Mechanisms of Uptake and Resistance to Troxacitabine, a Novel Deoxycytidine Nucleoside Analogue, in Human Leukemic and Solid Tumor Cell Lines

[CANCER RESEARCH 61, 7217–7224, October 1, 2001] Mechanisms of Uptake and Resistance to Troxacitabine, a Novel Deoxycytidine Nucleoside Analogue, in Human Leukemic and Solid Tumor Cell Lines

Henriette Gourdeau,1,2 Marilyn L. Clarke,1 France Ouellet, Delores Mowles, Milada Selner, Annie Richard, Nola Lee, John R. Mackey, James D. Young,3 Jacques Jolivet, Ronald G. Lafrenie`re, and Carol E. Cass4 Shire BioChem Inc., Laval, Que´bec, H7V 4A7 Canada [H. G., F. O., A. R., N. L., J. J., R. G. L.]; Departments of Oncology [M. L. C., J. R. M., C. E. C.] and Physiology [J. D. Y.], University of Alberta, Alberta T6G 1Z2 Canada; and Cross Cancer Institute, Edmonton, Alberta T6G 1Z2 Canada [M. L. C., D. M., M. S., J. R. M., C. E. C.]

ABSTRACT oxycytidine analogues such as gemcitabine (2Ј,2Ј-difluorodeoxycytidine; dFdC) and cytarabine (1-␤-D-arabinofuranosylcytosine; araC), -Troxacitabine (Troxatyl; BCH-4556; (-)-2؅-deoxy-3؅-oxacytidine), a de which are in the ␤-D configuration, troxacitabine has a nonnatural ␤-L oxycytidine analogue with an unusual dioxolane structure and nonnatural configuration. Troxacitabine, which shares the same intracellular ac- L-configuration, has potent antitumor activity in animal models and is in clinical trials against human malignancies. The current work was under- tivation pathway as gemcitabine and cytarabine, undergoes a series of taken to identify potential biochemical mechanisms of resistance to trox- phosphorylation reactions through a first rate-limiting step catalyzed 5 acitabine and to determine whether there are differences in resistance by dCK (EC 2.7.1.74) to form the active triphosphate nucleotide. mechanisms between troxacitabine, gemcitabine, and cytarabine in hu- Troxacitabine triphosphate is incorporated into DNA (2), which is man leukemic and solid tumor cell lines. The CCRF-CEM leukemia cell believed to be the main mechanism of cytotoxicity of deoxycytidine line was highly sensitive to the antiproliferative effects of troxacitabine, analogues, although there are differences in their DNA incorporation gemcitabine, and cytarabine with inhibition of proliferation by 50% ob- patterns. The incorporation of troxacitabine causes immediate chain served at 160, 20, and 10 nM, respectively, whereas a deoxycytidine kinase termination (6), whereas gemcitabine incorporation allows the addi- dCK)-deficient variant (CEM/dCK؊) was resistant to all three drugs. In) contrast, a nucleoside transport-deficient variant (CEM/ARAC8C) exhib- tion of a single deoxynucleotide (7, 8), and cytarabine incorporation ited high levels of resistance to cytarabine (1150-fold) and gemcitabine allows the addition of multiple deoxynucleotides before chain termi- (432-fold) but only minimal resistance to troxacitabine (7-fold). Analysis nation (9, 10). Troxacitabine also differs from gemcitabine and cyt- of troxacitabine transportability by the five molecularly characterized arabine in that it is resistant to CDA (EC 3.5.4.5; Ref. 2). Although human nucleoside transporters [human equilibrative nucleoside trans- cellular uptake of troxacitabine has been reported to occur via the porters 1 and 2, human concentrative nucleoside transporter (hCNT) 1, equilibrative nucleoside transport processes (11), it is not known hCNT2, and hCNT3] revealed that short- and long-term uptake of 10–30 whether functional plasma membrane nucleoside transporters are re- ␮ 3 M [ H]troxacitabine was low and unaffected by the presence of either quired for troxacitabine cytotoxicity, as has been reported for gem- nucleoside transport inhibitors or high concentrations of nonradioactive citabine (12, 13) and cytarabine (14, 15). troxacitabine. These results, which suggested that the major route of cellular uptake of troxacitabine was passive diffusion, demonstrated that Understanding of nucleoside transport processes of human cells has deficiencies in nucleoside transport were unlikely to impart resistance to advanced considerably during recent years (reviewed in Refs. 16–18). troxacitabine. A troxacitabine-resistant prostate cancer subline (DU145R; Seven distinct nucleoside transport processes have been demonstrated 6300-fold) that exhibited reduced uptake of troxacitabine was cross- in human cells on the basis of permeant selectivities, inhibition by resistant to both gemcitabine (350-fold) and cytarabine (300-fold). dCK diagnostic agents, and mechanisms of transport. The proteins respon- activity toward deoxycytidine in DU145R cell lysates was <20% of that in sible for the five major nucleoside transport processes have been DU145 cell lysates, and no activity was detected toward troxacitabine. identified by molecular cloning. The human transporter proteins com- Sequence analysis of cDNAs encoding dCK revealed a mutation of a prise two functionally and molecularly distinct groups: (a) the hENTs; highly conserved amino acid (Trp923Leu) in DU145R dCK, providing a and (b) the sodium-dependent hCNTs. The proteins and their func- possible explanation for the reduced phosphorylation of troxacitabine in DU145R lysates. Reduced deamination of deoxycytidine was also observed tional activities are: (a) hENT1, broadly selective and mediates es in DU145R relative to DU145 cells, and this may have contributed to the (equilibrative NBMPR-sensitive) activity; (b) hENT2, broadly selec- overall resistance phenotype. These results, which demonstrated a differ- tive and mediates ei (equilibrative NBMPR-insensitive) activity; (c) ent resistance profile for troxacitabine, gemcitabine, and cytarabine, sug- hCNT1, pyrimidine nucleoside selective and mediates cit (concentra- gest that troxacitabine may have an advantage over gemcitabine and tive, insensitive to NBMPR and thymidine-selective activity); (d) cytarabine in human malignancies that lack or have low nucleoside trans- hCNT2, purine nucleoside and uridine selective and mediates cif port activities. (concentrative, insensitive to NBMPR and formycin B-selective) activity; and (e) hCNT3, broadly selective and mediates cib (concentra- INTRODUCTION tive, insensitive to NBMPR and broadly selective) activity. The proteins responsible for csg (concentrative, sensitive to NBMPR and Ј Ј Troxacitabine (Troxatyl; BCH-4556; (-)-2 -deoxy-3 -oxacytidine) guanosine-selective)- and cs (concentrative, sensitive to NBMPR)- has potent antitumor activity against human leukemic (1) and solid mediated transport activities, which have only been reported in a few tumor (2–5) xenograft animal models and is in clinical trials against cell types, have not been identified. human malignancies. Unlike naturally occurring nucleosides and de- The current work was undertaken to identify potential biochemical mechanisms of resistance to troxacitabine. The antiproliferative ef- Received 5/4/01; accepted 7/26/01. fects of troxacitabine against human cell lines that either possessed or The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with lacked the capacity for nucleoside transport or phosphorylation of 18 U.S.C. Section 1734 solely to indicate this fact. deoxycytidine were compared to determine whether the loss of either 1 H. G. and M. L. C. contributed equally to this work. 2 To whom requests for reprints should be addressed, at Shire BioChem Inc., 275 Boulevard Armand-Frappier, Laval, Que´bec H7V 4A7, Canada. E-mail: hgourdeau@ 5 The abbreviations used are: dCK, deoxycytidine kinase; CDA, cytidine deaminase; ca.shire.com. hCNT, human concentrative nucleoside transporter; hENT, human equilibrative nucleo- 3 J. D. Y. is a Heritage Scientist of the Alberta Heritage Foundation for Medical side transporter; NBMPR, nitrobenzylmercaptopurine ribonucleoside; THU, 3,4,5,6- Research. tetrahydrouridine; DMDC, 2Ј-deoxy-2Ј-methylidenecytidine; HPLC, high-performance 4 C. E. C. is Canada Research Chair in Oncology. liquid chromatography. 7217

