Difluorodeoxyguanosine in Chinese Hamster Ovary Cells'

Home , Ribonucleotide reductase

[CANCER RESEARCH55, 1517-1524, April 1, 19951 Cytotoxicity, Metabolism, and Mechanisms of Action of 2',2'- Difluorodeoxyguanosine in Chinese Hamster Ovary Cells'

Varsha Gandhi,2 Shin Mineishi, Peng Huang, Amy J. Chapman, Yandan Yang, Feng Chen, Biffie Nowak, Shern Chubb, Larry W. Hertel, and William Plunkett Department of Clinical Investigation, The University of Texas M. D. Anderson Cancer Center, Houston, Texas 77030 (V. G., S. M., P. H.. A. J. C.. Y. 1'., F. C., B. N., S. C., w. p.]. andLillyResearchLaboratories,Indianapolis,indiana46285(L W.H.)

ABSTRACT treatment of hematological malignancies (4â€”6).Arabinosyladenine, although not successful in cancer therapy, has proven antiviral activity The emerging clinical success of gemcItabine (2',2'-difluorodeoxycytl dine) Stimulated interest in the synthesis and evaluation of purine conge (7). The importance of the 2'-pentose moiety for the design of chemo ners. The cytotoxicity, metabolism, and mechanisms of action of the lead therapeutic agents and the fact that fluorine has a van der Waals radius candidate, 2',2'-difluorodeoxyguanoslne (dFdGuo), were StUdied In ChI nese hamster ovary cells Unlike the natural nucleoside deoxyguanosine similar to that of hydrogen led to the synthesis and evaluation of (dGuo), dFdGuo was not a substrate for purine nucleoside phosphorylase 2'-fluorinated analogues. 2'-Deoxy-2'-fluorocytidine and 2'-deoxy Wild-type Chinese hamster ovary cells and a mutant line deficient in 2'-fluoroguanosine possess inhibitory activities against several deoxycytidine (dCyd) kinase were similarly affected by dFdGuo (50% lymphoid cell lines and influenza viruses, respectively (8, 9). Inter inhibitory concentration, 7.5 and 6.5 pM, respectively), suggesting that estingly, substitution of the 2'-hydrogen of dCyd3 with a fluorine unlike gemcitabine, dCyd kinase was not responsible for activation of atom in the arabinose configuration produced 10-fold greater cyto dFdGuo This was further confirmed by separation of nucleoside kinases toxicity than when placed in the ribose position (10). 2'-Fluoroarabi (adenosine kinase, dGuo ldnase, and dCyd kinase) of Chinese hamster nosylguanine is active against human leukemia cells in vitro (11). ovary cells on DEAE-cellulose column chromatography. The kinase ac Recently, substitution of both hydrogens of the 2'-carbon of dCyd tivity that phosphorylated dGuo also converted dFdGuo to its monophos. with fluorines generated dFdCyd (12), which is currently being phate, suggesting that dGuo kinase activated dFdGuo. Consistent with this result, coincubation with dGuo spared the dFdGuo-mediated toxicity; evaluated in the clinic under the generic name gemcitabine. however, addition of up to 10 mM dCyd did not reverse the toxicity of Comparative studies between ara-C and dFdCyd demonstrated that dFdGuo. Intracellularly, dFdGuo was phosphorylated to its mono-, di-, the difluoro anomer is transported more rapidly into cells and is a and triphosphates; dFdGuo triphosphate (dFdGTP) was the major me preferred substrate for phosphorylation to the active 5'-triphosphate, tabolite and accumulated to 45 jiM after a 6-h Incubation with 30 jaM which is relatively stable in cells (i3). Additionally, unlike ara-C, the dFdGuo. The elimination of dFdGTP was monophasic with a tÂ½ofabout diphosphate of dFdCyd inhibited ribonucleotide reductase, an action 6 h. Deoxynucleotides were decreased in cells Incubated with dFdGuo, that decreased deoxynucleotide pools and was the key to the suggesting that ribonudeotide reductase was Inhibited. dAT?, which de self-potentiation of the analogue (14â€”16). creased 78% after a 4-h Incubation with 30 jaM dFdGuo, was most This multitude of actions of gemcitabine metabolites in exerting affected@dFdGuo was a potent inhibitor of DNA synthesis. Extension of a DNA printer over a defined template in the presence of dFdGTP revealed cytotoxicity and its emerging clinical success (17) stimulated interest that dFdGTP was a good substrate for incorporation opposite C sites of in the synthesis and evaluation of purine congeners (18, 19). Meta the template by DNA polymerase a dFdGTP incorporation caused DNA bolic inactivation and growth inhibition studies in a variety of tumor polymerase a to pause after the polymerization of one additional de lines suggested that dFdGuo (Fig. 1) was the lead analogue among the oxynucleotide. This pattern oflnhlbltlon, which is shared by gemcltabine, purine series (18, 19). Using CHO cells, we investigated the cytotox distinguishes 2',2'-difluoronucleosldes from arabinosylnudeosides which icity, metabolism, and mechanisms of action of dFdGuo. Emphasis is halt primer extension at the Incorporation slte dGTP competed effectively placed on the enzymes likely to be involved in the catabolism (purine with dFdGTP for incorporation by DNA polymerase a The unique nucleoside phosphorylase) and anabolism (nucleoside kinases) of activation requirements and patterns of Inhibition of DNA synthesis din dFdGuo and on ribonucleotide reductase and DNA polymerase, which tinguish this promising new antimetaboilte from other nucleoside by analogy with dFdCyd are likely targets of dFdGuo nucleotides. The analogues. results are compared with those of dFdCyd and the base congener ara-G. INTRODUCTION MATERIALS AND METHODS The chemical constituents and their configuration at the 2'-position of purine and pynmidine nucleotides are critical determinants of the Chemicals. dFdGuo, its 5'-triphosphate, dFdGTP, and dFdlno were syn metabolic fate and therapeutic activity of these nucleic acid precur thesized at the Lilly Research Laboratories, Indianapolis, IN. [8-3HJdFdGuo sors. A hydroxyl group in the arabinose configuration results in the (specific activity, 6 Ci/mmol) was prepared by Amersham Radiochemicals, incorporation of these arabinosyl nucleosides into DNA and inhibition Arlington, IL. Bovine spleen purine nucleoside phosphorylase was purchased of further DNA synthesis, leading to cell death (1â€”3).Arabinosyl from Sigma Chemical Co., St. Louis, MO, as were natural nucleosides and analogues of both purine and pyrimidine are established agents in the nucleotides. All other chemicals were of the highest purity available. For nucleoside kinase assays, the natural nucleosides [2,8-3H]adenosine(39 Ci/mmol), [5-3H]dCyd (26 Ci/mmol), and [8-3H]dGuo (7 Ci/mmol) were Received 11/29/94; accepted 2/3/95. Thecostsof publicationofthisarticleweredefrayedinpartby thepaymentofpage charges. This article must therefore be hereby marked advertisement in accordance with 3 The abbreviations used are: dCyd, deoxycitidine; dAdo, deoxyadenosine; ara-C and 18 U.S.C. Section 1734 solely to indicate this fact. ara-cm', 1-@-t-arabinofuranosylcytosineandits 5'-triphosphate;ara-Gand ara-GTP, 1 Supported in part by Grants CA 28596 and CA 57629 from the National Cancer i-@-D-arabinofuranosylguanine and its 5'-triphosphate; dFdCyd, dFdCMP, dFdCDP, and Institute, Department of Health and Human Services, and Grant DHP-1 from the Amer dFdCFP, 2',2'-difluorodeoxycytidine (gemcitabine) and its 5'-mono-, -di-, and -triphos icanCancerSociety. phate;dFdGuo,dFdGMP,dFdGDP,anddFdGTP,2',2'-difluorodeoxyguanosineandits 2 To whom requests for reprints should be addressed, at Department of Clinical 5'-mono-, -di-, and -triphosphate; dGuo, deoxyguanosine; dFdlno, difluorodeoxyinosine; Investigation, Box 52, The University of Texas M. D. Anderson Cancer Center, Houston, PNP, purine nucleoside phosphorylase; t@, half-life of elimination; dNTP, deoxynucle TX 77030. otide triphosphate; CHO, Chinese hamster ovary; pol a, polymerase a. 1517

