after administration of Dox, rats were anesthetized with chloral hydrate (400 mg/kg, i.p.). The heart was exposed and animals were perfused transcardially with ice-cold saline (500 ml, 50 ml/min). At the end of perfusion, cerebral
The MDR phenotype and the P-gp expression were assessed in all cell lines (Fig. 1). P-gp was evidently expressed in PN1A and P1(0.5) cells; PSI-2, U87MG, A172 did not express P-gp whereas P5 clone showed light expression of the protein. Accordingly, we studied the cytotoxic effect of the Ond (10 and 30 μg/ml) exposure and the co-treatment Ond (10 and 30 μg/ml)/ Dox (0.1 μg/ml) to evaluate the responsiveness of the cell lines to the killing effect of these drugs. Our experiments demonstrated that Ond at 10 μg/ml is not toxic to all cell lines (Fig. 2A–C) while the higher concentration of 30 μg/ ml results to be slightly toxic to P5, PN1A, P1(0.5) cells that express P-gp protein (Fig. 2A, C). In addition, we showed that Ond reverses the MDR phenotype in PN1A and P1(0.5) (Fig. 3C). These cells that are resistant to Dox exhibit significant reduced survival when co-treated with Ond (10 and 30 μg/ml) plus Dox (0.1 μg/ml). The mean of P5 viable cells treated with Ond (30 μg/ml) was 0.4 ± 0.03 vs 0.7 ± 0.04 in the same untreated cells (Student’s t-test, P < 0.05). The mean of PN1A viable cells treated with Ond only (30 μg/ ml) was 0.95 ± 0.18 vs 1.85 ± 0.27 in the same untreated cells (Student’s t-test, P < 0.05) while the means of viable cells
244
I. Sardi et al./Cancer Letters 353 (2014) 242–247
Fig. 2. Effect of the Ond exposure (10 and 30 μg/ml). Ond (30 μg/ml) has a cytotoxic effect on P5 alone among the hepatocarcinoma (A) and glioma (B) drug sensitive cell lines. Panel C shows the responsiveness to Ond 30 μg/ml of the drug resistant P1(0.5) and PN1A cell lines (paradox cytotoxicity).
co-treated with Ond (10 and 30 μg/ml) plus Dox (0.1 μg/ml) were 0.8 ± 0.15 and 0.33 ± 0.08 respectively vs 1.48 ± 0.42 in the same untreated cells (Student’s t-test, P < 0.05). The mean of P1(0.5) viable cells treated with Ond only (30 μg/ ml) was 0.13 ± 0.02 vs 0.28 ± 0.03 in the same untreated cells (Student’s t-test, P < 0.05) while the means of viable cells cotreated with Ond (10 and 30 μg/ml) plus Dox (0.5 μg/ml) were 0.21 ± 0.03 and 0.14 ± 0.03 respectively vs 0.28 ± 0.03 in the same untreated cells (Student’s t-test, P < 0.05). The cytotoxic effect of Dox (0.1 μg/ml) on glioblastoma cell lines is shown in Fig. 3A. The mean of A172 viable cells treated with Dox was 0.21 ± 0.05 vs 1.18 ± 0.13 in the same untreated cells (Student’s t-test, P < 0.05) while the mean of U87MG viable cells treated with Dox was 0.23 ± 0.04 vs 1.1 ± 0.11 in the same untreated cells (Student’s t-test, P < 0.05). These cells remain highly sensitive to the killing effect of Dox even in the presence of Ond that does not change their responsiveness to the anticancer drug (Fig. 3B).
In vivo experiments According to our previous published data [17], the mean plasma levels of Dox in rats receiving the drug alone did not significantly differ from those found in animals pretreated with Ond only and Dox was undetectable in control samples from untreated rats (data not shown). We performed a time-course of the effect of pretreatment with Ond (2 mg/kg, i.p., three times in 24 h) on Dox penetration into the rat brain, administering the drug (12 mg/kg, i.p.) 1, 1.30 or 2 h after the last injection of Ond. Dox concentration was higher in all brain areas of animals pretreated with Ond (Fig. 4). The mean concentration of Dox in the brainstem of rats treated with it was 0.15 ± 0.03 ng/mg fresh tissue (n = 3) vs 0.25 ± 0.08 ng/ mg fresh tissue in those treated with Dox plus Ond at 1 h (+ 67%, n = 4), 0.31 ± 0.11 ng/mg fresh tissue in those treated with Dox plus Ond at 1.30 h (+ 106%, n = 4) and 0.42 ± 0.08 ng/mg fresh tissue in
I. Sardi et al./Cancer Letters 353 (2014) 242–247
245
Fig. 3. Effect of the co-treatment Ond/Dox. U87MG and A172 cell lines are very sensitive to Dox alone and Ond plus Dox treatment compared to untreated controls (A and B). The resistance to Dox disappears with complete reversion of the MDR phenotype in P1(0.5) and PN1A that are drug resistance cell lines. Ond 10 μg/ml or 30 μg/ml is equally effective for increasing Dox toxicity (C). *P < 0.05; **P < 0.01.
those treated with Dox plus Ond at 2 h (+ 180%, n = 4). This latter effect was statistically significant (one-way ANOVA F[3,11] = 6.294, P < 0.05; *P < 0.05 vs Dox only, Dunnett’s Multiple Comparison Test). The mean concentration of Dox in the cerebral hemispheres of rats treated with the anticancer agent only was 0.1 ± 0.01 ng/mg fresh tissue (n = 3) vs 0.19 ± 0.05 ng/mg fresh tissue in those treated with Dox plus Ond at 1 h (+ 90%, n = 5), 0.23 ± 0.07 ng/mg fresh tissue in those treated with Ond at 1.30 h (+ 130%, n = 4) and 0.35 ± 0.16 ng/mg fresh tissue in those treated with Ond at 2 h (+ 250%, n = 5). This latter effect was statistically significant (one-way ANOVA F[3,13] = 4.156, P < 0.05; *P < 0.05 vs Dox only, Dunnett’s Multiple Comparison Test). The mean concentration of Dox in the cerebellum of rats treated with this drug alone was 0.3 ± 0.07 ng/mg fresh tissue (n = 2) vs 0.5 ± 0.1 ng/mg fresh tissue in those treated with Dox plus Ond at 1 h (+ 67%, n = 5), 0.65 ± 0.06 ng/mg fresh tissue in those treated with Dox plus Ond at 1.30 h (+ 117%, n = 5) and 0.7 ± 0.2 ng/mg fresh tissue
in those treated with Dox plus Ond at 2 h (+ 133%, n = 5). The effect at 1.30 h was statistically significant (one-way ANOVA F[3,13] = 5.357, P < 0.05; *P < 0.05 vs Dox alone, Dunnett’s Multiple Comparison Test). The effect at 2 h was highly statistically significant (one-way ANOVA F[3,13] = 5.357, P < 0.05; **P < 0.01 vs Dox alone, Dunnett’s Multiple Comparison Test). Moreover, we investigated whether pretreatment with Ond might increase the Dox concentrations in the heart and kidney, thus predicting increased toxicity in these two target organs. Pretreatment with Ond did not increase the levels of Dox in either tissue (P > 0.05, Student’s t-test). We addressed the issue of cardiac and renal toxicity by analysis of plasma LDH activity and MDA levels. Dox induced a nonsignificant 60% increase of plasma LDH activity 1, 1.30 or 2 h after treatment in treated vs control rats (P > 0.05, Student’s t-test). No difference in LDH activity was found between rats treated with Dox alone and those receiving Ond plus Dox (P > 0.05, Student’s t-test) (data not shown).
