Characterization of a Gemcitabine-Resistant Murine Leukemic Cell Line

5614 Vol. 10, 5614 –5621, August 15, 2004

Clinical Cancer Research

Characterization of a Gemcitabine-Resistant Murine Leukemic Cell Line: Reversion of In vitro Resistance by a Mononucleotide Prodrug Lars Petter Jordheim,1 Emeline Cros,1 Marie-He´le`ne Gouy,2 Carlos M. Galmarini,1,3 Suzanne Peyrottes,2 John Mackey,4 Christian Perigaud,2 and Charles Dumontet1 1 INSERM U590, Laboratoire de Cytologie Analytique, Faculte´ de Me´decine Rockefeller, Universite´ Claude Bernard Lyon I, Lyon, France; 2UMR 5625 Centre National de la Recherche Scientifique– UM II, Universite´ Montpellier II, Montpellier, France; 3Centre d’Oncologie Mole´culaire, Centre Le´on Be´rard, Lyon, France; 4 Medical and Experimental Oncology, Department of Oncology, University of Alberta, Alberta, Canada

ABSTRACT Resistance to cytotoxic nucleoside analogues is a major problem in cancer treatment. The cellular mechanisms involved in this phenomenon have been studied for several years, and some factors have been identified. Various strategies to overcome resistance have been suggested, but none has yet shown efficacy in vivo. We developed a gemcitabineresistant cell line (L1210 10K) from the murine leukemic L1210 strain (L1210 wt) by continuous exposure to increasing concentrations of gemcitabine. L1210 10K is highly resistant to gemcitabine (14,833-fold), 1-␤-D-arabinofuranosylcytosine (ara-C; 2,100-fold), troxacitabine (>200-fold), and cladribine (160-fold) and slightly resistant to trimidox (7.22-fold), but does not display cross-resistance to fludarabine or nonnucleoside anticancer drugs. Deoxycytidine kinase mRNA was not detected by quantitative real-time reverse transcription-PCR in L1210 10K cells, whereas expression of thymidine kinase 1 and ribonucleotide reductase subunit R2 gene was moderately reduced. L1210 10K cells also demonstrated in vivo resistance to nucleoside analogues: gemcitabine- or ara-C-treated mice carrying L1210 10K had significantly shorter survival than gemcitabine- or ara-C-treated mice carrying L1210 wt (P < 0.05). UA911, a

Received 3/12/04; revised 4/22/04; accepted 4/28/04. Grant support: Grants from Ligue Contre le Cancer–Comite´ de Saoˆne et Loire, ARC, and EWS Stiftelsen (L. Jordheim). M. Gouy received a doctoral fellowship from the Ministe`re de l’Education Nationale, de la Recherche et de la Technologie, France. C. Galmarini is a Allocataire de Recherche at Centre Le´on Be´rard in Lyon, France. Part of this research was supported by the Association pour la Recherche sur le Cancer (ARC, France). 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 18 U.S.C. Section 1734 solely to indicate this fact. Requests for reprints: Charles Dumontet, Laboratoire de He´matologie, Hoˆpital Edouard Herriot, 5 Place d’Arsonval, 69003 Lyon, France. Phone: 04 78 77 72 36; E-mail: [email protected].

mononucleotide prodrug (pronucleotide) of ara-C was found to significantly sensitize L1210 10K cells in vitro. These results suggest that reduced deoxycytidine kinase expression is a mechanism of resistance to gemcitabine that is relevant in vivo and can be circumvented by a prodrug approach.

INTRODUCTION Gemcitabine (2⬘,2⬘-difluorodeoxycytidine) is a nucleoside analogue that shows activity against hematologic malignancies and solid tumors alone (1– 4) or in combination with other chemotherapeutic agents (5–7). The cellular metabolism of gemcitabine is similar to that of physiologic nucleosides (reviewed in Ref. 8). Once inside the cell, it is phosphorylated by deoxycytidine kinase (dCK) to its monophosphorylated form difluoro-dCMP, and is phosphorylated further by other intracellular kinases to the metabolically active forms difluoro-dCDP and difluoro-dCTP (9 –11). Diphosphorylated gemcitabine is an inhibitor of ribonucleotide reductase (12), thereby causing a decrease in dCTP pools and decreased feedback inhibition of dCK, leading to enhanced phosphorylation of gemcitabine (13, 14), a process designated as “selfpotentiation.” The triphosphorylated form is cytotoxic by incorporation into DNA, causing masked chain termination after the incorporation of an additional nucleotide after gemcitabine (15, 16). The cytotoxicity of gemcitabine in nondividing cells has been attributed to the fact that difluoro-dCTP can disturb RNA metabolism by inhibiting CTP synthetase (17) or by direct incorporation into newly synthesized RNA (18). Resistance to gemcitabine may involve a variety of mechanisms. In humans, cellular influx of gemcitabine involves different membrane nucleoside transporters, such as hENT1, hENT2, hCNT1, and hCNT3. It has been shown that nucleoside-transporter-deficient cells display resistance to gemcitabine (19). Various authors, including ourselves, have shown that dCK-deficient lines display a gemcitabine-resistant phenotype (20). Gemcitabine and its monophosphorylated form can be inactivated by intracellular enzymes such as cytidine deaminase and dCMP deaminase, producing difluorodeoxyuridine from 2⬘,2⬘-difluorodeoxycytidine and difluoro-dUMP from difluoro-dCMP, respectively (10, 21). Intracellular 5⬘-nucleotidases are also likely to be involved in the inactivation of 2⬘,2⬘-difluorodeoxycytidine by dephosphorylating difluoro-dCMP or by modifying pools of endogenous nucleotides. To develop a model of gemcitabine resistance suited for in vivo strategies of reversion of drug resistance, we have developed the L1210 10K model, which displays resistance both in vitro and in vivo.