purchased from ICN Biochemicals (Costa Mesa, CA). All nucleosides were 0 purified by reverse-phase HPLC to greater than 99% purity. Cell CUltUre and Cytotoxicity. CHO cell lines were maintained in loga rithmic growth on 60-mm plastic dishes or roller bottles in McCoy's 5A H2N@5@> medium with 20% horse serum (GIBCO, Grand Island, NY). Under these conditions, the cell population-doubling time was 12 h. The cell line aC'@7, HO@ Ho@c::@ kindly supplied by Priscilla P. Saunders (20), is a CHO line deficient in dCyd kinase. In several experiments, cells were trypsinized, washed into calcium free McCoy's 5A medium, and grown in spinner flasks for 24 to 36 h before OH F use. Clonogenicity after drug treatment was determined by two methods. In the Deoxyguanoslrss Difluorodooxyguano first, cells were seeded on plates, treated as indicated, washed into fresh (dGuo) (dFdG) medium, and incubated for 5 days to allow colony formation. In the second procedure, cells were incubated with drugs in suspension culture. At the Fig. 1. Chemical structures of dGuo and dFdOuo (dFdG). indicated times portions of each culture were removed, washed into drug-free medium, plated in multiple dilutions at densities of 100â€”1000cells/plate,and incubated for 7â€”8days. Colonies were fixed and stained with methanol calculation assumes that nucleotides are uniformly distributed in total cell containing 0.5% crystal violet. The mean of at least 3 plates was determined; water. SD values were less than 10% of the mean in every case. Determinations ofdNTP. CHO cells incubated with 10 or 30 g@MdFdGuo Phosphorolysis of dFdGuo. The activity of PNP was determined by for 3 h were extracted by 60% methanol. The DNA polymerase assay as spectrophotometric assay. Normal substrates guanosine and inosine at 10, 20, modified by Sherman and Fyfe (23) was used to quantitate dNTPs in the cell 50, 100, and 200 fLMwere incubated with 0.01 unit (1 unit = amount required extracts. DNA polymerase I (United States Biochemical Corporation, Cleve for phosphorolysis of 1 @molofdGuo per mm) of purified PNP from bovine land, OH) was used to start a reaction in a mixture that contained 100 m@ spleen (Sigma) for up to 10 mm. The reaction conditions were as suggested by HEPES buffer (pH 7.3), 10 mMMgC12,7.5 @.tgBSA,and synthetic oligonu the manufacturers. dFdGuo (100 @M)anddFdlno (50 @.tM)wereincubatedwith cleotides of defined sequences as templates annealed to a primer, [3H]dATPor 1 unit of enzyme. Xanthine oxidase (0.1 unit; Sigma) was added to the reaction [3H]dTrP, and either standard dNTP or the extract from 1 or 2 X 10@CHO mixtures containing inosine and dFdlno for liberation of uric acid. The prod cells before and after incubation with dFdGuo. Reactants were incubated for 1 ucts of reactions with dFdGuo and dFdlno were also analyzed by reverse-phase h and applied to filter discs; after a washing, the radioactivity on the discs was HPLC. determined by liquid scintillation counting. Reactions conducted using a syn Phosphorylation of dFdGuo by Nucleoside Kinases. Nucleoside kinase thetic oligonucleotide that did not require dGTP as a substrate demonstrated activities were determined in extracts of wild-type CHO cells. Cells harvested that addition of dFdGTP (30, 50, or 100 ,.@M)didnot affect the DNA polym from roller bottle cultures were suspended in 2 volumes of buffer A (50 m@i erase I activity. Hence the dFdGTP present in the extracts from CHO cells K2HPO4, pH 7.5, containing 10 mM 2-mercaptoethanol). The protease inhibitor would not interfere with the dNTP assay. phenylmethylsulfonyl fluoride was added at a final concentration of 250 @M Inhibition of DNA and RNA Synthesis. CHO cells (2 X 10's) in expo and the cells were then disrupted by bomb cavitation for 30 min at 1000 psi of nential growth phase were incubated with various concentrations of dFdGuo nitrogen. After centrifugation for 20 mm at 11,000 X g at 4Â°C,theextract was for the desired period before addition of 0.5 @Ci[3H]thymidineor [3H]uridine dialyzed against buffer A and applied to a column (1 x 15 cm) of DEAE for 30 min to determine DNA and RNA synthesis, respectively. The cells were cellulose (Whatman DE52) that had previously been equilibrated with buffer collected on a 25-mm glass fiber disc (prewetted with 1% sodium PP@)by A. The column was eluted with a 300-mI linear gradient of KC1(0â€”600mM) filtration