246
I. Sardi et al./Cancer Letters 353 (2014) 242–247
Fig. 4. Time course of the effect of pretreatment with Ond on Dox penetration into the brain of the rat. Dox (12 mg/kg, i.p.) is administered to different groups of rats either alone (white bars) or at different times (1, 1.30 or 2 h) after pretreatment with Ond (2 mg/kg, i.p.). Tissue levels of Dox are expressed as ng/mg fresh tissue (mean ± SEM). *P < 0.05; **P < 0.01.
Discussion The core result of this study is that pretreatment with therapeutic dose of Ond allows delivery of significantly higher amounts of Dox to the brain tissue in vivo, which is otherwise disallowed by the BBB. This effect of Ond is most likely due to the fact that Ond and Dox compete for P-gp mediated transport at BBB level. Dox is commonly used in the treatment of a wide range of tumors [18]. Interestingly, data on models of malignant glioma suggest that Dox is an effective anti-tumor agent in vitro and in vivo [19–21]. Eramo and co-workers have demonstrated that Dox displays high antitumor activity also against glioblastoma stem cells [22]. However, anthracyclines have very limited efficacy in the treatment of brain tumors because of their poor penetration through BBB due to the MDR efflux transporters [23]. Several approaches have been tested for improving anthracyclines delivery to CNS through circumvention of the BBB. Recently, intracerebral drug delivery of Dox via a polymer matrix in brain tumors of rat has been successfully performed [3]. A clinical study has shown promising results with intralesional Dox infusion via an Ommaya reservoir without significant systemic side effects for the therapy of malignant gliomas [24]. In vitro studies have recognized the MDR efflux pumps such as P-gp, BCRP and other MDR-associated proteins in primary cultures of cerebral endothelial cells. These molecules are located at both the luminal and apical surfaces of the BBB to protect the CNS against xenobiotic agents [25,26] and they are therefore capable of transporting a wide assortment of chemotherapeutic agents such as anthracyclines, methotrexate, etoposide, taxols and Vinca alkaloids [5,7]. Morphine and ondansetron are known substrates of P-gp [5,27]. Thus, it is conceivable that these compounds could increase the access of Dox to the brain competing with the same efflux transporters as P-gp and BCRP [5]. The ability of Ond to modulate the MDR efflux pumps is of clinical interest. Here, we confirmed our unpublished old observations on the effect of Ond on Dox efflux from cancer and not cancer cells that express P-gp and the MDR phenotype in vitro. In addition, we showed that Ond is not toxic to human hepatocellular carcinoma (PSI-2) and glioblastoma cell lines (U87MG, A172) that do not express P-gp and the MDR phenotype. Conversely, Ond is able to induce a cytotoxic effect on the P-gp and MDR positive P1(0.5) and PN1A cell lines (paradox cytotoxicity). Moreover, our data show that Ond reverses the MDR phenotype in cancer (HCC cell lines) and not
transformed cell lines (NIH 3T3) at a concentration that can be reached during cancer treatment, at least in vitro. These observations, taken together to the fact that U87MG and A172 cells are very sensitive to Dox treatment and the co-exposure to Ond does not change their sensitivity to the cytotoxic effect of the drug, suggest that Ond and Dox could be used together to treat high grade gliomas. Ond is a selective 5-hydroxytryptamine(3) (5-HT(3)) receptor antagonist, mainly used in clinical practice as an antiemetic for prophylaxis and treatment of nausea and vomiting related to chemotherapy and anesthesia. It acts on both the peripheral and the CNS. However, CNS concentrations are less than 15% of plasma concentrations, indicating that the rate of penetration of the blood– brain barrier by Ond is very low [28]. Accordingly to our experiments in vitro, in this study we demonstrated a significantly higher Dox level in all brain areas 2 h after pretreatment with Ond (cerebral hemispheres and brainstem P < 0.05; cerebellum P < 0.01). Since Ond is substrate of P-gp and increases Dox concentration in cells overexpressing P-gp these results suggest that Ond increases the Dox concentration inside the brain competing for P-gp mediated drug efflux at BBB level. This fact results in higher intratissutal concentration of the anticancer drug in cancer most likely reaching an efficacious concentration for inhibiting cancer cells growth. Anthracyclines may induce acute cardiotoxicity; in addition, high cumulative doses are associated with late-onset cardiomyopathy that is refractory to standard treatment [29]. Free radical generation and p53-dependent apoptosis by modulation of mTOR activity are supposed to contribute to Dox-induced cardiotoxicity [30,31]. On this basis, Ond could increase the toxicity of Dox on heart and kidney once given during cancer chemotherapy. So, we quantified Dox levels in heart and kidney of treated rats and we observed that the pre-treatment with Ond did not increase the levels of Dox in either tissue 1, 1.30 and 2 h after administration. We also evaluated plasma LDH activity and MDA levels, as markers of acute cardiotoxicity, and found no difference between rats treated with Dox alone and those also receiving Ond. These results are in agreement with the fact that cardiomyocytes do not express P-gp, while kidney has several transport proteins to efflux Dox into the urine. In conclusion, our data suggest that competition between Ond and Dox for P-gp mediated transport at BBB level increases significantly Dox penetration into the brain of the rat, without increasing cardiac or renal toxicity. These preliminary results on a rodent model will enable us to novel therapeutic approaches for refractory or
I. Sardi et al./Cancer Letters 353 (2014) 242–247