MATERIALS AND METHODS Reagents. Compounds used for in vitro and in vivo experiments were 1-␤-D-arabinofuranosylcytosine (ara-C; Aracytine; Pharmacia, Puurs, Belgium), cisplatin (Merck, Lyon,

Clinical Cancer Research 5615

Table 1 Gene

Forward primer (5⬘–3⬘)

ENT-1 MRP5 dCK dGK TK1 TK2 cN-II cN-III cdN mdN RNR R1 RNR R2

AACTCTCCACCCACCAACAG CCCAGGCAACAGAGTCTAACC TGAGGATTGGGAAGTGGTTC AAACCGGCTTTTGTTACACG CCAGATCGCCCAGTACAAGT TCACCTGTACGGTTGATGGA ACCGTAGCATGTTCCAGGAC CCAAATCTCACGGTTTGCTT GCGAGAGATGAACGACATGA CCTCTTTACCTCCTGCCACA CCCAATGAGTGTCCTGGTCT CCTACTAACCCCAGCGTTGA

Primers, probes, and PCR conditions used in reverse transcription-PCR studies Reverse primer (5⬘–3⬘)

Probe (5⬘-FAM–3⬘-TAMRA)

TTCCACCTCAGCAGTCACAG CGGTAATTCAATGCCCAAGTC CCACCGTTCAGGTTTCTCAT TGCAGAGCACTCAAGAGGAA GCGGCATTTTCAGAGAAGTC TTGGCCATCATTTCACAGAA GGGAGTGGTTCGCAGATAAA TTGCCATCCTTGACGACATA GGGCTCCCTTAAGGAGAAGA CCTGCCGGATAACTTCTTCA ATGAGCAGGTCTCCCATGAC ACTGGGATCCTCCAGGAAGT GTTCTGCTGGTTGCTCTTCC GTTTCAGAGCTTCCCAGTGC

MgCl2 (mM) Denaturation 2 2 2 2 2 4 2 2 2 3

95°C/15 min 95°C/15 min 95°C/5 min 95°C/0 min 95°C/0 min 95°C/0 min 95°C/5 min 95°C/0 min 95°C/15 min 95°C/15 min 95°C/0 min 95°C/15 min

Annealing/extension 58°C/5 min; 72°C/10 min 55°C/5 min; 72°C/9 min 55°C/5 min; 72°C/8 min 60°C/10 min; 72°C/12 min 60°C/10 min; 72°C/12 min 60°C/10 min; 72°C/12 min 55°C/5 min; 72°C/8 min 60°C/10 min; 72°C/12 min 58°C/5 min; 72°C/10 min 58°C/5 min; 72°C/10 min 57°C/10 min; 72°C/8 min 58°C/5 min; 72°C/10 min

Abbreviations: FAM, 6-carboxyfluorescein; TAMRA, 6-carboxytetramethylrhodamine; ENT-1, murine equilibrative nucleoside transporter; MRP5, multidrug resistance protein 5; dCK, deoxycytidine kinase; dGK, deoxyguanosine kinase; TK1, thymidine kinase 1; TK2, thymidine kinase 2; cN-II, cytoplasmic 5⬘-nucleotidase II; cN-III, cytoplasmic 5⬘-nucleotidase III; cdN, cytoplasmic 5⬘-deoxynucleotidase 1; mdN, mitochondrial 5⬘-deoxynucleotidase 2; RNR R1, large subunit of ribonucleotide reductase; RNR R2, small subunit of ribonucleotide reductase.

France), gemcitabine (Gemzar; Lilly, Indianapolis, IN), Triapine (3-aminopyridine-2-carboxaldehyde thiosemicarbazone; gift from Vion Pharmaceuticals, Inc., New Haven, CT), Didox (N3,4-trihydroxybenzamide) and Trimidox (N-3,4,5-tetrahydroxybenzenecarboximidamide; gifts from Molecules for Health, Inc., Richmond, VA), 9-␤-D-arabinofuranosyl-2-fluoro-AMP (Fludara; Schering Laboratories, Lys-les-Lannoy, France), cladribine (Leustatine; Janssen-Cilag, Boulogne-Billancourt, France), troxacitabine (Troxatyl; Shire, Hampshire, UK), etoposide (Ve´ pe´ side-Sandoz; Laboratoire Sandoz, Rueil-Malmaison, France), and doxorubicin (Adriablastine; Pharmacia & Upjohn, Saint-Quentin-en-Yvelines, France). 9-␤-D-Arabinofuranosyl-2fluoroadenine, hydroxyurea, methylthiazoletetrazolium, isopropanolol, and NaCl were purchased from Sigma Aldrich (SaintQuentin, France), and HCl was from Merck. The TRIzol RNA extraction kit, murine Moloney leukemia virus reverse transcriptase, TaqDNA polymerase, and RPMI 1640 cell culture medium were purchased from Invitrogen (Cergy Pontoise, France); L-glutamine and penicillin-streptomycin were from Gibco (Cergy Pontoise, France); and fetal bovine serum was from PAN Biotech GmbH (Aidenbach, Germany). UA911, the bis(S-pivaloyl-2-thioethyl)phosphotriester derivative of ara-C, was obtained by adaptation of a previously published procedure and dissolved in dimethyl sulfoxide (100%) to a final concentration of 10 mM (22). Cell Culture. L1210 murine leukemic cells were grown in RPMI 1640 containing L-glutamine, penicillin (200 IU/ml), streptomycin (200 ␮g/ml), and fetal bovine serum (10%) at 37°C in the presence of 5% CO2. Cytotoxicity Studies. Drug dilutions were prepared in RPMI medium at increasing concentrations (10⫻), and 100 ␮l were added to each well in a 24-well plate (Becton Dickinson, Franklin Lakes, NJ). L1210 cells (900 ␮l containing 20,000 cells) were added to the wells and incubated at 37°C for 72 h. Methylthiazoletetrazolium (500 ␮g) was added, and after 2 h of incubation at 37°C, the supernatant was removed and 300 ␮l of isopropanol–HCl–H2O (90:9:1, v/v/v) were added to solubilize the formazan crystals. Absorbance was measured spectrophotometrically with a microplate reader (Labsystem Multiskanner