and then washed twice with 4 ml of ice-cold 0.4 N HCIO4 and twice in 50 m@ Tris-HC1, pH 7.5, containing 10 mM 2-mercaptoethanol. Fractions of with 2 ml of ethanol. Radioactivity retained on the filter disc was determined 3 ml were collected, and every third fraction was assayed for its ability to by liquid scintillation counting. phosphorylate the indicated nucleosides. DNA Primer Extension Assay A 17-base oligodeoxynucleotide obtained The phosphorylation activities of natural nucleosides and dFdGuo were from Pharmacia LKB Biotechnology, Inc. (Piscataway, NJ), was labeled with determined by the anion exchange filter disc method (21). The reaction mixture 32pat the 5'-end, annealed to its complementary site on a defined sequence for each nucleoside substrate contained 50 mM Tris-HC1 (pH 7.8), 5 mM template (Genosys Biotechnologies, Inc., Woodlands, TX), and precipitated MgCl2, 8 m@iAlP, and a final concentration of 25 pM tritiated nucleoside with ethanol as described before (3). This template/primer hybrid substrate in a 96-well microtiter plate that was floating in a 37Â°Cwaterbath. 5'-.[32P ]GTAAAACGACGGCCAGT (1) An equal volume of each fraction from the DEAE-cellulose eluate was added 3 p CATTTTGCTGCCGGTCA( to start the reaction. After 60 mm (dCyd, adenosine) or 180 mm (all other substrates), 35 ,.tl of the reaction mixture were applied to a 14-mm disc of was used as a substrate for DNA p01a. The initial six nucleotides (sites 18â€”23) DE81 paper (Whatman) and washed with 1 mM ammonium formate (10 incorporated in the primer were alternate dCMPs and dAMPs. In this way p01 ml/disc). After successive washes in water and 70% ethanol, the discs were a had a â€œrunningstartâ€•(24) before it encountered the Csite on the template, dried and the radioactivity was quantitated by liquid scintillation counting. The where dFdGTP, ara-GTP, or dGTP were incorporated. activity was expressed as pmol product formed/mg of protein/h. Protein The primerextensionreactionmixturecontained20 mMTriS-HC1(pH7.4),03 content was determined with Bradford assay dye (Bio-Rad Laboratories, mM MgCI2, 0.25 mM DTF, 40 pg/mi BSA, indicated concentrations ofdNTPs and Richmond, CA) using bovine plasma y-globulin as a standard. dFdGTP or ara-GTP along with the labeled,primer/templatecomplexes(5), and Metabolism ofdFdGuo. Exponentially growing cells were incubated with DNA p01 a. The enzyme was purified from exponentially growing 010 cells as indicated concentrations of [3H]dFdGuo (specific activity, 30 g.@Ci/@.tmol)for 6 describedpreviously(25).The reactionswere carriedout at 37Â°Cfor10-20 mm, h, and samples were removed every h to determine the accumulation of stopped by addition ofan equal volume ofSO mM EDTA with 03% bromophenol dFdGTP. After nucleotides were extracted using perchloric acid (22), dFdGTP blue in 90% deionizedformamide,and analyzedby electrophoresisthrougha 10% was separated from natural nucleotides by HPLC using an anion exchange polyacrylamide sequencing gel. A control reaction (with 30 @MdNTPs) done Partisil-iO SAX column (Whatman, Clifton, NJ) and a gradient elution scheme under similar conditions for up to 30 mm demOnstrated that the reaction was linear as described before (22). The radioactivity associated with the dFdGTP peak with time (r = 0.99), and by 30 mm less than 20% of the primers were elongated was quantitated in at least two separate experiments by scintillation counter. beyond the 0 incorporation site. The identity of dFdGTP was confirmed by its coelution with authentic triphos Various concentrations of dGTP (0.005, 0.01, 0.025, 0.05, 0.075, 0.1, 0.25, phate, resistance to periodate oxidation, and UV absorbance characteristics. and 0.5 pM), dFdGTP, or ara-GTP (0.1, 0.3, 0.5, 1.0, 3.0, 5.0, and 10.0 @M) Quantitationof the nucleotides in HCIO4extracts was determined by electronic were used to determine the kinetic parameters. Radioactivity associated with integration and reference to external standards. The intracellular concentration the band at the 0 site (site 24 on the extended primer) and in all the bands of nucleotides was calculated by dividing the quantity of nucleotides per cell beyond this site (sites 25â€”31)wasquantitated using a Betascope 603 blot by the number of cells required to replace 1 ml volume of water. This analyzer (Betagen, Waltham, MA). The radioactivity was normalized based on 1518