Downloaded from cancerres.aacrjournals.org on September 25, 2021. © 2001 American Association for Cancer Research. UPTAKE AND RESISTANCE TO TROXACITABINE process gave rise to resistance to troxacitabine. The transportability of Resistance to troxacitabine was induced by exposing DU145 cells stepwise troxacitabine was then examined in human cell lines that either lacked to 2-fold increments of troxacitabine (0.01–10 ␮M) that were increased only nucleoside transport activity altogether or exhibited a single nucleo- when the proliferation rates of drug-treated cultures were similar to those of side transporter activity by measuring the rates of uptake of [3H]trox- untreated cultures. After 8 months of continuous exposures, the resistant R ␮ acitabine in the presence and absence of transport inhibitors or high variant (designated DU145 ) was maintained in 2 M troxacitabine. concentrations of nonradioactive troxacitabine. These studies indi- Chemosensitivity Testing. The relative cytotoxicities of troxacitabine, gemcitabine, and cytarabine against CCRF-CEM, CEM/ARAC8C, and CEM/ cated that troxacitabine, a poor permeant of nucleoside transporters, dCK cells were assessed using the CellTiter 96 proliferation assay. This assay enters cells primarily by passive diffusion, suggesting that a lack of is based on the reduction of a tetrazolium compound to a soluble formazan nucleoside transport capability was unlikely to be involved in the derivative by the dehydrogenase enzymes of metabolically active cells. The development of resistance. This was confirmed in a troxacitabine- absorbance (490 nm) is directly proportional to the number of living cells in R resistant prostate cancer cell line (DU145 ) that was evaluated for culture. Cells were added to 96-well tissue culture plates (105 cells/well; 8 cross-resistance to gemcitabine and cytarabine and for changes in replicates/condition) and exposed to graded concentrations (1–105 nM)of cellular uptake of troxacitabine. Reduced quantities of troxacitabine cytarabine, gemcitabine, or troxacitabine, after which the numbers of living metabolites were observed in the resistant cells that could not be cells remaining were determined according to the manufacturer’s instructions. explained by changes in troxacitabine permeation, and dCK and CDA Antiproliferative activity of drugs against DU145 and DU145R cells was 5 activities were therefore compared in the parental and resistant cells. determined by seeding cells in 12-well tissue culture plates (10 cells/well; 3 Because the resistant cells also exhibited reduced CDA activity (in replicates/condition) and initiating drug exposures 24 h later by a complete addition to reduced dCK activity), the impact of inhibition of CDA on change of medium. Cells were enumerated by trypsinization and electronic particle counting (Coulter Electronics Inc., Luton, United Kingdom) after troxacitabine toxicity was assessed by evaluating the sensitivity of exposure to drugs for 48 h. Chemosensitivity was expressed as the drug DU145 cells to troxacitabine in the presence and absence of THU, an concentration that inhibited cell proliferation by 50% (IC50 values) and deter- inhibitor of CDA (19). mined from concentration-effect relationships using GraphPad Prism 2.01 The results indicated that troxacitabine, a novel deoxycytidine (GraphPad Software, San Diego, CA). analogue, has a different uptake and metabolism profile in cultured Transient Expression of hCNT1 and hCNT2 in HeLa Cells. The cDNAs human leukemic and prostate cancer cell lines than either gemcitabine encoding the hCNT1 and hCNT2 proteins (GenBank accession number or cytarabine. Troxacitabine may thus have an advantage for treatment U62966 and AF036109, respectively) were subcloned from plasmids pMHK2 of human malignancies that lack or have low nucleoside transport (23) and pMH15 (24) into the mammalian expression vector, pcDNA3, to activities and may also be useful in human leukemias or solid tumors produce pcDNA3-hCNT1 (25) and pcDNA3-hCNT2.6 The plasmids were refractory to cytarabine or gemcitabine, respectively. transfected separately into actively proliferating HeLa cells as described previously (12, 25, 26). Cellular Uptake Assays. All assays were conducted at room temperature. MATERIALS AND METHODS For suspension cells, uptake assays were conducted in microfuge tubes (106 cells/tube) as described previously (27, 28). For adherent cells, assays were Materials. Troxacitabine was synthesized at Shire BioChem Inc. (20), and conducted in either sodium-containing transport buffer [20 mM Tris-HCl, 3 mM [3H]troxacitabine (3.9 Ci/mmol) was prepared by Moravek Biochemicals, Inc. ⅐ K2HPO4,1mM MgCl2 6H2O,2mM CaCl2,5mM glucose, and 130 mM NaCl (Brea, CA) from material provided by Shire Biochem Inc. Cytarabine and (pH 7.4); 300 Ϯ 15 mOsm] or a sodium-free transport buffer in which NaCl gemcitabine were gifts, respectively, from Bristol-Myers Squibb (Montre´al, was replaced by N-methyl-D-glucamine as described previously (25, 26). Canada) and Eli Lilly (Toronto, Canada). Uridine, NBMPR, dipyridamol, Intracellular Metabolism of [3H]Troxacitabine and [3H]Deoxycytidine. dilazep, tubercidin, and phorbol 12-myristate 13-acetate were purchased from Subconfluent cultures of DU145 and DU145R cells were exposed to [3H]trox- Sigma Chemical Co.-Aldrich Canada Ltd. (Oakville, Canada); THU was from acitabine or [3H]deoxycytidine for 4 and 24 h. Cells were then harvested by 3 Calbiochem (San Diego, CA); [methyl- H]thymidine (2 Ci/mmol) and trypsinization and recovered by centrifugation, and the resulting cell pellets [5-3H]deoxycytidine (18.4 Ci/mmol) were from Amersham Canada (Oakville, 3 were resuspended in 60% ice-cold methanol. After overnight incubation at Canada); [5- H]uridine (40 Ci/mmol) was from Moravek Biochemicals, Inc.; Ϫ20°C, the methanol extracts were centrifuged, and the pellets were used for [14C(U)]sucrose (0.6 Ci/mmol) was from New England Nuclear Life Science determination of protein levels with the Bradford protein assay (Bio-Rad Products, Inc. (Boston, MA); the CellTiter 96 proliferation kit was from Laboratories, Mississauga, Canada). The supernatants were collected, evapo- Promega Corp. (Madison, WI); the pcDNA3 mammalian expression vector rated to dryness, and resuspended in 200 ␮lofHO for HPLC analysis. was from Invitrogen (Carlsbad, CA); and DEAE-dextran was from Pharmacia 2 Samples were analyzed with a C reversed-phase column (YMC ODS-A: 5 Biotech Inc. (Baie d’Urfe´, Canada). All other reagents used were of analytical 18 ␮m, 120 A, and 250 ϫ 4.6 mm inside diameter; Waters Corp., Milford, MA) grade and were obtained from commercial sources. and a LC module 1 (486 detector and injector 715) connected to UV (254 nm) Cell Culture. The human CCRF-CEM leukemia, DU145 prostate cancer, and radioactivity monitors (Canberra Packard Canada Ltd., Montre´al, Canada). HeLa cervical carcinoma, and HL-60 promyelocytic leukemia cell lines were A linear gradient elution was started with CH CN (buffer A) to reach 10% of purchased from the American Type Culture Collection (Manassas, VA). CEM/ 3 buffer B [0.05 M NaPO (pH 6.7) containing 5 mM tetrabutyl ammonium ARAC8C, a nucleoside transport-deficient derivative of CCRF-CEM, was a 4 dihydrogen phosphate] in 25 min at a flow rate of 1 ml/min. Elution was gift from Dr. B. Ullman (14) and was routinely cultured with 0.5 ␮M cytarabine Ϫ continued for 15 min. Analysis of standards (15 nmol each of troxacitabine, and 0.25 ␮M tubercidin to maintain the mutant phenotype. CEM/dCK ,a troxacitabine monophosphate, troxacitabine diphosphate, and troxacitabine dCK-deficient derivative of CCRF-CEM, was a gift from Dr. A. Fridland (21). Ϫ triphosphate) gave retention times of 6.1, 10.4, 21.1, and 39.9 min, respec- Cells were grown in RPMI 1640 (CCRF-CEM, CEM/ARAC8C, CEM/dCK , tively. Peak areas were quantified using Millennium software (Waters Corp.). HeLa, and HL-60) or Eagle’s MEM with 0.1 mM nonessential amino acids (DU145 and DU145R) supplemented with 10% fetal bovine serum (Life ATP, ADP, and AMP levels were also determined to monitor the quality of the extraction procedure (29). Technologies, Inc., Burlington, Canada). Stock cell lines, which were demon- R strated to be free of Mycoplasma, were maintained as suspension (CCRF- dCK and CDA Assays. DU145 and DU145 cells were harvested from CEM, CEM/ARAC8C, CEM/dCKϪ, and HL-60) or adherent (DU145, actively proliferating cultures by trypsinization and centrifugation, and cell Յ Ϫ DU145R, and HeLa) cultures in the absence of antibiotics and incubated at pellets were stored ( 2 months) at 70°C until use. For analysis, the pellets ϫ 7 37°C in a humidified atmosphere (5% CO ). were thawed on ice, mixed (2.5 10 cells/ml) with 0.3 M Tris-HCl (pH 8.0) 2 ␮ ␤ Exponentially growing HL-60 cells (3 ϫ 106 cells) were induced to differ- containing 50 M -mercaptoethanol, sonicated, and centrifuged. The resulting entiate by plating in 100-mm Falcon Primaria tissue culture dishes (Becton extracts were used for enzyme assays (conducted at 37°C) and for determina- Dickinson, Missassuaga, Canada) in the presence of phorbol myristate acetate (200 ng/ml) as described previously (22). 6 T. Lang, J. D. Young, and C. E. Cass, unpublished results. 7218