RC). The IC50 was defined as the concentration inhibiting proliferation to a level equal to 50% of that of controls, and the resistance ratio was the ratio between the IC50 of the gemcitabine-resistant L1210 10K cell line and the IC50 of the sensitive parental cell line L1210 wt. IC50 values were determined from concentration– effect curves generated using Microsoft Excel. Statistical significance was determined using Student’s t test. In vivo Studies. On day 0, groups of 4- to 6-week-old female B6D2F1 mice (IFFA CREDO, Arbresle, France) received intraperitoneal injections containing 106 L1210 wt or 10K cells in exponential growth phase. Gemcitabine-treated mice were treated on days 1, 4, 7, and 10 at similar doses. Three dose levels were tested: 360, 240, and 120 mg/kg. For ara-C studies, several schedules were tested: administration of 1000 mg/kg on days 1 and 2 or 300, 150, or 75 mg/kg on days 1– 4 and 7–10, for a total of eight injections. The survival analysis was performed by constructing Kaplan–Meier survival curves. Drug solutions were prepared in distilled water at 24, 16, and 8 mg/ml for gemcitabine and 66.6, 20, 10, and 5 mg/ml for ara-C. Reverse Transcription-PCR. Total mRNA was extracted by TRIzol as described previously (23). mRNA was converted to cDNA by reverse transcription at 37°C for 1 h using Moloney leukemia virus reverse transcriptase as described in the manufacturer’s manual. cDNA levels were standardized by use of primers for ribosomal 18S (Applied Biosystems, Foster City, CA), and specific mRNA levels were quantified by real-time reverse transcription-PCR in a LightCycler thermal cycler (Roche, Mannheim, Germany) in a final reaction volume of 6.67 ␮l containing forward and reverse primers (300 nM each), MgCl2 (variable concentrations), deoxynucleotide triphosphates (500 ␮M) LC-FastStart DNA Master SYBRGreen or LC-FastStart DNA Master Hybridization Probes (Roche, Mannheim, Germany), and probes (130 nM). Cycling conditions and primer sequences are presented in Table 1. Results were analyzed with RelQuant software (Roche) as indicated in the user’s manual, and results are given as mean values of three experiments expressed in arbitrary units using the parental L1210 wt cell line as reference. The statistical significance was determined using Student’s t test.

5616 Study of Gemcitabine-Resistant L1210 Cell Line

Table 2

PCR primers used for sequencing the dCK gene (Locus ID: NT_039307)

Amplicon

Forward primer (5⬘–3⬘)

Reverse primer (5⬘–3⬘)

Amplicon size (bp)

Exon 1 Exon 2 Exon 3 Exon 4 Exon 5 Exon 6 3⬘UT

AGGAACCACAGACATGCAGA CTGCGCATGTGTTCTTCAGT AGCTAGAGCAACGTAGTGAG TCTAAGTAAGGCTCACAGGG AACGGCAGCAGACACAACTA GGGGTATTCAGCTCTTCTGT CTGACTCTGCAAGCCTTTCC

ACAGTTCCCAACAAAGGGTG GAGTTTTTGCTGTACTCCCC CAGCGCAGCCTCTAATTTGA AACGTGGGTCCTCTACAAGA AAGTACAGTGAGCAGCAACC CGGGAGGGACACAAGTTTTA CAAGGACTCTGGAGCAGGAC

458 263 402 408 207 322 166

NOTE. 3⬘-Untranscribed region (3⬘UT) corresponds to a region situated 2456 bp after the STOP codon of the dCK gene.

Sequencing of dCK Gene. Genomic DNA was prepared from L1210 wt and L1210 10K cells with a phenol– chloroform extraction method. dCK exons were amplified by PCR from 250 ng of genomic DNA using TaqDNA polymerase in a final volume of 25 ␮l containing forward and reverse primers (1.5 ␮M each; Table 2), deoxynucleotide triphosphates (200 ␮M) and MgCl2 (1.5 mM). PCR conditions included initial denaturation for 15 min at 94°C followed by 50 cycles of 94°C for 30 s, 60°C for 30 s, and 72°C for 30 s. Statistical Analysis. Statistical analyses were performed with STATISTICA version 6 (StatSoft, Inc., Tulsa, OK).

RESULTS Development and Characterization of a GemcitabineResistant Cell Line. Gemcitabine-resistant L1210 cells were developed by continuous exposure to increasing concentrations of gemcitabine over a period of 3 months, starting with an initial concentration of 1 nM, with 2–5-fold increases every 5 to 15 days. The cells obtained (L1210 10K) were viable in a medium containing 10 ␮M gemcitabine. L1210 10K cells showed no apparent morphologic differences or difference in growth rate compared with the parental L1210 wt cells. The IC50 values and resistance ratios for the gemcitabineresistant L1210 10K cell line and the sensitive parental cell line

Table 3

L1210 wt are listed in Table 3. L1210 10K cells were 15,000fold less sensitive to gemcitabine than the parental cell line. Cross-resistance was observed for other nucleoside analogues such as ara-C, cladribine, and troxacitabine, for which the resistance ratios were 2100, 160, and ⬎200, respectively. Interestingly, no cross-resistance was found for the purine nucleoside analogue fludarabine, regardless of whether cells were exposed to the clinical formulation (9-␤-D-arabinofuranosyl-2-fluoroAMP; Schering) or unphosphorylated fludarabine (Sigma). Among the ribonucleotide reductase inhibitors tested, Trimidox displayed stronger cytotoxic activity on L1210 wt than on L1210 10K cells (resistance ratio ⫽ 7.22 ⫾ 2.55), whereas Didox, hydroxyurea, and Triapine gave resistance ratios close to 1. Furthermore, L1210 10K showed no resistance to cisplatin, etoposide, doxorubicin, or homoharringtonine, suggesting that the resistance mechanism was specific to nucleoside analogues. In vivo Sensitivity to Nucleoside Analogues. The median survival of mice receiving injections of 1210 wt or L1210 10K were almost similar (17 and 14 days, respectively; P, not significant). Survival of L1210 wt-carrying mice was significantly enhanced by gemcitabine treatment, with median survivals of 28, 35, and 42 days for doses of 120, 240, and 360 mg/kg, respectively. Survival of L1210 10K-carrying mice was unaffected by gemcitabine treatment (Fig. 1). No major toxicity