recurrent brain tumors where anticancer drugs are usually cleared by the BBB. Whether this phenomenon may have a therapeutic impact remains to be elucidated. Conflict of interest None. Acknowledgements This work was supported by: Associazione Italiana per la Ricerca sul Cancro (AIRC), grant IG–12799; “Amicodivalerio” Onlus; “Noi per Voi” Onlus, Fondazione Tommasino Bacciotti. References [1] B.A. Barres, The mystery and magic of glia: a perspective on their roles in health and disease, Neuron 60 (2008) 430–440. [2] E.A. Neuwelt, B. Bauer, C. Fahlke, G. Fricker, C. Iadecola, D. Janigro, et al., Engaging neuroscience to advance translational research in brain barrier biology, Nat. Rev. Neurosci. 12 (2011) 169–182. [3] M.S. Lesniak, U. Upadhyay, R. Goodwin, B. Tyler, H. Brem, Local delivery of doxorubicin for the treatment of malignant brain tumors in rats, Anticancer Res. 25 (2005) 3825–3831. [4] A.C. Stan, S. Casares, D. Radu, G.F. Walter, T.D. Brumeanu, Doxorubicin-induced cell death in highly invasive human gliomas, Anticancer Res. 19 (1999) 941–950. [5] W. Löscher, H. Potschka, Role of drug efflux transporters in the brain for drug disposition and treatment of brain diseases, Prog. Neurobiol. 76 (2005) 22–76. [6] S. Agarwal, A.M. Hartz, W.F. Elmquist, B. Bauer, Breast cancer resistance protein and P-glycoprotein in brain cancer: two gatekeepers team up, Curr. Pharm. Des. 17 (2011) 2793–2802. [7] P. Breedveld, J.H. Beijnen, J.H. Schellens, Use of P-glycoprotein and BCRP inhibitors to improve oral bioavailability and CNS penetration of anticancer drugs, Trends Pharmacol. Sci. 27 (2006) 17–24. [8] I. Sardi, G. la Marca, M.G. Giovannini, S. Malvagia, R. Guerrini, L. Genitori, et al., Detection of doxorubicin hydrochloride accumulation in the rat brain after morphine treatment by mass spectrometry, Cancer Chemother. Pharmacol. 67 (2011) 1333–1340. [9] I. Sardi, Morphine facilitates doxorubicin penetration in the central nervous system: a new prospect for therapy of brain tumors, J. Neurooncol. 104 (2011) 619–620. [10] R. Mazzanti, J.M. Croop, Z. Gatmaitan, M. Budding, K. Steiglitz, R. Arceci, I.M. Arias, Benzquinamide inhibits P-glycoprotein mediated drug efflux and potentiates anticancer agent cytotoxicity in multidrug resistant cells, Oncol. Res. 4 (1992) 359–365. [11] G.M. MacNab, J.J. Alexander, G. Lecatsas, E.M. Bey, J.M. Urbanowicz, Hepatitis B surface antigen produced by a human hepatoma cell line, Br. J. Cancer 34 (1976) 509–515. [12] O. Fantappiè, E. Masini, I. Sardi, L. Raimondi, D. Bani, M. Solazzo, et al., The MDR phenotype is associated with the expression of COX-2 and iNOS in a human hepatocellular carcinoma cell line, Hepatology 35 (2002) 843–852.
247
[13] O. Fantappiè, M. Solazzo, N. Lasagna, F. Platini, L. Tessitore, R. Mazzanti, P-glycoprotein mediates celecoxib-induced apoptosis in multiple drug-resistant cell lines, Cancer Res. 67 (2007) 4915–4923. [14] J.M. Croop, B.C. Guild, P. Gros, D.E. Housman, Genetics of multidrug resistance: relationship of a cloned gene to the complete multidrug resistant phenotype, Cancer Res. 47 (1987) 5982–5988. [15] N. Lasagna, O. Fantappiè, M. Solazzo, L. Morbidelli, S. Marchetti, G. Cipriani, et al., Hepatocyte growth factor and inducible nitric oxide synthase are involved in multidrug resistance-induced angiogenesis in hepatocellular carcinoma cell lines, Cancer Res. 66 (2006) 2673–2682. [16] M. Uchiyama, M. Mihara, Determination of malondialdehyde precursor in tissues by thiobarbituric acid test, Anal. Biochem. 86 (1978) 271–278. [17] I. Sardi, G. la Marca, S. Cardellicchio, L. Giunti, S. Malvagia, L. Genitori, et al., Pharmacological modulation of blood-brain barrier increases permeability of doxorubicin into the rat brain, Am. J. Cancer Res. 3 (4) (2013) 424–432. [18] R.S. Benjamin, C.E. Riggs Jr., N.R. Bachur, Pharmacokinetics and metabolism of adriamycin in man, Clin. Pharmacol. Ther. 14 (1973) 592–600. [19] T. Abe, S. Hasegawa, K. Taniguchi, A. Yokomizo, T. Kuwano, M. Ono, et al., Possible involvement of multidrug-resistance-associated protein (MRP) gene expression in spontaneous drug resistance to vincristine, etoposide and adriamycin in human glioma cells, Int. J. Cancer 58 (1994) 860–864. [20] J.E. Wolff, T. Trilling, G. Mölenkamp, R.M. Egeler, H. Jürgens, Chemosensitivity of glioma cells in vitro: a meta analysis, J. Cancer Res. Clin. Oncol. 125 (1999) 481–486. [21] S.C. Steiniger, J. Kreuter, A.S. Khalansky, I.N. Skidan, A.I. Bobruskin, Z.S. Smirnova, et al., Chemotherapy of glioblastoma in rats using doxorubicin-loaded nanoparticles, Int. J. Cancer 109 (2004) 759–767. [22] A. Eramo, L. Ricci-Vitiani, A. Zeuner, R. Pallini, F. Lotti, G. Sette, et al., Chemotherapy resistance of glioblastoma stem cells, Cell Death Differ. 13 (2006) 1238–1241. [23] T. Ohnishi, I. Tamai, K. Sakanaka, A. Sakata, T. Yamashima, J. Yamashita, et al., In vivo and in vitro evidence for ATP-dependency of P-glycoprotein-mediated efflux of doxorubicin at the blood-brain barrier, Biochem. Pharmacol. 49 (1995) 1541–1544. [24] S. Voulgaris, M. Partheni, M. Karamouzis, P. Dimopoulos, N. Papadakis, H.P. Kalofonos, Intratumoral doxorubicin in patients with malignant brain gliomas, Am. J. Clin. Oncol. 25 (2002) 60–64. [25] S. Hori, S. Ohtsuki, M. Tachikawa, N. Kimura, T. Kondo, M. Watanabe, et al., Functional expression of rat ABCG2 on the luminal side of brain capillaries and its enhancement by astrocyte-derived soluble factor(s), J. Neurochem. 90 (2004) 526–536. [26] N.J. Abbott, A.A. Patabendige, D.E. Dolman, S.R. Yusof, D.J. Begley, Structure and function of the blood-brain barrier, Neurobiol. Dis. 37 (2010) 13–25. [27] A.H. Schinkel, E. Wagenaar, C.A. Mol, L. van Deemter, P-glycoprotein in the blood-brain barrier of mice influences the brain penetration and pharmacological activity of many drugs, J. Clin. Invest. 97 (1996) 2517– 2524. [28] K.H. Simpson, F.M. Hicks, Clinical pharmacokinetics of ondansetron. A review, J. Pharm. Pharmacol. 48 (1996) 774–781. [29] P.K. Singal, N. Iliskovic, Doxorubicin-induced cardiomyopathy, N. Engl. J. Med. 339 (1998) 900–905. [30] S. Zhou, A. Starkov, M.K. Froberg, R.L. Leino, K.B. Wallace, Cumulative and irreversible cardiac mitochondrial dysfunction induced by doxorubicin, Cancer Res. 61 (2001) 771–777. [31] W. Zhu, M.H. Soonpaa, H. Chen, W. Shen, R.M. Payne, E.A. Liechty, et al., Acute doxorubicin cardiotoxicity is associated with p53-induced inhibition of the mammalian target of rapamycin pathway, Circulation 119 (2009) 99–106.