Ϫ Table 1 A comparison of chemosensitivities of CCRF-CEM, CEM/dCK , and CEM/ respectively (Table 1). These results indicated that troxacitabine tox- ARAC8C cells to troxacitabine, gemcitabine, and cytarabine Ϫ icity was less affected than that of gemcitabine or cytarabine by the Suspension cultures (104 cells/well) of CEM, CEM/ARAC8C, and CEM/dCK cells 5 were exposed to graded concentrations (1–10 nM) of troxacitabine, gemcitabine, or absence of a functional nucleoside transporter. cytarabine for 48 h. Chemosensitivity was measured using the Promega CellTiter 96 cell Uptake of [3H]Troxacitabine and [3H]Uridine by CCRF-CEM proliferation assay and is expressed as the IC50 value determined from concentration- effect relationships as described in “Materials and Methods.” Each value represents the and CEM/ARAC8C Cells. The contribution of permeation to cellu- mean Ϯ SD of three separate experiments (8 replicates for each concentration/experi- lar uptake of troxacitabine was examined by comparing time courses ment). of uptake of [3H]troxacitabine by CCRF-CEM and CEM/ARAC8C Troxacitabine Gemcitabine Cytarabine cells (Fig. 1). Parallel measurements were also conducted with [3H] Cell line IC (nM) IC (nM) IC (nM) 50 50 50 uridine, a physiological substrate of hENT1. CCRF-CEM cells exhib- CCRF-CEM 160 Ϯ 12 20 Ϯ 0.4 10 Ϯ 2 Ϫ ␮ 3 CEM/dCK Ͼ100,000 Ͼ50,000 Ͼ50,000 ited a large difference in capacity for uptake of 30 M [ H]troxacit- CEM/ARAC8C 1,180 Ϯ 320 8,630 Ϯ 880 11,500 Ϯ 2,700 abine and [3H]uridine, with initial rates of uptake (2–10 s) of 0.073 and 0.585 pmol/106 cells/s, respectively. Although uptake (2–60 s) of [3H]uridine by CCRF-CEM cells was reduced by Ͼ99% to 0.002 and 6 tion of protein content by the Bio-Rad Bradford protein assay. dCK activity 0.005 pmol/10 cells/s by the presence of either 5 mM nonradioactive 3 was determined as described elsewhere (30, 31). Experiments were performed uridine or 100 nM NBMPR, respectively, uptake of [ H]troxacitabine in the presence and absence of 1 mM thymidine to assess the mitochondrial by CCRF-CEM cells was unaffected by 5 mM nonradioactive troxac- thymidine kinase contribution to dCK activity, and in some experiments, itabine or 100 nM NBMPR. Extended time courses of uptake of 3 3 [ H]troxacitabine was used as a substrate instead of [ H]deoxycytidine. For [3H]troxacitabine and [3H]uridine were also determined in CEM/ CDA activity, 100-␮l portions of extract were incubated for up to 60 min in the ARAC8C cells. In these cells, the uptake of troxacitabine was com- presence of 20 ␮lof5mM deoxycytidine and 80 ␮l of CDA buffer [0.1 M Tris-HCl (pH 8.0) containing 50 ␮M ␤-mercaptoethanol]. The deoxycytidine parable with that of uridine, with initial rates (2–60 s) of 0.057 and 6 was separated from the product (uridine) by HPLC analysis. The mobile phases 0.030 pmol/10 cells/s, respectively. Because the initial uptake rates ␮ used were 0.1% trifluoroacetic acid (pH 3.27; buffer A) and 0.1% CH3CN in (2–10 s) observed for 30 M troxacitabine were so low, uptake at buffer A (buffer B). A linear gradient elution was used (2% A320% B, 30 higher concentrations (40, 50, and 60 ␮M) was also examined in min, 1 ml/min). Retention times (UV detection at 270 nm) for deoxycytidine CCRF-CEM cells (data not shown). The total amounts of troxacitab- and uridine were 7.2 and 8.6 min, respectively. ine accumulated for all four concentrations at 60 s increased only Sequence Analysis of DU145R dCK Gene. Actively proliferating DU145 slightly, from 12.8 to 14.4 pmol/106 cells. These results were con- and DU145R cells were harvested by trypsinization and centrifugation as described above, and cytoplasmic RNA was isolated and incubated with sistent with uptake by passive diffusion and/or a low-affinity transport reverse transcriptase to yield cDNA using the GeneAmp RNA PCR kit (Roche process. Molecular Systems Inc., Branchburg, NJ). A 701-bp fragment corresponding to 89% of the dCK open reading frame (residues 218–917; GenBank accession number M60527) was amplified using sense (5Ј-CGCATCAAGAAAATCTC- CCAT-3Ј) and antisense (5Ј-ACCTTTTCAACCAGACTTTC-3Ј) primers. PCR products from two independent reverse transcription-PCR reactions for both DU145 and DU145R cells were cloned into pCR2.1-TOPO (Invitrogen). Plasmid DNA was prepared using the Qiagen-tip 100 kit (Qiagen, Mississauga, Canada) and sequenced from M13R and T7 primers using dye primer cycle sequencing (Bio S&T Inc., Lachine, Canada). Sequences were aligned and analyzed using Lasergene software (DNASTAR), and homology searches were conducted using the BLAST server from the National Center for Biotechnol- ogy Information (Bethesda, MD).

