A New O-Alkylguanine-DNA Alkyltransferase Inhibitor Associated with a Nitrosourea (Cystemustine) Validates a Strategy of Melanoma-Targeted Therapy in Murine B16 and Human-Resistant M4Beu Melanoma Xenograft Models

Chemoresistance to O-alkylating agents is a major barrier to successful treatment of melanoma. It is mainly due to a DNA repair suicide protein, O-alkylguanine-DNA alkyltransferase (AGT). Although AGT inactivation is a powerful clinical strategy for restoring tumor chemosensitivity, it was limited by increased toxicity to nontumoral cells resulting from a lack of tumor selectivity. Achieving enhanced chemosensitization via AGT inhibition preferably in the tumor should protect normal tissue. To this end, we have developed a strategy to target AGT inhibitors. In this study, we tested a new potential melanoma-directed AGT inhibitor [2-amino-6-(4-iodobenzyloxy)-9-[4-(diethylamino) ethylcarbamoylbenzyl] purine; IBgBZ] designed as a conjugate of O-(4-iododbenzyl)guanine (IBg) as the AGT inactivator and a N,N-diethylaminoethylenebenzamido (BZ) moiety as the carrier to the malignant melanocytes. IBgBZ demonstrated AGT inactivation ability and potentiation of O-alkylating agents (cystemustine, a chloroethylnitrosourea) in M4Beu highly chemoresistant human melanoma cells both in vitro and in tumor models. The biodisposition study on mice bearing B16 melanoma, the standard model for the evaluation of melanomadirected agents, and the secondary ion mass spectrometry imaging confirmed the concentration of IBgBZ in the tumor and in particular in the intracytoplasmic melanosomes. These results validate the potential of IBgBZ as a new, more tumorselective, AGT inhibitor in a strategy of melanoma-targeted therapy. Malignant melanoma remains an increasingly prevalent cancer worldwide, despite emphasis on prevention and recent advances in early diagnosis (Miller and Mihnn, 2006). When this highly metastatic cancer is disseminated, chemotherapy remains the most efficient modality of treatment commonly using O-alkylating agents (nitrosoureas and tetrazines). However, only 15 to 20% of patients are responsive, principally because melanoma is highly chemoresistant (Tsutsumida et al., 2005). Melanoma chemoresistance to O-alkylating agents is mainly mediated by the DNA repair protein O-alkylguanine-DNA alkyltransferase (AGT or ATase), also termed Omethylguanine-DNA methyltransferase (Passagne et al., 2006). AGT repairs the O-alkylguanine lesions by a stoichiometric and irreversible transfer of the O-alkyl group to the cysteine 145 residue in the active site (Juillerat and Juillerat-Jeanneret, 2007). The protein is thereby totally inactivated, and restoration of AGT activity then requires de novo synthesis. On the basis of this suicide mechanism of action, several pseudosubstrates derived from 6-substituted guanines were developed to inactivate AGT (McElhinney et al., 2003). Although various highly specific AGT inhibitors are now available [O-benzylguanine (Bg) and O-(4-bromothenyl)guanine (BTg)], clinical trials in association with an Oalkylating agent [chloroethylnitrosourea (CENU) or tetrazine] showed increased side effects, requiring the reduction Article, publication date, and citation information can be found at http://jpet.aspetjournals.org. doi:10.1124/jpet.108.137737. ABBREVIATIONS: AGT, O-alkylguanine-DNA alkyltransferase; Bg, O-benzylguanine; BTg, O-(4-bromothenyl)guanine; CENU, chloroethylnitrosourea; IBgBZ, 2-amino-6-(4-iodobenzyloxy)-9-[4-(diethylamino) ethylcarbamoylbenzyl] purine; IBg, O-(4-iododbenzyl)guanine; BZ, N,Ndiethylaminoethylenebenzamido; HPLC, high-pressure liquid chromatography; cystemustine, N -[2-chloroethyl]-N-[2-(methylsulfonyl) ethyl]-N nitrosourea; DMSO, dimethyl sulfoxide; ID, injected dose; SIMS, secondary ion mass spectrometry. 