Contents lists available at ScienceDirect
Cancer Letters j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / c a n l e t
Original Articles
Delivery of doxorubicin across the blood–brain barrier by ondansetron pretreatment: a study in vitro and in vivo Iacopo Sardi a,b,*, Ornella Fantappiè c, Giancarlo la Marca d, Maria Grazia Giovannini e, Anna Lisa Iorio a,b, Martina da Ros a,b, Sabrina Malvagia d, Stefania Cardellicchio a,b, Laura Giunti a,b, Maurizio de Martino a,b, Roberto Mazzanti c a
Neuro-oncology Unit, Department of Paediatric Medicine, Anna Meyer Children’s University Hospital, Florence, Italy Department of Health Sciences, University of Florence, Florence, Italy c Second Medical Oncology Unit, Azienda Ospedaliero-Universitaria Careggi, Florence University, Florence, Italy d Newborn Screening, Biochemistry and Pharmacology Laboratory, Paediatric Neurology Unit and Laboratories, Department of Neuroscience, Meyer Children’s Hospital, Florence, Italy e Department of Health Sciences, Section of Pharmacology and Clinical Oncology, University of Florence, Florence, Italy b
A R T I C L E
I N F O
Article history: Received 31 May 2014 Accepted 14 July 2014 Keywords: Blood–brain barrier Doxorubicin Ondansetron Multidrug resistance Brain tumors
A B S T R A C T
Doxorubicin (Dox) has got a limited efficacy in the treatment of central nervous system tumors because of its poor penetration through blood–brain barrier mediated by MDR efflux transporters. We investigated the possibility that ondansetron (Ond) enhances Dox cytotoxicity in cell lines interfering with P-glycoprotein and increases Dox concentration in rat brain tissues. The MDR phenotype was studied using human hepatocellular carcinoma cell line PLC/PRF/5 (P5 and P1(0.5) clones), two subclones of NIH 3T3 cells (PSI-2 and PN1A) and two glioblastoma cell lines (A172, U87MG). Rats were pretreated with Ond before injection of Dox. Quantitative analysis of Dox was performed by mass spectrometry. Our in vitro experiments demonstrated that Ond at 10 μg/ml is not toxic to all cell lines. However, Ond reverses the MDR phenotype in P1(0.5) and PN1A cell lines. In addition, we showed that pretreatment with Ond increases Dox concentration in rat brain tissues, without increasing acute heart and renal toxicity. © 2014 Elsevier Ireland Ltd. All rights reserved.
Introduction Prerequisite for the efficacy of antineoplastic agents is that they reach efficacious concentration inside the cancerous cells. In tumors of the central nervous system (CNS) the achievement of therapeutic concentration of chemical agents is hampered by the presence of the P-glycoprotein (P-gp) efflux transporter localized on the blood– brain barrier (BBB), a physiological barrier that protects the brain from toxic insults [1,2]. Consequently, cure rate of CNS tumors by chemotherapy is not possible because of limited diffusion and accumulation of anticancer agents within the nervous parenchyma. Among the variety of anticancer drugs, anthracyclines show a potent effect in inhibiting tumor cell growth in several cancer cell lines and are now used to treat many type of neoplasms.
* Corresponding author. Address: Neuro-oncology Unit, Department of Paediatric Medicine, Meyer Children’s Hospital, Viale G. Pieraccini 24, 50139 Florence, Italy. Tel.: +39 055 5662631; fax: +39 055 5662746. E-mail address: [email protected] (I. Sardi). http://dx.doi.org/10.1016/j.canlet.2014.07.018 0304-3835/© 2014 Elsevier Ireland Ltd. All rights reserved.
Doxorubicin (Dox) is a highly effective inhibitor of growth of glioma cells in vitro and is toxic to glioblastoma cell lines in vivo [3,4]. Unfortunately, Dox has a poor penetration into the brain when intravenously administered due to the lack of drug penetration through the BBB. It has lately emerged that the mechanism of BBB-mediated drug resistance is complicated by the cooperation of P-gp (ABCB1) and breast cancer resistance protein (BCRP, ABCG2). They very efficiently remove several molecules from the CNS, thus limiting their entry into the brain [5]. These two “gatekeeper” efflux pumps work in synergy on the BBB and are usually present on the plasma membrane of many brain tumors. Inhibition of P-gp or BCRP can be compensated by the respective other transporter [6]. Interestingly, P-gp and BCRP have broad substrate specificity and interact with a range of chemically assorted molecules, including chemotherapeutic agents, such as anthracyclines [5,7] and many other drugs as ondansetron (Ond). Thus, it is conceivable that Ond and other chemical agents could inhibit P-gp and BCRP localized on the BBB, neurons and glial cells increasing the access of doxorubicin to the brain. Interestingly, we have recently verified in an animal model that morphine pretreatment increases the level of doxorubicin into the brain [8,9].
I. Sardi et al./Cancer Letters 353 (2014) 242–247
Searching for substrates for P-gp, we observed that drugs such as benzquinamide and Ond were substrates for P-gp and could interfere with anticancer efflux from cancer cells that overexpress the MDR phenotype [10]. This fact is interesting as both benzquinamide and Ond are antiemetics and especially Ond is now commonly used to control cancer chemotherapy induced emesis. This study therefore was performed to assess whether Ond enhances Dox cytotoxicity in glioblastoma cell lines interfering with P-gp and possesses the ability to increase Dox concentration in brain tissue without increasing heart and kidney toxicity.