RESULTS Antiproliferative Activity of Troxacitabine, Gemcitabine, and Cytarabine against CCRF-CEM Cells and Drug-resistant Variants That Lack dCK or Nucleoside Transport Activity. The toxicity of troxacitabine against CCRF-CEM and CEM/dCKϪ cells was compared with that of gemcitabine and cytarabine by assessing their relative abilities to inhibit proliferation during continuous 48-h exposures (Table 1). Proliferation of CCRF-CEM cells was potently

inhibited, with IC50 values of 160, 20, and 10 nM, respectively, for troxacitabine, gemcitabine, and cytarabine. In contrast, the IC50 values for CEM/dCKϪ cells were substantially greater than the values observed for CCRF-CEM cells (Table 1). These results confirmed that Fig. 1. A comparison of uptake of troxacitabine and uridine by nucleoside transport- impaired dCK activity greatly decreased cellular sensitivity to trox- competent CCRF-CEM and nucleoside transport-deficient CEM/ARAC8C cells. Uptake 3 acitabine, as reported previously (2), and were consistent with the of 30 ␮M [ H]nucleoside was determined at the time intervals shown as described in “Materials and Methods” in the absence or presence of either 5 mM nonradioactive three drugs sharing the same initial activation pathway whereby nucleoside or 100 nM NBMPR. Each data point represents the mean Ϯ SD of three phosphorylation is catalyzed by dCK. determinations; error bars are not shown where values were small and therefore obscured by data points. A (CCRF-CEM cells): [3H]troxacitabine alone, ࡗ;[3H]uridine alone, f; CEM/ARAC8C cells, which exhibit high-level resistance to cytara- 3 and [ H]uridine plus 5 mM nonradioactive uridine () or 100 nM NBMPR (Œ). The values bine because of the genetic loss of hENT1-mediated transport, are for [3H]troxacitabine in the presence of nonradioactive troxacitabine or NBMPR were the cross-resistant to other nucleoside analogues that are permeants for same as those obtained for troxacitabine alone and therefore are not shown. B (CEM/ ARAC8C cells): [3H]troxacitabine alone, ࡗ; and [3H]uridine alone, f. The values in the this transporter (32). When exposed to troxacitabine, gemcitabine, and presence of nonradioactive nucleoside were the same as those obtained for [3H]nucleoside cytarabine, the levels of resistance differed by 7-, 431-, and 1150-fold, alone and therefore are not shown. 7219

DU145 cells from 0.35 to 0.02 pmol/106 cells/s, indicating that Ͼ95% of uridine uptake was mediated. In separate experiments (Fig. 2A), the 3 initial rate of uptake (2–60 s) of 10 ␮M [ H]uridine was reduced by 6 57% (from 0.60 to 0.26 pmol/10 cells/s) in the presence of 100 nM NBMPR and by Ͼ99% (from 0.60 to 0.01 pmol/106 cells/s) in the presence of 100 ␮M dilazep, indicating that the major routes of uridine entry were via hENT1- and hENT2-mediated transport processes, respectively. The concentrative nucleoside transporters (hCNT1, hCNT2, and hCNT3), which are sodium/nucleoside symporters, can be detected by comparing initial rates of uptake in sodium-containing assay mixtures with those in which sodium has been replaced with N-methyl-D- glucamine; the latter manipulation eliminates the extracellular sodium gradient required for functionality of the concentrative nucleoside transporters. Rates of uptake of [3H]uridine by DU145 cells were similar in medium with or without sodium (Fig. 2A), indicating that DU145 cells did not possess sodium-dependent nucleoside transport activity. Uptake of [3H]troxacitabine by DU145 cells was examined in the presence and absence of NBMPR at a concentration that is known to block hENT1-mediated but not hENT2-mediated transport processes (Fig. 2B). Uptake was barely detectable, even after 4-h exposures to [3H]troxacitabine in the presence or absence of NBMPR. These results, which indicated that hENT1-mediated transport in DU145 cells Fig. 2. The uptake characteristics of DU145 cells. Uptake was determined in the presence (closed symbols) or absence (open symbols) of sodium for the time intervals was not responsible for troxacitabine permeation, were consistent with shown as described in “Materials and Methods.” Each data point represents the the low and primarily nonmediated uptake of troxacitabine observed mean Ϯ SD of three determinations; error bars are not shown where values were small and 3 in CCRF-CEM cells. High concentrations of nonradioactive troxacit- therefore obscured by data points. A, 10 ␮M [ H]uridine alone (f and Ⅺ)orinthe 3 presence of 100 nM NBMPR (Œ and ‚)or100␮M dilazep (). Uridine uptake in abine and dilazep also had no effect on uptake of [ H]troxacitabine by dilazep-treated cells was identical in the presence () and absence of sodium (data not 3 DU145 cells (data not shown). shown). B, 10 ␮M [ H]troxacitabine alone (f) or in the presence of 100 nM NBMPR (Œ). The apparent absence of mediated permeation of troxacitabine in DU145 cells was confirmed in HeLa cells (Fig. 3), which possess high The Influence of Enantiomeric Configuration of Troxacitabine levels of both hENT1- and hENT2-mediated transport activities (34). on Its Ability to Inhibit Uridine Transport in CCRF-CEM Cells. Uptake of troxacitabine was low and unaffected by the addition of a Equilibrative transporters are stereoselective (33), which could ex- 100-fold excess of nonradioactive troxacitabine and reduced only plain the lack of recognition of the L-enantiomer of troxacitabine by hENT1. The influence of enantiomeric configuration on troxacitabine transportability was examined in CCRF-CEM cells by comparing the effects of 10 mMD-orL-troxacitabine on initial uptake rates of 10 ␮M 3 [ H]uridine (the natural D-enantiomer). The addition of 10 mM nonradioactive D-uridine to assay mixtures reduced transport of 10 ␮M [3H]uridine (2–8 s) from 1.02 to 0.09 pmol/106 cells/s. In contrast, the presence of D-orL-troxacitabine produced only partial inhibitions, reducing initial D-uridine transport rates from 1.02 to 0.56 (D-troxac- 6 itabine) and 0.64 (L-troxacitabine) pmol/10 cells/s. These results, which demonstrated that D- and L-troxacitabine enantiomers were not capable of effectively competing with D-uridine for interaction with hENT1 when present at 1000-fold greater concentrations than uridine, suggested that troxacitabine was poorly recognized by hENT1 because of its dioxolane structure rather than its enantiomeric configuration. Uptake of [3H]Troxacitabine and [3H]Uridine in DU145 Cells. Uptake of troxacitabine by DU145 cells was reported previously to be inhibited by NBMPR and dipyridamol, suggesting entry into DU145 cells via both hENT1- and hENT2-mediated transport processes (11). Because the nucleoside transport characteristics of DU145 cells have not been defined previously, the transportability of uridine, which is a permeant of both equilibrative and concentrative nucleoside transporters (16), was examined in experiments that assessed the effects of (a) Fig. 3. A comparison of uptake of troxacitabine and uridine by the hENT2-mediated excess nonradioactive uridine, (b) inhibitors of equilibrative nucleo- process of HeLa cells. Uptake was determined for the time intervals shown as described in “Materials and Methods.” Each data point represents the mean Ϯ SD of three side transport processes (NBMPR and dilazep), and (c) replacement of determinations; error bars are not shown where values were small and therefore obscured ␮ 3 f sodium with N-methyl-D-glucamine. by data points. A, uptake of 10 M [ H]troxacitabine in either the absence ( ) or presence of 100 nM NBMPR (Œ), 100 ␮M dilazep (), or 1 mM nonradioactive troxacitabine (ࡗ). M 3 3 Addition of nonradioactive uridine (1 m ) to uptake assay mixtures B, uptake of 10 ␮M [ H]troxacitabine (f]) or D-[ H]uridine (Ⅺ) in the presence of 100 nM 3 reduced the initial rate of uptake (2–60 s) of 5 ␮M [ H]uridine by NBMPR. 7220