IC50 values and resistance ratios in L1210 wt and L1210 10K cells as determined by methylthiazoletetrazolium assay IC50

Gemcitabine (␮M) Ara-C (␮M) Troxacitabine (␮M) Cladribine (␮M) Fara-AMP (␮M) Fara-A (␮M) Trimidox (␮M) Didox (␮M) HU (␮M) Triapine (␮M) Cisplatin (␮M) Etoposide (␮M) Doxorubicin (ng/ml) Homoharringtonine (ng/ml)

L1210 wt

L1210 10K

RR

P*

0.000217 ⫾ 0.000176 0.0266 ⫾ 0.0208 0.500 ⫾ 0.440 0.083 ⫾ 0.029 2.50 ⫾ 0.866 3.33 ⫾ 0.577 4.67 ⫾ 1.15 43.3 ⫾ 5.78 66.7 ⫾ 20.8 0.466 ⫾ 0.0577 18.7 ⫾ 10.0 5.00 ⫾ 4.24 7.75 ⫾ 3.89 0.400 ⫾ 0.100

23.7 ⫾ 11.0 53.3 ⫾ 35.1 ⬎100 12.67 ⫾ 6.43 3.00 ⫾ 0.00 3.67 ⫾ 1.12 33.3 ⫾ 11.5 46.7 ⫾ 5.78 80.0 ⫾ 17.3 0.566 ⫾ 0.0577 27.0 ⫾ 14.7 10.0 ⫾ 0.00 10.0 ⫾ 1.41 0.366 ⫾ 0.152

14,833 ⫾ 72,514 2,100 ⫾ 360 ⬎200 160 ⫾ 69.3 1.33 ⫾ 0,577 1.08 ⫾ 0.144 7.22 ⫾ 2.55 1.08 ⫾ 0.144 1.23 ⫾ 0.251 1.23 ⫾ 0.252 1.56 ⫾ 0.962 3.13 ⫾ 2.65 1.42 ⫾ 0.532 0.944 ⫾ 0.419

0.020† 0.058 0.00† 0.028† 0.37 0.68 0.013† 0.52 0.44 0.10 0.46 0.34 0.20 0.77

NOTE. Data are mean values of three independent single experiments. Abbreviations: RR, resistance ratio; ara-C, 1-␤-D-arabinofuranosylcytosine; Fara-AMP, 9-␤-D-arabinofuranosyl-2-fluoro-AMP; Fara-A, 9-␤-Darabinofuranosyl-2-fluoroadenine; HU, hydroxyurea. * Student’s t test. † Statistically significant.

Clinical Cancer Research 5617

Fig. 1 Effect of gemcitabine on survival of mice carrying L1210 wt or L1210 10K cells. Mice received 106 L1210 wt (solid lines) or L1210 10K (dotted lines) cells on day 0, and then 0.9% NaCl (F) or gemcitabine (‚) on days 1, 4, 7, and 10 (多). Gemcitabine doses: 120 mg/kg (A), 240 mg/kg (B); 360 mg/kg (C). Each group contained five 5-week-old B6D2F1 female mice. Kaplan–Meier test was significant for L1210 wt-carrying mice in A, B, and C (P ⬍ 0.05).

related to gemcitabine was observed during this experiment, and all mice died of progressive disease. Similar results were observed in mice treated with ara-C, with a median survival of 25 days in mice carrying L1210 wt

cells and treated with 1000 mg/kg ara-C (Fig. 2A). When we used different treatment schemes with a total of eight injections, median survival was 23, 21, and 20 days for animals receiving doses of 300, 150, and 75 mg/kg, respectively (Fig. 2, B–D). Survival of L1210 10K-carrying mice was unaffected by ara-C treatment in all cases (median survival ⫽ 14, 14, 13, and 13 days with the different treatment schemes). Lethal toxicity was observed in L1210 wt-carrying mice treated with 1000 mg/kg (2 of 10) and with 300 mg/kg ara-C (3 of 7). These mice died before developing tumors as a result of ara-C toxicity. Identification of Genes Involved in Resistance by Quantitative Reverse Transcription-PCR. To identify cellular mechanisms responsible for the resistance observed in L1210 10K cells, we quantified the expression of genes known to be involved in the metabolism and cytotoxicity of nucleoside analogues by quantitative reverse transcription-PCR (Fig. 3). The main difference was the total absence of dCK mRNA in L1210 10K cells, whereas this gene was found to be expressed in L1210 wt cells. Conversely, no significant difference was found in the expression of equilibrative nucleoside transporter 1, which is the main transporter of nucleoside analogues at the cell membrane level (24), or of MRP5, which has been described as a potential gemcitabine-efflux pump (25). Furthermore, we observed no differences in the levels of deoxyguanosine kinase; thymidine kinase 2; the 5⬘nucleotidases cN-II, cN-III, cdN, and mdN (see Ref. 26 for new nomenclature for 5⬘-nucleotidases); or of the large subunit of ribonucleotide reductase between L1210 10K and L1210 wt cells. A small difference in the expression of thymidine kinase 1 and in the small subunit of ribonucleotide reductase was observed, with a lower level of expression in L1210 10K cells. Analysis of the dCK Gene. Because dCK mRNA could not be detected in L1210 10K cells, we sought to determine whether the dCK gene was altered in these cells. We thus attempted to amplify the six exons of the murine dCK gene in both L1210 wt and L1210 10K cells. As shown in Fig. 4, all exons were amplified in L1210 wt cells. However, only the first five exons were amplified in the L1210 10K cells. The sixth exon as well as a 3⬘-untranscribed fragment of 166 bp starting 2456 bp after the STOP codon of the dCK gene could not be amplified in this line. These results suggest a genetic modification of chromosome 5 in the 3⬘ region of the dCK gene. The fragments corresponding to exons 2–5 were sequenced, and no difference was observed between the two cell lines. Reversion of In vitro Nucleoside Resistance by an Ara-C Prodrug. Different strategies are being studied to circumvent nucleoside analogue resistance, in particular that due to dCK deficiency. One possibility is to expose cells to compounds that are already phosphorylated. Because monophosphates are not transported by nucleoside transporters such as ENT1 and do not permeate freely through the plasma membranes, we have used in our experiments a mononucleotide prodrug (pronucleotide) of ara-C (UA911; Fig. 5). In such constructs, the phosphate negative charges were temporarily masked by two S-pivaloyl-2thioethyl groups. We have previously demonstrated that such pronucleotides are able to selectively deliver their parent 5⬘monophosphates inside cells through an esterase-mediated process, leading to the accumulation of active triphosphorylated forms (27).