(Fig. 3). Incubations with 1, 10, and 100 p.MdFdGuo for 4 h resulted in about 20, 45, and >80% cell kill, respectively. This loss of clono genicity was not reversed by coincubation with dCyd (Fig. 3). These results indicated that dCyd kinase does not appear to be responsible for the phosphorylation of dFdGuo. This notion was supported by the data showing that wild-type CHO cells and aC'@7 (deficient in dCyd .5 kinase) were equally sensitive to dFdGuo; 50% inhibitory concentra C tions were 7.5 and 6.5 p.M, respectively (data not shown). The obser 0 C.) vation that 20 @LMdGuospared CHO cells from the cytotoxicity of dFdGuo (data not shown) indicated involvement of dGuo kinase in the phosphorylation of dFdGuo. U Nucleoside Kinase Activity in Cell Extracts. To identify the enzyme activity responsible for the phosphorylation of dFdGuo, major nucleoside kinases of CHO cells were separated. Fig. 4A shows a I typical elution profile of kinase activities from cell extracts fraction ated by DEAE-cellulose chromatography. adenosine kinase eluted first from the column. The activity that phosphorylated dGuo was eluted as a single peak, termed dGuo kinase. This peak was distinct and separable from the peak that eluted at the highest KCI concen 2 3 4 trations, which phosphorylated dCyd. The same fractions from the Hours DEAE column were then used to identify the activities responsible for Fig. 2. Cytotoxicity of dFdGuo to CHO cells. Cells growing in suspension cultures the phosphorylation of dFdGuo (Fig. 4B). dFdGuo (used at a conc were incubated with either 10 (A) or 30 (â€¢) @MdFdGuo.At the indicated times, portions of each culture were washed into drug-free medium and plated for colony formation. centration of 25 @LM)wasphosphorylated exclusively by the dGuo Results are expressed as percentage of control (untreated) cells. Points are the means of kinase peak. Addition of 8-aminoguanosine in the reaction mixture to three determinations. inhibit purine nucleoside phosphorylase did not change this pattern (data not shown). dGTP (50 SM), a potent inhibitor of dGuo kinase, the average total radioactivity in each lane and the background value (counts almost completely inhibited phosphorylation of dFdGuo and dGuo in the bands at the 0 site and beyond, in a reaction lacking dGTP) was (data not shown). The affinity of dGuo kinase for dFdGuo as a subtracted. These data (reaction velocity) were plotted against the concentra substrate was analyzed by using the peak fraction (fraction 23; see tion of the substrate. The apparent Km values for incorporation were then Fig. 4) of dGuo kinase and various concentrations of dFdGuo. Kinetic calculated based on the Michaelis-Menten equation, using a computer-assisted studies showed an apparent Km of 40 @Mwitha Vmas of 12 nmol/ program (26). The apparent V,,,@,,valuesobtained through these calculations mg/h. were converted to percentage of product/min based on the total radioactivity in Intracellular Accumulation and Retention of dFdGTP. The ac each lane. To determine the inhibition of DNA primer elongation by analogues cumulation of dFdGTP was quantitated after incubation of CHO cells in the presence of dGTP, the radioactivity above the pause site (bands at sites with 10 and 30 @.LM[3H@dFdGuo,acid extraction, and separation on anion 26â€”31 for dFdGTP or at sites 25â€”31for ara-GTP) was quantitated and exchange HPLC. The levels of dFdGuo and its mono-, di-, and triphos compared with the control lane (reaction containing all four dNTP). The final values represent a mean Â±SDof at least four sets of data points generated by phates were 22, 6, 8, and 63% of total radioactivity. This distribution two separate reactions each run twice by gel electrophoresis.