243
Fig. 1. Western blot analysis for P-gp. The drug sensitive cell lines PSI-2, U87MG, A172 do not express P-gp. PN1A and P1(0.5) are P-gp positive cells. P5 has a lower expression of the protein than its MDR positive clone P1(0.5).
hemispheres, cerebellum, brainstem were collected, immediately frozen on dry ice and kept at −80 °C until assay.
Materials and methods Dox determination by mass spectrometry Cell lines and treatment in vitro Experiments were performed using different cell lines: human hepatocellular carcinoma drug sensitive cell line PLC/PRF/5 [11], clone (P5), and highly MDR positive P1(0.5) clone that was developed by prolonged serial exposures to increasing concentrations of Dox (Santa Cruz Biotechnology) starting from parental P5 cells as explained elsewhere [12]. P1(0.5) cell clone was cultured in DMEM containing 0.5 μg/ ml Dox and 10% fetal bovine serum (FBS), as previously reported [12]. Two subclones of NIH 3T3 cells were used; one was the parental drug sensitive clone (PSI-2) and the other was the MDR expressing clone (PN1A). These cells were cultured in DMEM supplemented with 10% FBS and 0.1 μg/ml Dox was added to the PN1A cell medium, as previously described [13]. Conferring the MDR phenotype to PN1A cell clone was obtained by transfection of PSI-2 cells with the pBA-mdr vector that contained mouse mdr1 cDNA as explained before [14]. Finally, we used U87MG and A172 glioblastoma cell lines, purchased from the American Type Culture Collection (ATCC) and cultured in EMEM or DMEM containing 10% FBS according to the manufacturer’s instructions (ATCC). Western blot analysis Preparation of total protein lysates and western blot analysis was performed as previously described [12] by using anti-P-gp (C219) (Gene Tex) and anti-β-actin (Sigma Aldrich) monoclonal antibody. Evaluation of toxic effect on cell survival Cells were plated in 60-mm Petri dishes (2 × 105 per dish) and cultured in serumcontained medium for 24 h. The next day the medium was removed and serumcontained medium with or without Dox (0.1 or 0.5 μg/ml) and Ond (10 or 30 μg/ ml) in combination or respectively, was added. After 3 days, the cells were washed twice with PBS, trypsinized, and counted by using the trypan blue exclusion method [15].
Dox was quantified in plasma and tissues by LC-MS/MS. An Applied BiosystemsSciex (Toronto, Canada) API 4000 bench-top TurboIonSpray-Triple-Quad Mass Spectrometer (MS/MS) was used for this study. The ion source operated under positive ion mode (5500 V). Declustering Potential (DP), Collision Exit Potential (CXP) and Collision Energy (CE) were automatically optimized for doxorubicin and 13C4-Dox. The resulting DP was 35 V. Optimal CE and CXP were found at 16 V and 10 V, respectively. From these experiments, the resulting most selective ion-pair transitions for the quantitative experiment (SRM) are 544.3 > 397.2 for doxorubicin and 548.3 > 401.2 for 13C4Dox. We have chosen one additional transition as qualifier: 544.3 > 379.2 for Dox and 548.3 > 383.2 for 13C4-Dox. Quantitative analysis was undertaken using an HPLC Series 1100 Agilent Technologies (Waldbronn, Germany). Liquid chromatography was performed using a Phenomenex Synergi 4 μm Polar-RP 80A 4 μm, 2 × 150 mm HPLC column (Phenomenex, Anzola Emilia, Italy) injecting 5 μl of the extracted sample. Column flow was 0.2 ml/min using an isocratic aqueous solution of 80% methanol containing 0.1% formic acid. The eluent from the column was directed to the TurboIonSpray probe without split ratio. LDH activity and tissue lipid peroxidation Lactate dehydrogenase (LDH) activity was measured in rat plasma using the Cytotoxicity Detection Kit (Roche Diagnostics GmbH, Mannheim, Germany) and absorbance was read at 490 nm. Tissue lipid peroxidation in terms of malondialdehyde (MDA) production (thiobarbituric acid-reactive species) was determined as previously described [16]. Statistics Data were analyzed using one-way ANOVA or Student’s t-test, as appropriate. Significance was set at P < 0.05.
Drugs and reagents
Results
Doxorubicin was purchased from Sandoz Spa, Varese, Italy; Ondansetron was (Zofran®) from GlaxoSmithKline Spa, Verona, Italy. The internal standard Doxorubicin-13C4 (98% D purity) was purchased from Medical Isotopes Inc. (Pelham, NY, USA). All other chemicals and solvents were of the highest purity grade commercially available.
In vitro experiments
Animals Male adult Wistar rats (260–280 g, Harlan Italia, Milano, Italy) were used. All animal manipulations were carried out according to the European Community guidelines for animal care (DL 116/92, application of the European Communities Council Directive 86/609/EEC). All efforts were made to minimize animal sufferings and to use only the number of animals necessary to produce reliable scientific data. Rats were randomly allocated to one of the following experimental groups: Group 1 (Ond + Dox): on day 1 rats were treated twice with Ond (2 mg/kg, i.p., at 10.00 a.m. and again at 5.00 p.m.); on day 2 at 10.00 a.m., the rats received a third dose of Ond; 1, 1.30 and 2 h later they received Dox (12 mg/kg, i.p.). Group 2 rats (Dox): rats were treated with Dox alone (12 mg/kg, i.p.). Group 3 rats (Control): naïve rats
Collection of brain tissue samples One hour after administration of Dox, rats were anesthetized with chloral hydrate (400 mg/kg, i.p.). The heart was exposed and animals were perfused transcardially with ice-cold saline (500 ml, 50 ml/min). At the end of perfusion, cerebral
The MDR phenotype and the P-gp expression were assessed in all cell lines (Fig. 1). P-gp was evidently expressed in PN1A and P1(0.5) cells; PSI-2, U87MG, A172 did not express P-gp whereas P5 clone showed light expression of the protein. Accordingly, we studied the cytotoxic effect of the Ond (10 and 30 μg/ml) exposure and the co-treatment Ond (10 and 30 μg/ml)/ Dox (0.1 μg/ml) to evaluate the responsiveness of the cell lines to the killing effect of these drugs. Our experiments demonstrated that Ond at 10 μg/ml is not toxic to all cell lines (Fig. 2A–C) while the higher concentration of 30 μg/ ml results to be slightly toxic to P5, PN1A, P1(0.5) cells that express P-gp protein (Fig. 2A, C). In addition, we showed that Ond reverses the MDR phenotype in PN1A and P1(0.5) (Fig. 3C). These cells that are resistant to Dox exhibit significant reduced survival when co-treated with Ond (10 and 30 μg/ml) plus Dox (0.1 μg/ml). The mean of P5 viable cells treated with Ond (30 μg/ml) was 0.4 ± 0.03 vs 0.7 ± 0.04 in the same untreated cells (Student’s t-test, P < 0.05). The mean of PN1A viable cells treated with Ond only (30 μg/ ml) was 0.95 ± 0.18 vs 1.85 ± 0.27 in the same untreated cells (Student’s t-test, P < 0.05) while the means of viable cells
244
I. Sardi et al./Cancer Letters 353 (2014) 242–247
Fig. 2. Effect of the Ond exposure (10 and 30 μg/ml). Ond (30 μg/ml) has a cytotoxic effect on P5 alone among the hepatocarcinoma (A) and glioma (B) drug sensitive cell lines. Panel C shows the responsiveness to Ond 30 μg/ml of the drug resistant P1(0.5) and PN1A cell lines (paradox cytotoxicity).