5618 Study of Gemcitabine-Resistant L1210 Cell Line

Fig. 2 Effect of 1-␤-D-arabinofuranosylcytosine (ara-C) on survival of mice carrying L1210 wt or L1210 10K cells. Mice received 106 L1210 wt (solid lines) or L1210 10K (dotted lines) cells on day 0, and then 0.9% NaCl (F) or ara-C (‚) on days 1 and 2 (多 in A) or on days 1– 4 and 7–10 (多 in B, C, and D). Ara-C doses: 1000 mg/kg (A); 300 mg/kg (B); 150 mg/kg (C); 75 mg/kg (D). Each group contained ten 6-week-old B6D2F1 female mice (A) or seven 4-week-old B6D2F1 female mice (B–D). Kaplan–Meier test was significant for L1210 wt-carrying mice in C and D (P ⬍ 0.05).

To study whether the nucleoside resistance of L1210 10K cells could be circumvented by the bis-S-pivaloyl-2-thioethyl approach, we compared the cytotoxic effect of ara-C to that of UA911 in methylthiazoletetrazolium assays (Table 4). L1210 wt cells were less sensitive to UA911 cytotoxicity than to ara-C cytotoxicity (IC50 values of 0.413 ␮M versus 0.0550 ␮M, respectively), whereas L1210 10K cells were significantly more sensitive to UA911 than to ara-C (6.88 ␮M versus 78.8 ␮M, respectively). The resistance

ratios were 25.8 for UA911 and 1629 for ara-C (P ⫽ 0.00023), respectively, showing that UA911 partially circumvented the resistance of L1210 10K cells to ara-C.

DISCUSSION Gemcitabine is now widely used in the treatment of various solid tumors such as pancreatic cancer, bladder cancer, and lung

Table 4 IC50 values and resistance ratios of 1-␤-D-arabinofuranosylcytosine and UA911 in L1210 wt and L1210 10K cells as determined by methylthiazoletetrazolium assay IC50 (␮M)

Ara-C UA911

L1210 wt

L1210 10K

RR

P*

0.0550 ⫾ 0.0370 0.413 ⫾ 0.400

78.8 ⫾ 69.7 6.88 ⫾ 4.09

1629 ⫾ 410 25.8 ⫾ 20.5

0.0074† 0.020†

NOTE. Data are mean values of four single experiments. Abbreviations: RR, resistance ratio; Ara-C, 1-␤-D-arabinofuranosylcytosine. * Student’s t test. † Statistically significant. P value for comparison of resistance ratios ⫽ 0.00023.

Clinical Cancer Research 5619

Fig. 3 mRNA expression levels in L1210 wt and L1210 10K cells. Total mRNA was isolated from L1210 wt and L1210 10K cells, and the expression of genes involved in nucleotide metabolism was analyzed as described in “Materials and Methods.” Expression levels are expressed as arbitrary PCR units. ⴱ, P ⬍ 0.05, Student’s t test. All experiments were done in triplicate. (ENT-1, murine equilibrative nucleoside transporter-1; MRP5, multidrug resistance protein 5; dCK, deoxycytidine kinase; dGK, deoxyguanosine kinase; TK1, thymidine kinase 1; TK2, thymidine kinase 2; cN-II, cytoplasmic 5⬘-nucleotidase II; cN-III, cytoplasmic 5⬘-nucleotidase III; cdN, cytoplasmic 5⬘-deoxynucleotidase; mdN, mitochondrial 5⬘deoxynucleotidase; RNR R1, large subunit of ribonucleotide reductase; RNR R2, small subunit of ribonucleotide reductase).

Fig. 4 PCR amplification of exons 1– 6 and a 3⬘-untranscribed region (3⬘UT) of the murine deoxycytidine kinase (dCK) gene in L1210 wt and L1210 10K cells. Genomic DNA was isolated from L1210 wt and L1210 10K cells and analyzed by PCR as described in “Materials and Methods.” Sizes are in bp.

cancer. More recent indications include breast cancers and certain hematologic malignancies. Unfortunately, response to gemcitabine treatment is limited by resistance (28). To better understand the cellular mechanisms involved in this resistance and to construct circumvention strategies, we have developed a murine cellular model for gemcitabine resistance that can be studied both in vitro and in vivo. The L1210 10K cell line is highly resistant to gemcitabine and two other pyrimidine nucleoside analogues, ara-C and troxacitabine, as well as the purine derivative cladribine. This cross-resistance is in accordance with results from similar studies on nucleoside-analogue-resistant cell lines (20, 27, 29 –35). Surprisingly, L1210 10K was not resistant to the purine analogue fludarabine. This result indicates that there exist differences in the metabolism and/or mechanism of action of gemcitabine and fludarabine. We chose the L1210 line because it is easily transposed into a syngeneic murine model. Our animal studies on the gemcitabine-resistant cell line showed that the resistance to nucleoside analogues observed in vitro was transposable in vivo. Mice receiving injections of L1210 wt cells and treated with gemcitabine or ara-C had significantly longer survival than