RESULTS 100 Purine Nucleoside Phosphoryla.se. The substrate specificity of purine nucleoside phosphorylase was determined using guanosine, inosine, dFdGuo, and dFdlno. The Km values of purine nucleoside phosphorylase for guanosine and inosine were 34 and 37 p.M,respec 0 C tively, values that compare favorably with the literature (27). The 0 C.) Vm@ (p.moles/min/0.1 unit) was 0.26 and 0.17 for guanosine and â€˜B inosine, respectively. Phosphorolysis of up to 100 p.M dFdGuo and dFdlno incubated with 1 unit of enzyme was not detectable by our assay. The lower limit of detection was 5 pmol/min/0.1 unit. HPLC .2 analysis of the products did not reveal any base liberation after a 15-mm reaction. Effect of dFdGuo on CHO Cell Clonogemclty. Wild-type CHO I cells growing in suspension cultures were incubated in the absence of drug (control) or with 10 or 30 @MdFdGuofor up to 6 h. Every h cells were removed and plated for colony formation. After a 6-h incubation with 10 @LMdFdGuo,70% of the cells had lost clonogenic capacity ,. I I I t (Fig. 2). The effect was greater with 30 @.tMdFdGuo,resulting in only 0 â€˜@ 1 10 100 1000 10000 10% survival after a 6-h incubation. To determine whether dFdGuo Deoxycytidine, @M mediated toxicity could be reversed by coincubation with dCyd, cells Fig. 3. Effect of dCyd on the cytotoxicity of dFdGuo. CHO cells growing as monolayer were incubated for 4 h with 0, 1, 10, or 100 p@MdFdGuo in the were exposed to 0 (â€¢),i ( â€¢),10(A), or 100 (U) @u.tdFdGuoin the presence of various concentrations of dCyd. After 4 h, cells were washed free of nucleosides, trypsinized, and presence of 1â€”10mM dCyd. Controls (without dFdGuo incubation) plated for colony formation in fresh medium. Results are expressed as percentage of indicated that up to 10 m@tdCyd was not cytotoxic to the CHO cells control (untreated) cells. Points are the means of three determinations. 1519