Fig. 4. A comparison of uptake of troxacitabine and uridine by recombinant hCNT1 and hCNT2 produced in HeLa cells. HeLa cells were transiently transfected with pcDNA3 without an insert () or with an insert containing either the coding sequence for hCNT1 (A and B) or hCNT2 (C and D), and uptake 3 3 of 10 ␮M [ H]troxacitabine (B and D)or[H]uridine (A and C) was determined in the presence of 100 ␮M dilazep for the time intervals shown as described in “Materials and Methods.” Up- take values for recombinant hCNT1 and hCNT2 were obtained in the presence (f) or absence (Œ) of sodium. Each data point represents the mean Ϯ SD of three determinations; error bars are not shown where values were small and therefore obscured by data points.

slightly by NBMPR or dilazep (Fig. 3A). The extended time courses abilities to inhibit proliferation during continuous 48-h exposures. for uridine uptake shown in Fig. 3B, which were obtained in cells Proliferation of DU145 cells was inhibited, with IC50 values of 10, 20, treated with 100 nM NBMPR to block hENT1, revealed a large and 100 nM, respectively, for troxacitabine, gemcitabine and cytara- difference in uptake of uridine and troxacitabine. At 4 h, HeLa cells bine. In contrast, DU145R cells exhibited 6300-fold resistance to had accumulated 1221 pmol/106 cells of uridine and only 9 pmol/106 troxacitabine and 350- and 300-fold cross-resistance to gemcitabine cells of troxacitabine. and cytarabine, respectively (data not shown). Transportability of [3H]Troxacitabine by the Sodium-depen- The metabolism of [3H]troxacitabine and [3H]deoxycytidine by dent Concentrative Nucleoside Transporters. The transportability DU145 and DU145R cells was compared after 4- and 24-h exposures of troxacitabine by each of the three known human members of the (Fig. 5). Parental DU145 cells converted troxacitabine to its mono-, CNT protein family was assessed in experiments that used either di-, and triphosphorylated forms (Fig. 5A), and the diphosphate was recombinant hCNT1 or hCNT2 that was introduced into HeLa cells by the major metabolite, which was consistent with the findings of a transient transfection or native hCNT3 that was activated by stimu- previous report (2). The ATP:ADP ratios in the DU145 cell extracts lation of HL-60 cells to differentiate. HeLa and HL-60 cells were varied between 2 and 4, indicating that the higher levels of diphos- treated with 100 ␮M dilazep to block the equilibrative transport phorylated troxacitabine relative to triphosphorylated troxacitabine 3 processes, and the uptake of 10 ␮M [ H]troxacitabine was compared were not due to degradation by phosphatase(s) during extraction 3 3 with that of 10 ␮M [ H]uridine during 4-h exposures. Sodium-depen- procedures. Furthermore, when [ H]deoxycytidine metabolism was dent uptake of uridine was evident for each of the three transporters, examined in DU145 cells, dCTP was the major phosphorylated me- whereas troxacitabine uptake was low and independent of sodium. In tabolite (Fig. 5B). Troxacitabine-derived radioactivity in DU145R the experiments shown in Fig. 4, hCNT1- and hCNT2-mediated cells was low compared with that in DU145 cells (5 versus 254 uptake of uridine at 4 h (1077 and 939 pmol, respectively) was much pmol/mg protein, respectively, at 24 h) and was mostly present as greater than uptake of troxacitabine (12 and 7 pmol, respectively). The unmetabolized troxacitabine (Fig. 5C). Whereas deoxycytidine- uptake of uridine observed in the absence of sodium may represent derived radioactivity in DU145R cells was also low compared with uncoupled Naϩ/uridine fluxes, or “slippage,” a process that has been that in DU145 cells (3 versus 34 pmol/mg protein, respectively), all observed previously for recombinant CNT family members (26). A three phosphorylated metabolites (mono-, di-, and triphosphates) were similar result was obtained for hCNT3 in experiments (data not detected (Fig. 5D). These results, which demonstrated reduced accu- shown) in which sodium-dependent, dilazep-insensitive uptake of 10 mulation and an altered pattern of phosphorylation in the troxacitab- 3 3 ␮M [ H]uridine and [ H]troxacitabine was compared in differentiated ine-resistant variant, suggested a change in dCK activity. HL-60 cells. After 4-h exposures, uridine uptake was 54-fold greater Demonstration of Altered dCK Activity in DU145R Cells. The than troxacitabine uptake (644 versus 12 pmol, respectively). activity of dCK toward troxacitabine and its natural substrate, deoxy- Characterization of Troxacitabine Resistance and Metabolism cytidine, was examined in extracts from DU145R and DU145 cells in DU145R Cells. To further examine mechanisms of troxacitabine and, for comparison, in extracts from CCRF-CEM cells and its cyt- resistance, a drug-resistant variant (DU145R) of cultured human pros- arabine-resistant variants (Table 2). As reported previously (2, 35), tate DU145 cells was developed by continuously growing the cells in phosphorylation of deoxycytidine by dCK was greatly reduced in increasing concentrations of drug over an 8-month period. The tox- CEM/dCKϪ cells relative to either CCRF-CEM or CEM/ARAC8C icity of troxacitabine against DU145 and DU145R cells was compared cells, and the almost 2-fold difference observed in dCK activity with that of gemcitabine and cytarabine by assessing their relative between CCRF-CEM and CEM/ARAC8C cells was likely due to 7221

in the homologous domain of deoxyguanosine and thymidine kinases (Fig. 6). DU145R Cells Exhibit Reduced CDA Activity. Although CDA activity is normally low in human tissues, aside from liver and placenta (36), it is elevated in solid tumors (31). CCRF-CEM, CEM/ dCKϪ, and CEM/ARAC8C cells, consistent with their hemapoietic origin, lacked detectable CDA activity, whereas DU145 cells exhibited high CDA activity (Table 2). Unexpectedly, the level of CDA activity was significantly reduced in DU145R cells (Table 2). Because troxacitabine is resistant to deamination (2), the 11–12-fold reduction observed in DU145R cells suggested that reduced deamination of deoxycytidine might have contributed to troxacitabine resistance in DU145R cells. To provide evidence that a difference in CDA activity might be involved in resistance to troxacitabine, DU145 cells were exposed to graded concentrations of troxacitabine (0–10 ␮M) in the absence or presence of 50 ␮M THU, a competitive inhibitor of CDA (37). The

IC50 values for inhibition of proliferation of DU145 cells by troxacitabine in the presence of THU was increased in two separate experiments from 15 to 140 nM (9.3-fold) and from 16 to 104 nM (6.5-fold).

DISCUSSION The purpose of this study was to identify potential mechanisms of resistance to troxacitabine. Troxacitabine cytotoxicity was first compared with that of cytarabine and gemcitabine in dCK-deficient Ϫ Fig. 5. Intracellular levels of troxacitabine and deoxycytidine and their metabolites in (CEM/dCK ) and nucleoside transport-deficient (CEM/ARAC8) leu- DU145 and DU145R cells. Exponentially growing cells (5 ϫ 106 cells/T75 flasks) were 3 kemia cell lines. Although previous research has shown that cells incubated in the presence of 1 ␮M [ H]troxacitabine (specific activity, 3.9 Ci/mmol; A and 3 C)or1␮M [ H]deoxycytidine (specific activity, 18.4 Ci/mmol; B and D)for4or24h. deficient in dCK are resistant to troxacitabine (2), gemcitabine (38, Results are expressed as the relative percentage of each peak of the total counts loaded 39), and cytarabine (2, 40), a comparative analysis of all three deoxy- onto the column, determined by HPLC as described in “Materials and Methods.” Data points for troxacitabine metabolism studies represent the mean Ϯ SD of three determi- cytidine analogues has not been reported. Although CCRF-CEM cells nations. deficient in dCK activity were cross-resistant to troxacitabine, gem-

Table 2 A comparison of the dCK and CDA activities of parental and drug-resistant human leukemic and prostate cancer cell lines Enzyme activities were measured in cell extracts obtained from exponentially growing cells as described in “Materials and Methods.” Activities are expressed as nmol of product formed/h/mg protein and are the mean Ϯ SD of three experiments. dCK