untreated controls, whereas mice receiving injections of L1210 10K cells and treated with gemcitabine or ara-C did not live longer than untreated littermates. These results show that we have developed a simple animal model to study resistance to nucleoside analogues and, particularly, to evaluate different strategies aiming to revert this resistance. The major mechanism of resistance to nucleoside analogues in the L1210 10K line appears to involve drug activation. We observed that expression of the two cytoplasmic nucleoside kinases dCK and thymidine kinase 1 was altered in L1210 10K. Using quantitative reverse transcription-PCR we detected 2-fold less thymidine kinase 1 mRNA in L1210 10K than in L1210 wt cells, whereas dCK mRNA remained undetectable in this model. dCK is the kinase responsible for the phosphorylation of most of the nucleoside analogues, and its down-regulation is frequently observed in nucleoside-analogue-resistant cell lines (29, 32– 41) or in nonresponding patients treated with nucleoside analogues (42, 43). An associated down-regulation of thymidine kinase 1 had also been observed earlier in ara-C-resistant MOLT-4/8 cells (36). These observations led us to conclude that dCK deficiency plays a significant role in the resistance observed in this model. To identify combined mechanisms of resistance, which are

Fig. 5 Chemical structure of UA911, the bis(S-pivaloyl-2-thio-ethyl)phosphotriester derivative of 1-␤-D-arabinofuranosylcytosine.

5620 Study of Gemcitabine-Resistant L1210 Cell Line

likely to be present because UA911 does not allow complete reversal of the resistant phenotype, we explored other parameters involved in transport, metabolism, or targeting. We observed no or small differences in the expression of the nucleoside transporter equilibrative nucleoside transporter 1, the efflux pump MRP5, and the 5⬘-nucleotidases cN-II, cN-III, cdN, and mdN as well as for the mitochondrial nucleoside kinases deoxyguanosine kinase and thymidine kinase 2 and the ribonucleotide reductase R1 and R2 subunits. We determined the expression of these genes because of their potential role in gemcitabine resistance. It has been shown that equilibrative nucleoside transporter 1 is the major transporter of nucleoside analogues at the cell membrane (24), and MRP5 has been shown to efflux gemcitabine out of cells (25). 5⬘-Nucleotidases may be involved in drug resistance because they are likely to cause dephosphorylation of monophosphorylated nucleoside analogues and there are clinical observations showing that these proteins could play a role in nucleoside analogue resistance (44). However, recent in vitro studies with recombinant enzymes have not shown direct degradation of monophosphates (45), suggesting that they induce this resistance by modifying intracellular levels of endogenous nucleotides. Finally, ribonucleotide reductase is a target for diphosphorylated gemcitabine, and a modification of its expression could be involved in gemcitabine resistance. The mechanism leading to a down-regulation of the expression of the dCK gene in nucleoside-analogue-resistant cells remains under discussion. One hypothesis involves a disturbance in the machinery responsible for the post-translational modifications of dCK (46). Another possibility is induced hypermethylation of the dCK promoter that would lead to downregulated transcription. This phenomenon has been studied in other models resistant to nucleoside analogues and could be involved in this mechanism (47, 48). However, treatment of L1210 10K cells with 5-azacytidine did not induce dCK expression (data not shown). Finally, the direct effect of nucleoside analogues on genomic DNA might lead to mutations of the dCK gene (49). Because we were unable to detect any expression of dCK mRNA in L1210 10K cells, we tried to sequence the corresponding gene. As shown in Fig. 4, we identified a genomic modification on chromosome 5 in this cell line. In fact, prolonged exposure to gemcitabine seems to have led to a partial deletion of the 3⬘ portion of the dCK gene, as shown by the fact that exon 6 was not amplified in this cell line. This result was confirmed by the fact that a portion of the 3⬘-untranscribed region of the dCK gene was also absent in the L1210 10K line. We have obtained similar results in other dCK-deficient cell lines, confirming this effect of gemcitabine on genomic DNA (50). Different strategies to circumvent nucleoside analogue resistance have included dCK stimulation (51–54), inhibition or stimulation of DNA repair (55), modulation of membrane transport (56), or intracellular deoxynucleotide triphosphate pools. UA911, a pronucleotide of ara-C, is more efficient than ara-C against L1210 10K cells, thereby demonstrating that this approach allows at least partial sensitization of totally dCK-deficient tumor cells. The fact that we did not observe total sensitization of L1210 10K cells with UA911 might be due to differences in the mechanism of action between gemcitabine and ara-C. Future studies will exploit the in vivo model to

explore whether prodrugs are also able to sensitize resistant cells in vivo.