Effect of dFdGuo on Deoxynucleotide Pools. Gemcitabine is a 600 potent inhibitor of ribonucleotide reductase, an action that decreases dNTP pools in cells in culture. To determine whether dFdGuo inca C bation has a similar effect on deoxynucleotide pools, dNTPs were 0 â€˜B compared in control and dFdGuo-treated cells (Table 1). During

@ 0 400 incubation with 30 p.MdFdGuo, there was a time-dependent decrease .c 0. @l in the dATP pool; less than 15% of control amounts of dATP re .c 2 mainedafter3h.Incubationswith10p.MdFdGuoalsoresultedina 0. 0 @. similar patternof inhibition; the dATP pool was reducedto 70 and 3 C, .@. 85% of the controlby 3 and 4 h, respectively(data not shown).The 200 @8levelofthedCrPpoolwasalsoreduced,butthiswasnotasmarked .5

E 0 0 E 40 C

30 C 0 â€˜B a:

20. @20

â€˜C 0. C, @0 IL .5 10

0 1 2 3 4 5 6 0 10 20 30 40 50 60 70 Hours Fractions Fig. 5. Accumulation of dFdGTP by CHO cells. Cells growing in suspension culture Fig. 4. Nucleoside kinase activity of CEO cell extracts by DEAE-cellulose chroma were incubated with either 10 (A) or 30 (â€¢) @t@i[3H]dFdGuo.At the indicated times, tography. Cell extract was applied to a column of DEAE-cellulose. Elution was with a portions of each culture were harvested and nucleotide analysis was conducted as 300-mi linear gradient of KG. Each third fraction was assayed for phosphorylation of(A) described in â€œMaterialsandMethods.â€•Pointsare the means of two separate experiments natural nucleosides adenosine (A), dGuo (â€¢),and dCyd (U) or (B) dFdGuo (dFdG). with SD of less than 15% of the mean. Points are from a single experiment, which is representative of three similar experiments.

suggests that the conversion of dFdGuo to its monophosphate in the cells 300 was rate limiting to the accumulation of dFdGTP, a presumed active metabolite. The concentration of dFdGTP was between 10 and 15 p.M after a 1-h incubation with 10 and 30 p.M dFdGuo, respectively. At these 100 exogenous concentrations of dFdGuo, the accumulation of triphosphate was not linear with time; however, there was a gradual increase in the a: I- triphosphate which reached to approximately 30 and 45 p.M,respectively, C, 30 A after a 6-h incubation with the drug (Fig. 5). The accumulation of .5 dFdGTP was dose dependent; a 2-h incubation with 100 and 300 p.Mof dFdGuo resulted in 80 and 160 p.r@iintracellular dFdGTP (data not A shown). 10 To determine the elimination kinetics of dFdGTP, cells were incu bated with 10 p.MdFdGuo for 3 h and washed into drug-free medium, 3 and the cellular concentration of dFdGTP was determined as indicated (Fig. 6). dFdGTP elimination was monophasic with a tÂ½of4.2 h. To

evaluate whether higher intracellular concentrations of dFdGTP ,@1' would be eliminated with a different pattern, as occurs for dFdCTP 0 5 10 15 (28), cells were incubated with 200 p.MdFdGuo for 3 h to accumulate Hours >100 p.M dFdGTP. The elimination remained monophasic at this Fig. 6. Retention of dFdGTP by CHO cells. Cells incubated with [3H)dFdOuo [10 ga@i intracellular dFdGTP level, but the rate was slightly slower (tÂ½ 6.4 (A) or 200 @LM(â€¢)]for 3 h were washed into drug-free medium. At the indicated limes, portions of each culture were harvested and extracted and the cellular concentration of h) than that observed at the lower cellular dFdGTP concentration dFdGTP was determined as described in â€œMaterialsandMethods.â€•Pointsare the mean (Fig. 6). of twoseparateexperimentswithSDof lessthan15%. 1520