Cell line Deoxycytidine Troxacitabine CDA CCRF-CEM 5.49 Ϯ 1.55 0.98 NDa CEM/ARAC8C 3.17 Ϯ 0.67 0.10 ND Ϫ CEM/dCK 0.04 Ϯ 0.01 ND ND DU145 0.93 Ϯ 0.38 1.49 Ϯ 0.11 162 Ϯ 69 DU145R 0.14 Ϯ 0.03 ND 14.26 Ϯ 0.4 a ND, not detected. metabolic differences that have arisen since the selection of the resistant variant in 1980 (21). Although phosphorylation of deoxycytidine was also reduced in DU145R cells relative to DU145 cells, there was only a 7-fold difference in dCK activity between DU145 and DU145R cells, whereas there was a 137-fold difference between CEM/dCK- and CCRF-CEM cells. Phosphorylation of troxacitabine, which was observed in extracts from both CCRF-CEM and DU145 cells, was not detected in either of the resistant variants (Table 2). To determine whether a mutation in dCK of DU145R cells was present and could be responsible for the observed differences in Fig. 6. PCR amplification and sequence analysis of dCK cDNA from cytoplasmic RNA activities, a partial cDNA encoding dCK was amplified from both of DU145 and DU145R cells. A, PCR products were analyzed on 1% agarose gels and DU145 and DU145R cells and sequenced. A single base change was stained with ethidium bromide. PCR amplification of GAPDH was used to control for mRNA levels. B, partial sequence of human and mouse dCK and other human de- detected in which a G residue at position 434 in DU145 dCK was oxynucleoside kinases as well as the unrelated Drosophila melanogaster nucleoside converted to T in DU145R dCK, resulting in a nonconserved change kinase. Accession numbers in SwissProt database are as follows: human dCK, P27707; murine dCK, P43346; human dGK, Q16854; human TK2, O00142. Accession number in from tryptophan to leucine at position 92 in the protein. This trypto- the European Molecular Biology Laboratory database is as follows: Drosophila melano- phan is conserved in all known sequences of dCK and is also present gaster dNK, Y18048.1. 7222

Downloaded from cancerres.aacrjournals.org on September 25, 2021. © 2001 American Association for Cancer Research. UPTAKE AND RESISTANCE TO TROXACITABINE citabine, and cytarabine, CCRF-CEM cells deficient in nucleoside of deoxycytidine (the natural substrate for CDA and dCK) would transport activity were cross-resistant to gemcitabine and cytarabine confer resistance by increasing deoxycytidine levels, thereby increas- but exhibited only low-level resistance to troxacitabine. These results ing its capacity to compete with troxacitabine for phosphorylation by suggested a different mechanism of cellular uptake for troxacitabine dCK. Interestingly, the antiproliferative activity of DMDC, an anti- than that of cytarabine and gemcitabine and a common initial activa- tumor nucleoside analogue resistant to deamination, is also modulated tion pathway whereby phosphorylation is catalyzed by dCK. Although by changes in CDA activity. Cells treated with THU, a competitive troxacitabine was reported previously to be transported by hENT1 inhibitor of CDA (47), exhibited reduced sensitivity toward gemcit- (11), the low level of troxacitabine resistance observed in CEM/ abine but increased sensitivity toward DMDC (19). The finding that ARAC8C cells as compared with CCRF-CEM cells, which possess THU protected DU145 cells from troxacitabine toxicity was consist- only hENT1-mediated nucleoside transport activity (41, 42), sug- ent with the conclusion that the coincidental reduction in CDA activ- gested that a functional nucleoside transporter is not required for ity observed in DU145R cells contributed to the overall resistance manifestation of troxacitabine toxicity. mechanism to troxacitabine, similar to what has been demonstrated The demonstration of very low rates of uptake of troxacitabine recently for DMDC (48). relative to uridine in transport-competent CCRF-CEM cells and sim- In conclusion, troxacitabine exhibited mechanisms of cellular up- ilar rates of uptake of troxacitabine and uridine in transport-deficient take, metabolism, and resistance that differed from those of the other cells suggested that the primary mode of uptake of troxacitabine in deoxycytidine analogues, cytarabine and gemcitabine. These differ- both CCRF-CEM cell lines was passive diffusion. In studies of uptake ences give troxacitabine properties that may be beneficial to patients of radiolabeled troxacitabine by cell lines with defined nucleoside who are refractory to cytarabine and/or gemcitabine. Indeed, signifi- transport activities, troxacitabine was shown to be a poor permeant for cant antileukemic activity has been observed with troxacitabine in a the molecularly characterized equilibrative (hENT1 and hENT2) and Phase I clinical trial in patients with primary refractory or relapsed sodium-dependent (hCNT1, hCNT2, and hCNT3) nucleoside trans- acute myeloid leukemia after prior therapy with cytarabine (49). porters, and thus its uptake was probably not mediated by these Troxacitabine is currently being evaluated in Phase II studies for acute nucleoside transporters. Although minor contributions from other myeloid leukemia, chronic myeloid leukemia (blast phase), and pan- uncharacterized transporters have not been ruled out, the low uptake creatic cancer. observed for troxacitabine was consistent with a diffusion model and with results of previous studies with L-nucleosides, which indicated that the L-enantiomers of several natural nucleosides (thymidine, ACKNOWLEDGMENTS uridine, and adenosine) are poor permeants of hENT1 and nonper- We thank Cheryl Santos, Louise Bernier, and Jose´e Dugas for technical meants of hENT2- and hCNT2-mediated processes (27, 33, 43–45). assistance with cytotoxicity and kinase assays; Thack Lang for providing The basis for the lack of agreement between the results reported here CNT-containing expression vectors; David Bouffard and Lorraine Leblond for for DU145 cells and those of a previous report (11), which suggested helpful discussions; and librarian Linda Harris for assistance with the prepa- that the permeation of troxacitabine was mediated by equilibrative ration of the manuscript. transporters, is unclear. To further explore potential mechanisms of resistance to troxacitabine, a troxacitabine-resistant cell line was developed by growing REFERENCES DU145 cells in the presence of stepwise increasing concentrations of 1. Gourdeau, H., Bibeau, L., Ouellet, F., Custeau, D., Bernier, L., and Bowlin, T. troxacitabine. Although DU145 and DU145R cells exhibited similar Comparative study of a novel nucleoside analogue (TroxatylTM, troxacitabine, BCH- 4556) and AraC against leukemic human tumor xenografts expressing high or low low and apparently nonmediated uptake of radiolabeled troxacitabine, cytidine deaminase activity. Cancer Chemother. Pharmacol., 47: 236–240, 2001. metabolism studies using radiolabeled troxacitabine or deoxycytidine 2. Grove, C. L., Guo, X., Liu, S-H., Gao, Z., Chu, C. K., and Cheng, Y-C. Anticancer indicated that a drastic reduction in the dCK activity of resistant activity of a novel nucleoside analogue with the unnatural L configuration. Cancer R Res., 55: 3008–3011, 1995. DU145 cells toward troxacitabine was the major mechanism of 3. Kadhim, S. A., Bowlin, T. L., Waud, W. R., Angers, E. G., Bibeau, L., DeMuys, J-M., resistance. DU145R cells were unable to phosphorylate troxacitabine Bednarski, K., Cimpoia, A., and Attardo, G. Potent antitumor activity of a novel ␤ and had significantly reduced intracellular levels of phosphorylated nucleoside analogue: BCH-4556 ( -L-dioxolane-cytidine) in human renal cell carcinoma xenograft tumor models. Cancer Res., 57: 4803–4810, 1997. deoxycytidine species. These results and the cross-resistance toward 4. Rabbini, S. A., Harakidas, P., Bowlin, T. L., and Attardo, G. Effect of nucleoside cytarabine and gemcitabine suggested a deficiency in and a change in analogue BCH-4556 on prostate cancer growth and metastasis in vitro and in vivo. Cancer Res., 58: 3461–3465, 1998. specificity of dCK, the common rate-limiting enzyme for the activa- 5. Weitman, S., Marty, J., Jolivet, J., Locas, C., and Von Hoff, D. The new dioxolane, tion of these three deoxycytidine analogues. A change in dCK spec- (-)-2Ј-deoxy-3Ј-oxacytidine (BCH-4556, troxacitabine), has activity against pancre- ificity has been reported previously for a cytarabine-resistant rat atic human xenografts. Clin. Cancer Res., 6: 1574–1578, 2000. 6. Kukhanova, M., Liu, S-H., Mozzherin, D., Lin, T-S., Chu, C. K., and Cheng, Y. C. leukemia cell line (38). PCR analysis confirmed that the amount of L- and D-enantiomers of 2Ј,3Ј-dideoxycytidine 5Ј-triphosphate analogs as substrates dCK mRNA isolated from DU145R cells was significantly reduced for human DNA polymerases. J. Biol. Chem., 270: 23055–23059, 1995. compared with that isolated from the parental cells, suggesting that 7. Plunkett, W., Huang, P., and Gandhi, V. Preclinical characteristics of gemcitabine. Anticancer Drugs, 6: 7–13, 1995. only one allele of the dCK gene was expressed. It is also possible that 8. Huang, P., Chubb, S., Hertel, L. W., Grindey, G. B., and Plunkett, W. Action of the mutation, which resulted in a change from a highly conserved 2Ј2-difluorodeoxycytidine on DNA synthesis. Cancer Res., 51: 6110–6117, 1991. 9. Grant, S. Ara-C: cellular and molecular pharmacology. Adv. Cancer Res., 72: tryptophan residue to a leucine residue at position 92 in dCK, may 197–233, 1998. have destabilized the dCK mRNA. These observations are similar to 10. Ohno, Y., Spriggs, D., Matsukage, A., Ohno, T., and Kufe, D. Effects of 1-␤-D- those obtained in a CCRF-CEM cell line resistant to dideoxycytidine arabinofuranosylcytosine incorporation on elongation of specific DNA sequences by DNA polymerase ␤. Cancer Res., 48: 1494–1498, 1988. (46). 11. Grove, K. L., and Cheng, Y. C. Uptake and metabolism of the new anticancer R Because DU145 cells retained the capacity to phosphorylate de- compound ␤-L-(-)-dioxolane-cytidine in human prostate carcinoma DU-145 cells. oxycytidine, although at reduced levels compared with parental cells, Cancer Res., 56: 4187–4191, 1996. 12. Mackey, J. R., Mani, R. S., Selner, M., Mowles, D., Young, J. D., Belt, J. A., the capacity of the resistant cells for deamination of deoxycytidine Crawford, C. R., and Cass, C. E. Functional nucleoside transporters are required for was examined. CDA activity in the troxacitabine-resistant DU145R gemcitabine influx and manifestation of toxicity in cancer cell lines. Cancer Res., 58: Ͻ 4349–4357, 1998. cells was reduced to 10% of that observed in the sensitive parental 13. Mackey, J. R., Yao, S. Y., Smith, K. M., Karpinski, E., Baldwin, S. A., Cass, C. E., DU145 cells. One possible explanation is that decreased deamination and Young, J. D. Gemcitabine transport in Xenopus oocytes expressing recombinant 7223