REFERENCES 1. Markman M, Webster K, Zanotti K, et al. Phase 2 trial of singleagent gemcitabine in platinum-paclitaxel refractory ovarian cancer. Gynecol Oncol 2003;90:593– 6. 2. Fowler WC, Jr, Van Le L. Gemcitabine as a single-agent treatment for ovarian cancer. Gynecol Oncol 2003;90:S21–3. 3. Hochster HS. Newer approaches to gemcitabine-based therapy of pancreatic cancer: fixed-dose-rate infusion and novel agents. Int J Radiat Oncol Biol Phys 2003;56:24 –30. 4. Vansteenkiste JF, Vandebroek JE, Nackaerts KL, et al. Clinicalbenefit response in advanced non-small-cell lung cancer: A multicentre prospective randomised phase III study of single agent gemcitabine versus cisplatin-vindesine. Ann Oncol 2001;12:1221–30. 5. Heinemann V. Role of gemcitabine in the treatment of advanced and metastatic breast cancer. Oncology 2003;64:191–206. 6. Dabrow MB, Francesco MR, Gilman PB, et al. Combined therapy with topotecan and gemcitabine in patients with inoperable or metastatic non-small cell lung cancer. Cancer Investig 2003;21:517–25. 7. Hussain SA, James ND. The systemic treatment of advanced and metastatic bladder cancer. Lancet Oncol 2003;4:489 –97. 8. Jordheim L, Galmarini CM, Dumontet C. Drug resistance to cytotoxic nucleoside analogues. Curr Drug Targets 2003;4:443– 60. 9. Krishnan P, Fu Q, Lam W, et al. Phosphorylation of pyrimidine deoxynucleoside analog diphosphates: selective phosphorylation of Lnucleoside analog diphosphates by 3-phosphoglycerate kinase. J Biol Chem 2002;277:5453–9. 10. Bouffard DY, Laliberte J, Momparler RL. Kinetic studies on 2⬘,2⬘difluorodeoxycytidine (Gemcitabine) with purified human deoxycytidine kinase and cytidine deaminase. Biochem Pharmacol 1993;45: 1857– 61. 11. Liou JY, Dutschman GE, Lam W, Jiang Z, Cheng YC. Characterization of human UMP/CMP kinase and its phosphorylation of D- and L-form deoxycytidine analogue monophosphates. Cancer Res 2002;62: 1624 –31. 12. Heinemann V, Xu YZ, Chubb S, et al. Inhibition of ribonucleotide reduction in CCRF-CEM cells by 2⬘,2⬘-difluorodeoxycytidine. Mol Pharmacol 1990;38:567–72. 13. Kim MY, Ives DH. Human deoxycytidine kinase: kinetic mechanism and end product regulation. Biochemistry 1989;28:9043–7. 14. Datta NS, Shewach DS, Mitchell BS, Fox IH. Kinetic properties and inhibition of human T lymphoblast deoxycytidine kinase. J Biol Chem 1989;264:9359 – 64. 15. Ross DD, Cuddy DP. Molecular effects of 2⬘,2⬘-difluorodeoxycytidine (Gemcitabine) on DNA replication in intact HL-60 cells. Biochem Pharmacol 1994;48:1619 –30. 16. Huang P, Chubb S, Hertel LW, Grindey GB, Plunkett W. Action of 2⬘,2⬘-difluorodeoxycytidine on DNA synthesis. Cancer Res 1991;51: 6110 –7. 17. Heinemann V, Schulz L, Issels RD, Plunkett W. Gemcitabine: a modulator of intracellular nucleotide and deoxynucleotide metabolism. Semin Oncol 1995;22:11– 8. 18. Ruiz van Haperen VW, Veerman G, Vermorken JB, Peters GJ. 2⬘,2⬘-Difluoro-deoxycytidine (gemcitabine) incorporation into RNA and DNA of tumour cell lines. Biochem Pharmacol 1993;46:762– 6. 19. Mackey JR, Mani RS, Selner M, et al. Functional nucleoside transporters are required for gemcitabine influx and manifestation of toxicity in cancer cell lines. Cancer Res 1998;58:4349 –57. 20. Dumontet C, Fabianowska-Majewska K, Mantincic D, et al. Common resistance mechanisms to deoxynucleoside analogues in variants of the human erythroleukaemic line K562. Br J Haematol 1999;106: 78 – 85.

Clinical Cancer Research 5621

21. Heinemann V, Xu YZ, Chubb S, et al. Cellular elimination of 2⬘,2⬘-difluorodeoxycytidine 5⬘-triphosphate: a mechanism of self-potentiation. Cancer Res 1992;52:533–9. 22. Lefebvre I, Perigaud C, Pompon A, et al. Mononucleoside phosphotriester derivatives with S-acyl-2-thioethyl bioreversible phosphateprotecting groups: intracellular delivery of 3⬘-azido-2⬘,3⬘-dideoxythymidine 5⬘-monophosphate. J Med Chem 1995;38:3941–50. 23. Galmarini CM, Thomas X, Calvo F, et al. Potential mechanisms of resistance to cytarabine in AML patients. Leuk Res 2002;26:621–9. 24. Mackey J, Baldwin SA, Young JD, Cass CE. Nucleoside transport and its significance for anticancer drug resistance. Drug Resist Updat 1998;1:310 –24. 25. Davidson JD, Ma L, Iverson PW, et al. Human multidrug resistance protein 5 (MRP5) confers resistance to gemcitabine. Proc Am Assoc Cancer Res 2002;43:780. 26. Bianchi V, Spychala J. Mammalian 5⬘-nucleotidases. J Biol Chem 2003;278:46195– 8. 27. Galmarini CM, Clarke ML, Santos CL, et al. Sensitization of ara-C-resistant lymphoma cells by a pronucleotide analogue. Int J Cancer 2003;107:149 –54. 28. Bergman AM, Pinedo HM, Peters GJ. Determinants of resistance to 2⬘,2⬘-difluorodeoxycytidine (gemcitabine). Drug Resist Updat 2002;5: 19 –33. 29. Bergman AM, Pinedo HM, Jongsma AP, et al. Decreased resistance to gemcitabine (2⬘,2⬘-difluorodeoxycitidine) of cytosine arabinosideresistant myeloblastic murine and rat leukemia cell lines: role of altered activity and substrate specificity of deoxycytidine kinase. Biochem Pharmacol 1999;57:397– 406. 30. Mansson E, Spasokoukotskaja T, Sallstrom J, Eriksson S, Albertioni F. Molecular and biochemical mechanisms of fludarabine and cladribine resistance in a human promyelocytic cell line. Cancer Res 1999;59:5956 – 63. 31. Mansson E, Flordal E, Liliemark J, et al. Down-regulation of deoxycytidine kinase in human leukemic cell lines resistant to cladribine and clofarabine and increased ribonucleotide reductase activity contributes to fludarabine resistance. Biochem Pharmacol 2003;65:237– 47. 32. Orr RM, Talbot DC, Aherne WG, et al. 2⬘-Deoxycytidine kinase deficiency is a major determinant of 2-chloro-2⬘-deoxyadenosine resistance in lymphoid cell lines. Clin Cancer Res 1995;1:391– 8. 33. Obata T, Endo Y, Tanaka M, Matsuda A, Sasaki T. Development and biochemical characterization of a 2⬘-C-cyano-2⬘-deoxy-1-beta-Darabino-pentofuranosylcytosine (CNDAC)-resistant variant of the human fibrosarcoma cell line HT-1080. Cancer Lett 1998;123:53– 61. 34. Ruiz van Haperen VW, Veerman G, Eriksson S, et al. Development and molecular characterization of a 2⬘,2⬘-difluorodeoxycytidine-resistant variant of the human ovarian carcinoma cell line A2780. Cancer Res 1994;54:4138 – 43. 35. van Bree C, Castro Kreder N, Loves WJ, et al. Sensitivity to ionizing radiation and chemotherapeutic agents in gemcitabine-resistant human tumor cell lines. Int J Radiat Oncol Biol Phys 2002;54:237– 44. 36. Groschel B, Cinatl J, Perigaud C, et al. S-Acyl-2-thioethyl (SATE) pronucleotides are potent inhibitors of HIV-1 replication in T-lymphoid cells cross-resistant to deoxycytidine and thymidine analogs. Antiviral Res 2002;53:143–52. 37. Groschel B, Kaufmann A, Hover G, et al. 3⬘-Azido-2⬘,3⬘-dideoxythymidine induced deficiency of thymidine kinases 1, 2 and deoxycytidine kinase in H9 T-lymphoid cells. Biochem Pharmacol 2002;64: 239 – 46. 38. Funato T, Satou J, Nishiyama Y, et al. In vitro leukemia cell models of Ara-C resistance. Leuk Res 2000;24:535– 41. 39. Bai L, Yamaguchi M, Tatsumi M, Kon K, Brautigam M. Mechanisms responsible for resistance of sublines derived from leukemia cell