0022-3565/08/3261-171–177$20.00 THE JOURNAL OF PHARMACOLOGY AND EXPERIMENTAL THERAPEUTICS Vol. 326, No. 1 Copyright © 2008 by The American Society for Pharmacology and Experimental Therapeutics 137737/3352828 JPET 326:171–177, 2008 Printed in U.S.A. 171 at A PE T Jornals on A ril 2, 2017 jpet.asjournals.org D ow nladed from of a therapeutic agent dose to suboptimal levels (Quinn et al., 2002; Ranson et al., 2006). This emerging clinical limitation led to the development of innovative strategies to reduce the dose-limiting toxicity of the treatments. Our approach is based on O-alkylating agent chemotherapy in combination with a melanoma-selective AGT inhibitor. Our aim was to sensitize tumor cells to the therapeutic O-alkylating agents while preventing toxic side effects by maintaining the DNA repair process (i.e., normal AGT levels) in healthy tissues. We had previously developed a series of AGT inhibitors derived from O-alkyl/aralkyl guanosine or 2 -deoxyguanosine that enhanced resistant malignant melanoma sensitivity to CENUs both in vitro and in animal models (Cussac et al., 1994a; Debiton et al., 1997; Mounetou et al., 1997; Buchdahl et al., 1998). Like others inhibitors, they were widely distributed in the organism and could deplete AGT in various normal tissues (Cussac et al., 1994b). To achieve melanoma-selectivity, we designed new AGT inhibitors [2-amino-6-(4-iodobenzyloxy)-9-[4-(diethylamino) ethylcarbamoylbenzyl] purine (IBgBZ)], conjugating the O(4-iododbenzyl)guanine (IBg) and the N,N-diethylaminoethylenebenzamido (BZ) moiety as a carrier to the malignant melanocytes. Indeed, we developed [I]benzamide derivatives as imaging agents, currently under Phase III clinical evaluation as radiopharmaceuticals for the diagnosis of disseminated melanoma (Michelot et al., 1993; Moreau et al., 1993). They form a reversible BZ-melanin complex, resulting in a high concentration in malignant melanocytes (Labarre et al., 2002; Guerquin-Kern et al., 2004). This property suggested that the BZ moiety would be a useful tool as a carrier to melanoma. The present study was conducted as a proof-of-concept for melanoma-targeted therapy and assessed, in vitro and in vivo, the AGT inhibition, the O-alkylating agent (CENU) antineoplastic activity enhancement, and the tumoral melanic tissue affinity of the new conjugate (IBgBZ) in murine and human-resistant melanoma. Materials and Methods Drugs. [I]BgBZ was labeled with high specific activity by radioiododestannylation in the presence of 37 MBq (1 mCi) of nocarrier-added sodium [I]iodide (GE Healthcare, Little Chalfont, Buckinghamshire, UK) and peracetic acid in aqueous ethanol adapted from a previously described procedure (Mounetou et al., 1995). [I]BgBZ was obtained in 85% radiochemical yield and 98% purity after high-pressure liquid chromatography (HPLC) purification. The reaction was initiated by nonradioactive sodium iodide, and IBgBZ was fully identified by proton and carbon nuclear magnetic resonance and mass spectrometry. The detailed syntheses will be described in a forthcoming article. Cystemustine was synthesized by previously described procedures (Madelmont, 1994) and supplied by Orphachem (Clermont-Ferrand, France). Cell Culture. The human melanoma M4Beu cells (obtained from Dr. J. F. Doré) (Jacubovich and Doré, 1979) and the B16-F0 melanoma (no. CRL-6322; American Type Culture Collection, Manassas, VA) were cultured in GlutaMAX (Fisher Scientific Bioblock, Illkirch, France) (Eagle’s minimum essential medium with glutamine) supplemented with 10% fetal calf serum and solutions of vitamins, sodium pyruvate, nonessential amino acids, and gentamicin at 37°C in a 5% CO2-humidified incubator. Determination of AGT Activity. The AGT assay was performed on protein extracts of M4Beu cells treated in vitro by IBgBZ, using a previously described procedure (Marchenay et al., 2001). Increasing amounts of M4Beu cell protein extracts were incubated with [H]Nmethyl-N-nitrosourea-methylated calf thymus DNA substrate for 1 h at 37°C. The substrate was then acid hydrolyzed, and the alkylated bases released in the sample were separated by reverse-phase HPLC (HP 1100 series; Hewlett Packard, Les Ulis, France) using a C18 Uptisphere column (5 m, 25 4.6 mm) eluted at a flow rate of 1.5 ml/min by a 20 mM NH4H2PO4 (pH 5.0) mobile phase containing 6% methanol (v/v) for 11 min and 15% methanol (v/v) for 14 min. The radioactivity of column effluents was automatically measured in a Packard flow scintillation analyzer (500TR series; PerkinElmer, Courtabeuf, France). AGT activity was determined by measuring the disappearance of O-[H]methylguanine from the substrate and was expressed in femtomoles of methyl groups transferred to protein per milligram of total protein present in the sample. To measure in vivo AGT activity after drug exposure in vivo, a group of mice with tumors that reached approximately 500 mm were treated with IBgBZ by i.p. injection. The animals were euthanized by CO2 inhalation at different times after injection. The tumors were removed and