co-treated with Ond (10 and 30 μg/ml) plus Dox (0.1 μg/ml) were 0.8 ± 0.15 and 0.33 ± 0.08 respectively vs 1.48 ± 0.42 in the same untreated cells (Student’s t-test, P < 0.05). The mean of P1(0.5) viable cells treated with Ond only (30 μg/ ml) was 0.13 ± 0.02 vs 0.28 ± 0.03 in the same untreated cells (Student’s t-test, P < 0.05) while the means of viable cells cotreated with Ond (10 and 30 μg/ml) plus Dox (0.5 μg/ml) were 0.21 ± 0.03 and 0.14 ± 0.03 respectively vs 0.28 ± 0.03 in the same untreated cells (Student’s t-test, P < 0.05). The cytotoxic effect of Dox (0.1 μg/ml) on glioblastoma cell lines is shown in Fig. 3A. The mean of A172 viable cells treated with Dox was 0.21 ± 0.05 vs 1.18 ± 0.13 in the same untreated cells (Student’s t-test, P < 0.05) while the mean of U87MG viable cells treated with Dox was 0.23 ± 0.04 vs 1.1 ± 0.11 in the same untreated cells (Student’s t-test, P < 0.05). These cells remain highly sensitive to the killing effect of Dox even in the presence of Ond that does not change their responsiveness to the anticancer drug (Fig. 3B).
In vivo experiments According to our previous published data [17], the mean plasma levels of Dox in rats receiving the drug alone did not significantly differ from those found in animals pretreated with Ond only and Dox was undetectable in control samples from untreated rats (data not shown). We performed a time-course of the effect of pretreatment with Ond (2 mg/kg, i.p., three times in 24 h) on Dox penetration into the rat brain, administering the drug (12 mg/kg, i.p.) 1, 1.30 or 2 h after the last injection of Ond. Dox concentration was higher in all brain areas of animals pretreated with Ond (Fig. 4). The mean concentration of Dox in the brainstem of rats treated with it was 0.15 ± 0.03 ng/mg fresh tissue (n = 3) vs 0.25 ± 0.08 ng/ mg fresh tissue in those treated with Dox plus Ond at 1 h (+ 67%, n = 4), 0.31 ± 0.11 ng/mg fresh tissue in those treated with Dox plus Ond at 1.30 h (+ 106%, n = 4) and 0.42 ± 0.08 ng/mg fresh tissue in
I. Sardi et al./Cancer Letters 353 (2014) 242–247
245
Fig. 3. Effect of the co-treatment Ond/Dox. U87MG and A172 cell lines are very sensitive to Dox alone and Ond plus Dox treatment compared to untreated controls (A and B). The resistance to Dox disappears with complete reversion of the MDR phenotype in P1(0.5) and PN1A that are drug resistance cell lines. Ond 10 μg/ml or 30 μg/ml is equally effective for increasing Dox toxicity (C). *P < 0.05; **P < 0.01.
those treated with Dox plus Ond at 2 h (+ 180%, n = 4). This latter effect was statistically significant (one-way ANOVA F[3,11] = 6.294, P < 0.05; *P < 0.05 vs Dox only, Dunnett’s Multiple Comparison Test). The mean concentration of Dox in the cerebral hemispheres of rats treated with the anticancer agent only was 0.1 ± 0.01 ng/mg fresh tissue (n = 3) vs 0.19 ± 0.05 ng/mg fresh tissue in those treated with Dox plus Ond at 1 h (+ 90%, n = 5), 0.23 ± 0.07 ng/mg fresh tissue in those treated with Ond at 1.30 h (+ 130%, n = 4) and 0.35 ± 0.16 ng/mg fresh tissue in those treated with Ond at 2 h (+ 250%, n = 5). This latter effect was statistically significant (one-way ANOVA F[3,13] = 4.156, P < 0.05; *P < 0.05 vs Dox only, Dunnett’s Multiple Comparison Test). The mean concentration of Dox in the cerebellum of rats treated with this drug alone was 0.3 ± 0.07 ng/mg fresh tissue (n = 2) vs 0.5 ± 0.1 ng/mg fresh tissue in those treated with Dox plus Ond at 1 h (+ 67%, n = 5), 0.65 ± 0.06 ng/mg fresh tissue in those treated with Dox plus Ond at 1.30 h (+ 117%, n = 5) and 0.7 ± 0.2 ng/mg fresh tissue
in those treated with Dox plus Ond at 2 h (+ 133%, n = 5). The effect at 1.30 h was statistically significant (one-way ANOVA F[3,13] = 5.357, P < 0.05; *P < 0.05 vs Dox alone, Dunnett’s Multiple Comparison Test). The effect at 2 h was highly statistically significant (one-way ANOVA F[3,13] = 5.357, P < 0.05; **P < 0.01 vs Dox alone, Dunnett’s Multiple Comparison Test). Moreover, we investigated whether pretreatment with Ond might increase the Dox concentrations in the heart and kidney, thus predicting increased toxicity in these two target organs. Pretreatment with Ond did not increase the levels of Dox in either tissue (P > 0.05, Student’s t-test). We addressed the issue of cardiac and renal toxicity by analysis of plasma LDH activity and MDA levels. Dox induced a nonsignificant 60% increase of plasma LDH activity 1, 1.30 or 2 h after treatment in treated vs control rats (P > 0.05, Student’s t-test). No difference in LDH activity was found between rats treated with Dox alone and those receiving Ond plus Dox (P > 0.05, Student’s t-test) (data not shown).
246
I. Sardi et al./Cancer Letters 353 (2014) 242–247
Fig. 4. Time course of the effect of pretreatment with Ond on Dox penetration into the brain of the rat. Dox (12 mg/kg, i.p.) is administered to different groups of rats either alone (white bars) or at different times (1, 1.30 or 2 h) after pretreatment with Ond (2 mg/kg, i.p.). Tissue levels of Dox are expressed as ng/mg fresh tissue (mean ± SEM). *P < 0.05; **P < 0.01.