lines to an antitumor agent 9-beta-D-arabinofuranosyl-2-fluoroadenine. J Cancer Res Clin Oncol 1998;124:367–73. 40. Grant S, Turner A, Nelms P, Yanovich S. Characterization of a multidrug resistant human erythroleukemia cell line (K562) exhibiting spontaneous resistance to 1-beta-D-arabinofuranosylcytosine. Leukemia 1995;9:808 –14. 41. Stegmann AP, Honders MW, Kester MG, Landegent JE, Willemze R. Role of deoxycytidine kinase in an in vitro model for AraC- and DAC-resistance: substrate-enzyme interactions with deoxycytidine, 1-beta-D-arabinofuranosylcytosine and 5-aza-2⬘-deoxycytidine. Leukemia 1993;7:1005–11. 42. Galmarini CM, Thomas X, Graham K, et al. Deoxycytidine kinase and cN-II nucleotidase expression in blast cells predict survival in acute myeloid leukaemia patients treated with cytarabine. Br J Haematol 2003;122:53– 60. 43. Kakihara T, Fukuda T, Tanaka A, et al. Expression of deoxycytidine kinase (dCK) gene in leukemic cells in childhood: decreased expression of dCK gene in relapsed leukemia. Leuk Lymphoma 1998; 31:405–9. 44. Galmarini CM, Graham K, Thomas X, et al. Expression of high Km 5⬘-nucleotidase in leukemic blasts is an independent prognostic factor in adults with acute myeloid leukemia. Blood 2001;98:1922– 6. 45. Mazzon C, Rampazzo C, Scaini MC, et al. Cytosolic and mitochondrial deoxyribonucleotidases: activity with substrate analogs, inhibitors and implications for therapy. Biochem Pharmacol 2003;66:471–9. 46. Kobayashi T, Kakihara T, Uchiyama M, et al. Low expression of the deoxycytidine kinase (dCK) gene in a 1-beta-D-arabinofuranosylcytosine-resistant human leukemic cell line KY-Ra. Leuk Lymphoma 1994;15:503–5. 47. Leegwater PA, De Abreu RA, Albertioni F. Analysis of DNA methylation of the 5⬘ region of the deoxycytidine kinase gene in CCRFCEM-sensitive and cladribine (CdA)- and 2-chloro-2⬘-arabino-fluoro2⬘-deoxyadenosine (CAFdA)-resistant cells. Cancer Lett 1998;130: 169 –73. 48. Antonsson BE, Avramis VI, Nyce J, Holcenberg JS. Effect of 5-azacytidine and congeners on DNA methylation and expression of deoxycytidine kinase in the human lymphoid cell lines CCRF/CEM/0 and CCRF/CEM/dCk-1. Cancer Res 1987;47:3672– 8. 49. Flasshove M, Strumberg D, Ayscue L, et al. Structural analysis of the deoxycytidine kinase gene in patients with acute myeloid leukemia and resistance to cytosine arabinoside. Leukemia 1994;8:780 –5. 50. Jordheim L, Cros E, Galmarini CM, van Bree C, Dumontet C. Deoxycytidine kinase gene alterations in gemcitabine-resistant cell lines. Proc Am Assoc Cancer Res 2004;45:618. 51. Csapo Z, Sasvari-Szekely M, Spasokoukotskaja T, Staub M. Modulation of human deoxycytidine kinase activity as a response to cellular stress induced by NaF. Acta Biochim Pol 2001;48:251– 6. 52. Csapo Z, Sasvari-Szekely M, Spasokoukotskaja T, et al. Activation of deoxycytidine kinase by inhibition of DNA synthesis in human lymphocytes. Biochem Pharmacol 2001;61:191–7. 53. Elgie AW, Sargent JM, Alton P, et al. Modulation of resistance to ara-C by bryostatin in fresh blast cells from patients with AML. Leuk Res 1998;22:373– 8. 54. Spasokoukotskaja T, Sasvari-Szekely M, Keszler G, et al. Treatment of normal and malignant cells with nucleoside analogues and etoposide enhances deoxycytidine kinase activity. Eur J Cancer 1999; 35:1862–7. 55. Sargent JM, Elgie AW, Williamson CJ, Lewandowicz GM, Taylor CG. Circumvention of ara-C resistance by aphidicolin in blast cells from patients with AML. Br J Cancer 2001;84:680 –5. 56. Chan TC. Augmentation of 1-beta-D-arabinofuranosylcytosine cytotoxicity in human tumor cells by inhibiting drug efflux. Cancer Res 1989;49:2656 – 60.