Table 1 Effect of dFdGuo incubation on deoxynucleotides exonuclease activity associated with this enzyme, the polymerization 010 cells were incubated with 30 gxMdFdGuo for the indicated times. The cells were process would represent solely the action of analogue incorporation. methanol extracted and dNTP pools were quantitated as described under â€œMaterialsand The extension of a DNA primer was studied in the absence of a Methods.â€•Datapoints represent the mean Â±SDof three separate experiments, each done in duplicate. competing natural deoxynucleotide (dGTP) and was compared with Deoxynucleotides (@tM) that of ara-GTP as shown in Fig. 8. Lane 1 represents the reaction mixture containing no analogues. The intense band at the pre-G site Time(h)cIATPdCI'PdGTPdTFP017.1 (site 23) indicated that pol a paused before the 0 site (site 24) because 6.3111.5 Â±5.639.1 Â±12.33.1 Â±1.632.1 Â± no dGTP was available for polymerization. The faint band at the 0 4.724.2 Â±3.518.3 Â±4.75.1 Â±2.427.7 Â± 0.732.3 Â±0.624.6 Â±3.36.6 Â±3.540.6 Â± site represents a small amount of noncognate dNTP incorporated into 15.642.4 Â±0.223.1 Â±18.64.9 Â±0.941.6 Â± this site; this may permit some further extension as is evident from the Â±0.925.1 Â±7.76.2 Â±0.551.2 Â±8.1 faint full length product band. When increasing concentrations of dFdGTP were included in the absence of dGTP (Lanes 2â€”8),the density of the band at the pre-G site (site 23) decreased in inverse proportion to the dFdGTP concentration, indicating that dFdGTP can serve as an alternate substrate for incorporation into the G site. Although some primers were elongated by pol a to make 29- to 31-mer products, the incorporation of dFdGTP at the 0 site resulted in a major pause at the site one nucleotide after the 0 site (site 25). If the nucleotide dTTP, which is necessary for extension beyond the 0 @30 site, was omitted, the pause site shifted to the 0 site (data not shown). â€˜B This analysis confirms the observation that the penultimate incorpo S ration of dFdGMP requires insertion of next normal and correct dNTP (dTTP for the present sequence) for the characteristic chain termina tion pattern. When this pattern was compared with the same concen @â€˜10 trations of ara-GTP in the reaction mixture (Lanes 9â€”15), three z differences were evident: (a) ara-GTP appeared to be incorporated more efficiently as indicated by the lighter band at the pre-G site compared to that at the same concentrations of dFdGTP; (b) the incorporation of ara-GTP at the 0 site strongly impaired the polym erization process by pol a at the site of incorporation as indicated by the intense band at the 0 site (site 24). In contrast, dFdGTP incorpo 0 10 20 30 40 50 ration lead to a pause after the incorporation of one additional nude otide; (c) ara-GTP was a more potent chain terminator, as suggested dFdGTP, @tM by the relative lack of primers extending beyond the analogue Fig. 7. Relationship between dFdGTP and inhibition ofDNA and RNA synthesis. Cells incorporation site. were incubated with either 10 or 30 @MdFdGuofor 2, 4, and 6 h. For the last 30 mm, they were also incubated with [3H]thymidine(O) or [3H]uridine (A). The radioactivity incor The data obtained from this gel and similar gels were used to porated into the acid-insoluble material was determined as described in â€œMaterialsand calculate the kinetic parameters for incorporation. The natural nude Methods.â€•Pointsare the mean of two experiments (SD less than 10%) and are plotted otide, dGTP, was a preferred substrate for incorporation opposite the against the dFdGTP concentrations as determined before (see Fig. 5). C site (apparent Km 0.02 Â±0.01 p.M).The apparent Km values of pol a for incorporation of dFdGTP and ara-GTP were 1.37 Â±0.78 and 0.24 Â±0.11 p.M,respectively. The apparent velocities of incorporation as the effect on dATP. Interestingly, there was no significant effect on were similar for all three substrates; apparent Vmax calculated as the dGTP pool, and the cellular concentration of dTTP increased slightly. Inhibition of DNA and RNA Synthesis. DNA and RNA synthesis 31 â€” were measured as incorporation of [3H]thymidine and [3H]uridine in CHO cells incubated with 10 p.Mand 30 p.MdFdGuo for 2, 4, and 6 h. The data were plotted against the intracellular dFdGTP levels 25- achieved at these exogenous concentrations of dFdGuo (see Fig. 5). 24â€” . .@ . . There was a direct relationship between levels of dFdGTP and the 23-@Oâ€¢â€¢@â€¢â€¢â€¢SS.. extent of DNA synthesis inhibition (Fig. 7). At about 45 p.MdFdGTP, which was achieved after a 6-h incubation with 30 p.MdFdGuo, DNA S â€¢ â€¢ â€¢ â€¢ â€¢ . . - synthesis was inhibited by >95%, suggesting a potent inhibitory action associated with dFdGTP. At these levels of dFdGTP, however, RNA synthesis was not significantly affected (Fig. 7). 17 Action of dFdGTP on DNA Strand Elongation in Vitro. The I 2 3 4 5 5 7 8 1 10 11 12 13 14 15 inhibition of DNA synthesis by dFdGuo in intact CHO cells depends 0 I I on at least two factors: (a) incorporation of dFdGTP by DNA poly merases and its action on further polymerization; and (b) the compe Fig. 8. Autoradiogram of a polyacrylamide gel showing action of dFdGTP and ara GTP on DNA primer extension in vitro by DNA pol a. The ability of CHO DNA pol a tition between analogue triphosphate and endogenous natural de to extend the 32P-labeledDNA primer in the presence of 30 @LMdCTP,dTl'P, or dATP oxynucleotide for incorporation by DNA polymerase. An assay based and with varying concentrations of dFdGTP (Lanes 2â€”8)orara-GTP (Lanes 9â€”15).Lane 1, no analogues.The concentrationsofanaloguetriphosphatesis0.1 (Lanes 2 and 9), 0.3 on extension of a DNA primer by DNA pol a from CHO cells was (Lanes 3 and 10), 0.5 (Lanes 4 and 11), 1.0 (Lanes 5 and 12), 3.0 (Lanes 6 and 13), 5.0 used to characterize these two processes. Because there is no 3'â€”Ã·S'- (Lanes 7 and 14), and 10 p@M(Lanes 8 and 15). 1521