flash frozen in liquid nitrogen and stored at 80°C. Increasing amounts of M4Beu cell protein extracts were analyzed as described above. Cytotoxicity Tests. For colony-forming assays, M4Beu cells were plated in six-well multidishes (150 cells/dish) and allowed to adhere for 20 h before treatment. The cells were treated with a culture medium containing increasing concentrations of IBgBZ (1–25 M) dissolved in dimethyl sulfoxide (DMSO) before dilution in fresh culture medium, and they were incubated for various times. For combination treatment tests, cystemustine (50 M) diluted in fresh culture medium was added 4 h after IBgBZ (2.5 or 5 M) and incubated for 2 h. After the last treatment, the treated medium was replaced by fresh medium, and the cells were grown for 14 days at 37°C in a CO2 incubator. Dishes were rinsed with phosphate-buffered saline, and the cells were fixed with methanol and stained with 0.2% crystal violet solution. The plating efficiency of cells not treated with drugs was approximately 70%. Colonies ( 50 cells) were counted. The surviving fraction (percentage of survival) was calculated as the ratio of treated to untreated cells (control) 100. In Vivo Biodistribution Study. For tissue distribution analyses, 6-week male C57BL/6Jco mice (Charles River Laboratories, L’Arbresle, France) were inoculated with 5 10 melanoma B16 cells by dorsal s.c. injection. When the tumor diameter reached approximately 0.5 cm, the animals received an i.v. injection of IBgBZ (0.74 MBq, 20 Ci, 10 mg/kg) dissolved in DMSO and 0.9% sodium chloride. Mice were euthanized by CO2 inhalation at 1 h, 3 h, 5 h 30 min, 6 h, and 48 h after drug administration and were immediately frozen by immersion in liquid nitrogen. They were then embedded in a 2% carboxymethylcellulose gel, and the block was sagittally sectioned at 22°C with a Reichert-Jung Cryopolycut cryomicrotome (Reichert Jung, Heidelberg, Germany). The selected 40m-thick slices were taken using no. 810 Scotch band tape (3M, St. Paul, MN), dried for 48 h at 22°C, fixed on a hard-bound sheet, and introduced into an Ambis 4000 detector (Scanalytics/CSPI, San Diego, CA) to quantify the radioactivity in selected organs (n 6 for each time studied) (Labarre et al., 1999). The final results are expressed as percentages of injected dose (ID) per gram of tissue. Excretion Study. The excretion study was conducted using 6-week-old male C57BL/6Jco mice given [I]BgBZ as described above and housed in metabolic cages (Charles River Laboratories), allowing separate collection of feces and urine at 24, 48, and 72 h after administration. Radioactivity was measured in a gamma counter (Minaxi 5530; Packard, Rungis, France). In Vivo Stability. Plasma, urine, rehydrated feces, and organs homogenized in a Potter blender (Fisher Scientific Bioblock) were sampled, extracted three times with methanol, and centrifuged at 2000g for 10 min at 4°C. The recovery of radioactivity ranged from 80 to 85%. After evaporation, the dry residues were dissolved in the mobile phase. Chromatography (HPLC) analysis was performed on 172 Rapp et al. at A PE T Jornals on A ril 2, 2017 jpet.asjournals.org D ow nladed from an HP 1100 system with a UV/visible detector (Hewlett Packard) equipped with a reverse-phase Kromasil column (C18, 5 m, 250 4.6 mm; Hewlett Packard) and connected to a 500TR flow scintillation analyzer (Packard Instrument S.A., Rungis, France). The flow rate was 1 ml/min with a gradient mobile phase starting from a methanol/0.4% NH4OH mixture:70/30 (v/v) to 100/0 at 10 min to 15 min. Secondary Ion Mass Spectrometry Analysis. Secondary ion mass spectrometry (SIMS) imaging was performed using a NanoSIMS-50 Ion Microprobe (CAMECA, Gennevilliers, France). B16 murine melanoma cells were intravenously injected into male C57BL/ 6Jco mice to obtain, within 12 days, tumor cell colonies in lungs mimicking pulmonary micrometastases. After IBgBZ administration, mice were euthanized at 1 h, 3 h, and 5 h 30 min. Typical sample preparation for SIMS analysis was described previously (Guerquin-Kern JL et al., 2005). Image processing was performed using ImageJ (W. S. Rasband, United States National Institutes of Health, Bethesda, MD, http://rsb.info.nih.gov/ij/), a public-domain Java image-processing program, to obtain proper colocalization of the observed structures on the processed maps for all of the ion species, and allowed further correlation with light microscope im-