plasma membrane mammalian nucleoside transporters. J. Natl. Cancer Inst. (Bethes- 30. Ruiz van Haperen, V. W., Veerman, G., Eriksson, S., Stegmann, A. P., and Peters, da), 91: 1876–1881, 1999. G. J. Induction of resistance to 2Ј,2Ј-difluorodeoxycytidine in the human ovarian 14. Ullman, B. Dideoxycytidine metabolism in wild type and mutant CEM cells deficient cancer cell line A2780. Semin. Oncol., 22: 35–41, 1995. in nucleoside transport or deoxycytidine kinase. Adv. Exp. Med. Biol., 253B: 415– 31. Ruiz van Haperen, V. W. T., Veerman, G., Braakhuis, B. J. M., Vermorken, J. B., 420, 1989. Boven, E., Leyva, A., and Peters, G. J. Deoxycytidine kinase and deoxycytidine 15. Gati, W. P., Paterson, A. R., Larratt, L. M., Turner, A. R., and Belch, A. R. Sensitivity deaminase activities in human tumour xenografts. Eur. J. Cancer, 29A: 2132–2137, of acute leukemia cells to cytarabine is a correlate of cellular es nucleoside transporter 1993. site content measured by flow cytometry with SAENTA-fluorescein. Blood, 90: 32. Ullman, B. Mutational analysis of nucleoside and nucleobase transport. In: D. Kessel 346–353, 1997. (ed.), Resistance to Antineoplastic Drugs, pp. 293–315. CRC Press, 1989. 16. Cass, C. E., Young, J. D., Baldwin, S. A., Cabrita, M. A., Graham, K. A., Griffiths, 33. Paterson, A. R. P., Clanachan, A. S., Craik, J. D., Gati, W. P., Jakobs, E. S., Wiley, M., Jennings, L. L., Mackey, J. R., Ng, A. M., Ritzel, M. W., Vickers, M. F., and Yao, J. S., and Cass, C. E. Plasma membrane transport of nucleosides, nucleobases and S. Y. Nucleoside transporters of mammalian cells. In: Amidon and Sadee (eds.), nucleotides: an overview. In: S. Imai and M. Nakazawa (eds.), Role of Adenosine and Membrane Transporters as Drug Targets, pp. 313–352. New York: Kluwer Academic/ Adenine Nucleotides in the Biological System, pp. 133–149. Elsevier Science Pub- Plenum Publishers, 1999. lishers BV, 1991. 34. Boumah, C. E., Hogue, D. L., and Cass, C. E. Expression of high levels of nitroben- 17. Baldwin, S. A., Mackey, J. R., Cass, C. E., and Young, J. D. Nucleoside transporters: zylthioinosine-sensitive nucleoside transport in cultured human choriocarcinoma molecular biology and implications for therapeutic development. Mol. Med. Today, (BeWo) cells. Biochem. J., 288: 987–996, 1992. 5: 216–224, 1999. 35. Johnson, M. A., and Fridland, A. Phosphorylation of 2Ј,3Ј-dideoxyinosine by cyto- 18. Vickers, M. F., Young, J. D., Baldwin, S. A., Mackey, J. R., and Cass, C. E. solic 5Ј-nucleotidase of human lymphoid cells. Mol. Pharmacol., 36: 291–295, 1989. Nucleoside transporter proteins: emerging targets for drug discovery. Emerging 36. Laliberte, J., and Momparler, R. L. Human cytidine deaminase: purification of the Therapeutic Targets, 4: 515–539, 2000. enzyme, cloning, and expression of its complimentary DNA. Cancer Res., 54: 19. Eda, H., Ura, M., F-Ouchi, K., Tanaka, Y., Miwa, M., and Ishitsuka, H. The 5401–5407, 1994. antiproliferative activity of DMDC is modulated by inhibition of cytidine deaminase. 37. Riva, C., Barra, Y., Carcassonne, Y., Cano, J-P., and Rustum, Y. Effect of tetrahy- Cancer Res., 58: 1165–1169, 1998. drouridine on metabolism and transport of 1-␤-D-arabinofuranosylcytosine in human 20. Mansour, T. S., Jin, H., Wang, W., Dixit, D. M., Evans, C. A., Tse, H. L. A., Belleau, cells. Exp. Chemother., 38: 358–366, 1992. B., Gillard, J. W., Hooker, E., Ashman, C., Cammack, N., Salomon, H., Belmonte, 38. Bergman, A. M., Pinedo, H. M., Jongsma, A. P., Brouwer, M., Ruiz van Haperen, A. R., and Wainberg, M. A. Structure-activity relationship among a new class of V. W., Veerman, G., Leyva, A., Eriksson, S., and Peters, G. J. Decreased resistance Ј Ј antiviral and heterosubstituted 2 ,3 -dideoxynucleoside analogues. Nucleosides Nu- to gemcitabine (2Ј,2Ј-difluorodeoxycytidine) of 1-␤-D-arabinofuranosylcytosine- cleotides, 14: 627–635, 1995. resistant myeloblastic murine and rat leukemia cell lines: role of altered activity and 21. Dow, L. W., Bell, D. E., Poulakos, L., and Fridland, A. Differences in metabolism and substrate specificity of deoxycytidine kinase. Biochem. Pharmacol., 57: 397–406, cytotoxicity between 9-␤-D-arabinofuranosyladenine and 9-␤-D-arabinofuranosyl-2- 1999. fluoradenine in human leukemic lymphoblasts. Cancer Res., 40: 1405–1410, 1980. 39. Bergman, A. M., Giaccone, G., van Moorsel, C. J., Mauritz, R., Noordhuis, P., 22. Ritzel, M. W. L., Ng, A. M. L., Yao, S. Y. M., Graham, K. A., Loewen, S. K., Smith, Pinedo, H. M., and Peters, G. J. Cross-resistance in the 2Ј,2Ј-difluorodeoxycytidine K. M., Ritzel, G. W., Mowles, D. A., Carpenter, P., Chen, X-Z., Karpinski, E., Hyde, (gemcitabine)-resistant human ovarian cancer cell line AG6000 to standard and R. J., Baldwin, S. A., Cass, C. E., and Young, J. D. Molecular identification and investigational drugs. Eur. J. Cancer, 36: 1974–1983, 2000. characterization of novel human and mouse concentrative Naϩ-nucleoside cotrans- 40. Verhoef, V., Sarup, J., and Fridland, A. Identification of the mechanism of activation porter proteins (hCNT3 and mCNT3) broadly selective for purine and pyrimidine of 9-␤-D-arabinofuranosyladenine in human lymphoid cells using mutants deficient in nucleosides (system. cib). J. Biol. Chem., 276: 2914–2927, 2001. nucleoside kinases. Cancer Res., 41: 4478–4483, 1981. 