Discussion The core result of this study is that pretreatment with therapeutic dose of Ond allows delivery of significantly higher amounts of Dox to the brain tissue in vivo, which is otherwise disallowed by the BBB. This effect of Ond is most likely due to the fact that Ond and Dox compete for P-gp mediated transport at BBB level. Dox is commonly used in the treatment of a wide range of tumors [18]. Interestingly, data on models of malignant glioma suggest that Dox is an effective anti-tumor agent in vitro and in vivo [19–21]. Eramo and co-workers have demonstrated that Dox displays high antitumor activity also against glioblastoma stem cells [22]. However, anthracyclines have very limited efficacy in the treatment of brain tumors because of their poor penetration through BBB due to the MDR efflux transporters [23]. Several approaches have been tested for improving anthracyclines delivery to CNS through circumvention of the BBB. Recently, intracerebral drug delivery of Dox via a polymer matrix in brain tumors of rat has been successfully performed [3]. A clinical study has shown promising results with intralesional Dox infusion via an Ommaya reservoir without significant systemic side effects for the therapy of malignant gliomas [24]. In vitro studies have recognized the MDR efflux pumps such as P-gp, BCRP and other MDR-associated proteins in primary cultures of cerebral endothelial cells. These molecules are located at both the luminal and apical surfaces of the BBB to protect the CNS against xenobiotic agents [25,26] and they are therefore capable of transporting a wide assortment of chemotherapeutic agents such as anthracyclines, methotrexate, etoposide, taxols and Vinca alkaloids [5,7]. Morphine and ondansetron are known substrates of P-gp [5,27]. Thus, it is conceivable that these compounds could increase the access of Dox to the brain competing with the same efflux transporters as P-gp and BCRP [5]. The ability of Ond to modulate the MDR efflux pumps is of clinical interest. Here, we confirmed our unpublished old observations on the effect of Ond on Dox efflux from cancer and not cancer cells that express P-gp and the MDR phenotype in vitro. In addition, we showed that Ond is not toxic to human hepatocellular carcinoma (PSI-2) and glioblastoma cell lines (U87MG, A172) that do not express P-gp and the MDR phenotype. Conversely, Ond is able to induce a cytotoxic effect on the P-gp and MDR positive P1(0.5) and PN1A cell lines (paradox cytotoxicity). Moreover, our data show that Ond reverses the MDR phenotype in cancer (HCC cell lines) and not
transformed cell lines (NIH 3T3) at a concentration that can be reached during cancer treatment, at least in vitro. These observations, taken together to the fact that U87MG and A172 cells are very sensitive to Dox treatment and the co-exposure to Ond does not change their sensitivity to the cytotoxic effect of the drug, suggest that Ond and Dox could be used together to treat high grade gliomas. Ond is a selective 5-hydroxytryptamine(3) (5-HT(3)) receptor antagonist, mainly used in clinical practice as an antiemetic for prophylaxis and treatment of nausea and vomiting related to chemotherapy and anesthesia. It acts on both the peripheral and the CNS. However, CNS concentrations are less than 15% of plasma concentrations, indicating that the rate of penetration of the blood– brain barrier by Ond is very low [28]. Accordingly to our experiments in vitro, in this study we demonstrated a significantly higher Dox level in all brain areas 2 h after pretreatment with Ond (cerebral hemispheres and brainstem P < 0.05; cerebellum P < 0.01). Since Ond is substrate of P-gp and increases Dox concentration in cells overexpressing P-gp these results suggest that Ond increases the Dox concentration inside the brain competing for P-gp mediated drug efflux at BBB level. This fact results in higher intratissutal concentration of the anticancer drug in cancer most likely reaching an efficacious concentration for inhibiting cancer cells growth. Anthracyclines may induce acute cardiotoxicity; in addition, high cumulative doses are associated with late-onset cardiomyopathy that is refractory to standard treatment [29]. Free radical generation and p53-dependent apoptosis by modulation of mTOR activity are supposed to contribute to Dox-induced cardiotoxicity [30,31]. On this basis, Ond could increase the toxicity of Dox on heart and kidney once given during cancer chemotherapy. So, we quantified Dox levels in heart and kidney of treated rats and we observed that the pre-treatment with Ond did not increase the levels of Dox in either tissue 1, 1.30 and 2 h after administration. We also evaluated plasma LDH activity and MDA levels, as markers of acute cardiotoxicity, and found no difference between rats treated with Dox alone and those also receiving Ond. These results are in agreement with the fact that cardiomyocytes do not express P-gp, while kidney has several transport proteins to efflux Dox into the urine. In conclusion, our data suggest that competition between Ond and Dox for P-gp mediated transport at BBB level increases significantly Dox penetration into the brain of the rat, without increasing cardiac or renal toxicity. These preliminary results on a rodent model will enable us to novel therapeutic approaches for refractory or
I. Sardi et al./Cancer Letters 353 (2014) 242–247
recurrent brain tumors where anticancer drugs are usually cleared by the BBB. Whether this phenomenon may have a therapeutic impact remains to be elucidated. Conflict of interest None. Acknowledgements This work was supported by: Associazione Italiana per la Ricerca sul Cancro (AIRC), grant IG–12799; “Amicodivalerio” Onlus; “Noi per Voi” Onlus, Fondazione Tommasino Bacciotti. References [1] B.A. Barres, The mystery and magic of glia: a perspective on their roles in health and disease, Neuron 60 (2008) 430–440. [2] E.A. Neuwelt, B. Bauer, C. Fahlke, G. Fricker, C. Iadecola, D. Janigro, et al., Engaging neuroscience to advance translational research in brain barrier biology, Nat. Rev. Neurosci. 12 (2011) 169–182. [3] M.S. Lesniak, U. Upadhyay, R. Goodwin, B. Tyler, H. Brem, Local delivery of doxorubicin for the treatment of malignant brain tumors in rats, Anticancer Res. 25 (2005) 3825–3831. [4] A.C. Stan, S. Casares, D. Radu, G.F. Walter, T.D. Brumeanu, Doxorubicin-induced cell death in highly invasive human gliomas, Anticancer Res. 19 (1999) 941–950. [5] W. Löscher, H. Potschka, Role of drug efflux transporters in the brain for drug disposition and treatment of brain diseases, Prog. Neurobiol. 76 (2005) 22–76. [6] S. Agarwal, A.M. Hartz, W.F. Elmquist, B. Bauer, Breast cancer resistance protein and P-glycoprotein in brain cancer: two gatekeepers team up, Curr. Pharm. Des. 17 (2011) 2793–2802. [7] P. Breedveld, J.H. Beijnen, J.H. Schellens, Use of P-glycoprotein and BCRP inhibitors to improve oral bioavailability and CNS penetration of anticancer drugs, Trends Pharmacol. Sci. 27 (2006) 17–24. [8] I. Sardi, G. la Marca, M.G. Giovannini, S. Malvagia, R. Guerrini, L. Genitori, et al., Detection of doxorubicin hydrochloride accumulation in the rat brain after morphine treatment by mass spectrometry, Cancer Chemother. Pharmacol. 67 (2011) 1333–1340. [9] I. Sardi, Morphine facilitates doxorubicin penetration in the central nervous system: a new prospect for therapy of brain tumors, J. Neurooncol. 104 (2011) 619–620. [10] R. Mazzanti, J.M. Croop, Z. Gatmaitan, M. Budding, K. Steiglitz, R. Arceci, I.M. Arias, Benzquinamide inhibits P-glycoprotein mediated drug efflux and potentiates anticancer agent cytotoxicity in multidrug resistant cells, Oncol. Res. 4 (1992) 359–365. [11] G.M. MacNab, J.J. Alexander, G. Lecatsas, E.M. Bey, J.M. Urbanowicz, Hepatitis B surface antigen produced by a human hepatoma cell line, Br. J. Cancer 34 (1976) 509–515. [12] O. Fantappiè, E. Masini, I. Sardi, L. Raimondi, D. Bani, M. Solazzo, et al., The MDR phenotype is associated with the expression of COX-2 and iNOS in a human hepatocellular carcinoma cell line, Hepatology 35 (2002) 843–852.