23. Ritzel, M. W. L., Yao, S. Y. M., Huang, M. Y., Elliott, J. F., Cass, C. E., and Young, 41. Belt, J. A., Marina, N. M., Phelps, D. A., and Crawford, C. R. Nucleoside transport J. D. Molecular cloning and functional expression of cDNAs encoding a human in normal and neoplastic cells. Adv. Enzyme Regul., 33: 235–252, 1993. Naϩ-nucleoside cotransporter (hCNT1). Am. J. Physiol., 272: C707–C714, 1997. 42. Crawford, C. R., Ng, C. Y., Ullman, B., and Belt, J. A. Identification and reconsti- 24. Ritzel, M. W., Yao, S. Y., Ng, A. M., Mackey, J. R., Cass, C. E., and Young, J. D. tution of the nucleoside transporter of CEM human leukemia cells. Biochim. Biophys. Molecular cloning, functional expression and chromosomal localization of a cDNA Acta, 1024: 289–297, 1990. ϩ 43. Foga, I. O., Geiger, J. D., and Parkinson, F. E. Nucleoside transporter-mediated encoding a human Na /nucleoside cotransporter (hCNT2) selective for purine 3 nucleosides and uridine. Mol. Membr. Biol., 15: 203–211, 1998. uptake and release of [ H]L-adenosine in DDT1 MF-2 smooth muscle cells. Eur. 25. Graham, K. A., Leithoff, J., Coe, I. R., Mowles, D., Mackey, J. R., Young, J. D., and J. Pharmacol., 318: 455–460, 1996. 44. Gati, W. P., Dagnino, L., and Paterson, A. R. Enantiomeric selectivity of adenosine Cass, C. E. Differential transport of cytosine-containing nucleosides by recombinant transport systems in mouse erythrocytes and L1210 cells. Biochem. J., 263: 957–960, human concentrative nucleoside transporter protein hCNT1. Nucleosides Nucleotides 1989. Nucleic Acids, 19: 415–434, 2000. 45. Casillas, T., Delicado, E. G., Garcia-Carmona, F., and Miras-Portugal, M. T. Kinetic 26. Fang, X., Parkinson, F. E., Mowles, D. A., Young, J. D., and Cass, C. E. Functional and allosteric cooperativity in L-adenosine transport in chromaffin cells. A mnemon- characterization of a recombinant sodium-dependent nucleoside transporter with ical transporter. Biochemistry, 32: 14203–14209, 1993. selectivity for pyrimidine nucleosides (cNT1rat) by transient expression in cultured 46. Owens, J. K., Shewach, D. S., Ullman, B., and Mitchell, B. S. Resistance to mammalian cells. Biochem. J., 317: 457–465, 1996. 1-␤-D-arabinofuranosylcytosine in human T-lymphoblasts mediated by mutations 27. Boleti, H., Coe, I. R., Baldwin, S. A., Young, J. D., and Cass, C. E. Molecular within the deoxycytidine kinase gene. Cancer Res., 52: 2389–2393, 1992. identification of the equilibrative NBMPR-sensitive (es) nucleoside transporter and 47. Camiener, G. W. Studies of the enzymatic deamination of ara-cytidine. V. Inhibition demonstration of an equilibrative NBMPR-insensitive (ei) transport activity in human in vitro and in vivo by tetrahydrouridine and other reduced pyrimidine nucleosides. erythroleukemia (K562) cells. Neuropharmacology, 36: 1167–1179, 1997. Biochem. Pharmacol., 17: 1981–1991, 1968. 28. Harley, E. R., Paterson, A. R., and Cass, C. E. Initial rate kinetics of the transport of 48. Miwa, M., Eda, H., Ura, M., Ouchi, K. F., Keith, D. D., Foley, L. H., and Ishitsuka, adenosine and 4-amino-7-(␤-D-ribofuranosyl)pyrrolo[2,3-d]pyrimidine (tubercidin) in H. High susceptibility of human cancer xenografts with higher levels of cytidine cultured cells. Cancer Res., 42: 1289–1295, 1982. deaminase to a 2Ј-deoxycytidine antimetabolite, 2Ј-deoxy-2Ј-methylidenecytidine. 29. Ruiz van Haperen, V. W., Veerman, G., Boven, E., Noordhuis, P., Vermorken, J. B., Clin. Cancer Res., 4: 493–497, 1998. and Peters, G. J. Schedule dependence of sensitivity to 2Ј,2Ј-difluorodeoxycytidine 49. Giles, F. J., Cortes, J. F., Beran, M., Estey, E., Proulx, L., Jolivet, J., Bivins, C., Rios, (Gemcitabine) in relation to accumulation and retention of its triphosphate in solid M. B., and Kantarjian, H. Phase I study of BCH-4556 in patients with advanced tumour cell lines and solid tumours. Biochem. Pharmacol., 48: 1327–1339, 1994. leukemia. Proc. Am. Soc. Clin. Oncol., 18: 28a, 1999.

7224

Downloaded from cancerres.aacrjournals.org on September 25, 2021. © 2001 American Association for Cancer Research. Mechanisms of Uptake and Resistance to Troxacitabine, a Novel Deoxycytidine Nucleoside Analogue, in Human Leukemic and Solid Tumor Cell Lines

Henriette Gourdeau, Marilyn L. Clarke, France Ouellet, et al.

Cancer Res 2001;61:7217-7224.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/61/19/7217

Cited articles This article cites 41 articles, 20 of which you can access for free at: http://cancerres.aacrjournals.org/content/61/19/7217.full#ref-list-1

Citing articles This article has been cited by 14 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/61/19/7217.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/61/19/7217. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.