247
[13] O. Fantappiè, M. Solazzo, N. Lasagna, F. Platini, L. Tessitore, R. Mazzanti, P-glycoprotein mediates celecoxib-induced apoptosis in multiple drug-resistant cell lines, Cancer Res. 67 (2007) 4915–4923. [14] J.M. Croop, B.C. Guild, P. Gros, D.E. Housman, Genetics of multidrug resistance: relationship of a cloned gene to the complete multidrug resistant phenotype, Cancer Res. 47 (1987) 5982–5988. [15] N. Lasagna, O. Fantappiè, M. Solazzo, L. Morbidelli, S. Marchetti, G. Cipriani, et al., Hepatocyte growth factor and inducible nitric oxide synthase are involved in multidrug resistance-induced angiogenesis in hepatocellular carcinoma cell lines, Cancer Res. 66 (2006) 2673–2682. [16] M. Uchiyama, M. Mihara, Determination of malondialdehyde precursor in tissues by thiobarbituric acid test, Anal. Biochem. 86 (1978) 271–278. [17] I. Sardi, G. la Marca, S. Cardellicchio, L. Giunti, S. Malvagia, L. Genitori, et al., Pharmacological modulation of blood-brain barrier increases permeability of doxorubicin into the rat brain, Am. J. Cancer Res. 3 (4) (2013) 424–432. [18] R.S. Benjamin, C.E. Riggs Jr., N.R. Bachur, Pharmacokinetics and metabolism of adriamycin in man, Clin. Pharmacol. Ther. 14 (1973) 592–600. [19] T. Abe, S. Hasegawa, K. Taniguchi, A. Yokomizo, T. Kuwano, M. Ono, et al., Possible involvement of multidrug-resistance-associated protein (MRP) gene expression in spontaneous drug resistance to vincristine, etoposide and adriamycin in human glioma cells, Int. J. Cancer 58 (1994) 860–864. [20] J.E. Wolff, T. Trilling, G. Mölenkamp, R.M. Egeler, H. Jürgens, Chemosensitivity of glioma cells in vitro: a meta analysis, J. Cancer Res. Clin. Oncol. 125 (1999) 481–486. [21] S.C. Steiniger, J. Kreuter, A.S. Khalansky, I.N. Skidan, A.I. Bobruskin, Z.S. Smirnova, et al., Chemotherapy of glioblastoma in rats using doxorubicin-loaded nanoparticles, Int. J. Cancer 109 (2004) 759–767. [22] A. Eramo, L. Ricci-Vitiani, A. Zeuner, R. Pallini, F. Lotti, G. Sette, et al., Chemotherapy resistance of glioblastoma stem cells, Cell Death Differ. 13 (2006) 1238–1241. [23] T. Ohnishi, I. Tamai, K. Sakanaka, A. Sakata, T. Yamashima, J. Yamashita, et al., In vivo and in vitro evidence for ATP-dependency of P-glycoprotein-mediated efflux of doxorubicin at the blood-brain barrier, Biochem. Pharmacol. 49 (1995) 1541–1544. [24] S. Voulgaris, M. Partheni, M. Karamouzis, P. Dimopoulos, N. Papadakis, H.P. Kalofonos, Intratumoral doxorubicin in patients with malignant brain gliomas, Am. J. Clin. Oncol. 25 (2002) 60–64. [25] S. Hori, S. Ohtsuki, M. Tachikawa, N. Kimura, T. Kondo, M. Watanabe, et al., Functional expression of rat ABCG2 on the luminal side of brain capillaries and its enhancement by astrocyte-derived soluble factor(s), J. Neurochem. 90 (2004) 526–536. [26] N.J. Abbott, A.A. Patabendige, D.E. Dolman, S.R. Yusof, D.J. Begley, Structure and function of the blood-brain barrier, Neurobiol. Dis. 37 (2010) 13–25. [27] A.H. Schinkel, E. Wagenaar, C.A. Mol, L. van Deemter, P-glycoprotein in the blood-brain barrier of mice influences the brain penetration and pharmacological activity of many drugs, J. Clin. Invest. 97 (1996) 2517– 2524. [28] K.H. Simpson, F.M. Hicks, Clinical pharmacokinetics of ondansetron. A review, J. Pharm. Pharmacol. 48 (1996) 774–781. [29] P.K. Singal, N. Iliskovic, Doxorubicin-induced cardiomyopathy, N. Engl. J. Med. 339 (1998) 900–905. [30] S. Zhou, A. Starkov, M.K. Froberg, R.L. Leino, K.B. Wallace, Cumulative and irreversible cardiac mitochondrial dysfunction induced by doxorubicin, Cancer Res. 61 (2001) 771–777. [31] W. Zhu, M.H. Soonpaa, H. Chen, W. Shen, R.M. Payne, E.A. Liechty, et al., Acute doxorubicin cardiotoxicity is associated with p53-induced inhibition of the mammalian target of rapamycin pathway, Circulation 119 (2009) 99–106.