METABOLISMANDACtiONS OF DIFLUORODEOXYGUANOSINE

No analogs dFdGTP ara-GTP DISCUSSION I II ii The present studies involving metabolism and mechanisms of action of 31- !!!0 dFdGuo demonstrate the similarities of this agent with its base and caibohy drate analogues, ara-G and dFdCyti The enzymes involved in the metabo lism of dFdGuo make this congener a close kin to am-U having similar 25- activation properties. The unique actions of dFd0uo, including the target enzymes and pattern of inhibition ofDNA synthesis, however, are shared by @ 23â€”@â€¢ . â€” . its carbohydrate analogue, dFdCyd. Hence it is desirable to compare dFdGuo with both ara-G and dFdCycL A problem in the clinical use of analogues of dGuo is their degra * . * _ *@ dation to free bases by PNP. The 2'-deoxynucleoside of 6-thiogua nine, synthesized as a strategy to direct the base analogue to PNP, was evaluated in the clinic but was found to be inactivated by phospho rolysis (29). However, the resistance of several 2'-substituted inosines to phosphorolysis by PNP suggested that 2'-substituted analogues of I 2 3 4 5 6 7 8 9 10 guanosine would also be resistant to PNP (30). The in vitro data obtained by using PNP on a 2'-substituted dGuo analogue, such as Fig. 9. Autoradiogram of a polyacrylamide gel showing competition between dGTP ara-G (31), are consistent with this notion. However, studies have and dFdGTP or ara-GTP for incorporation into DNA primer by pol a. A 32P-labeled DNA primer was extended over a defined template by DNA pol a in the presence of 30 @LM indicated that when ara-G was incubated with whole blood, the drug dCFP, d1TP, or dATP and with 0.1, 0.25, and 0.5 @tMdGTPwithout any analogues (Lanes was degraded within a few h (31). Similarly when ara-G was injected 2â€”4),orwith 10 p@MdFdGTP(Lanes5â€”7)orara-GTP(Lanes8â€”11).Lane1, a 0 reaction into mice, drug was eliminated quickly and levels of uric acid in without any analogue. plasma increased, suggesting that ara-G is hydrolyzed to guanine and then deaminated to xanthine (32). The 2'-fluoro-substituted analogue of dGuo, 2'-fluoro-ara-G (1 1), was neither a substrate nor an inhibitor percentage of product/mm, were 0.22 Â± 0.09 for dFdGTP, of PNP. Consistent with that, the 2',2'-difluoro analogues dFdGuo 0.16 Â±0.01 for ara-GTP, and 0.24 Â±0.05 for dO'!?. The apparent Km and dFdlno were resistant to degradation by PNP; compared to and Vmax of extension after incorporation of dGTP remained similar guanosine and inosine, the velocity of the reaction with analogues was and were 0.02 Â±0.01 p.M and 0.32 Â±0.01% product/mm. The <0.005% (present study). Although the stability of this drug must be velocities of extension after incorporation of the analogue, however, evaluated in vivo, the present data suggest that difluoro modifications could not be determined because analogue insertion halted further would render the nucleoside resistant to phosphorolysis, thereby extension, and the radioactivity above the target site remained con increasing the availability of dFdGuo for bioactivation. stant at different concentrations of dFdGTP or ara-GTP. Because the van der Waals radius of fluorine is similar to that of the To visualize and quantitate the competition between dGTP and hydrogen atom, it is likely that analogue nucleosides that contain either dFdGTP or ara-GTP, various concentrations of dGTP (0.1, 2'-deoxy-2'-fluororibose will resemble natural 2'- deoxyribonucleo 0.25, 0.5 p.M)were added to the reaction mixtures containing either no sides and will be accepted as substrates for deoxyribonucleoside G analogues or 10 p.M concentrations of either dFdGTP or ara-GTP kinases (30). Indeed, three lines of evidence obtained from the present (Fig. 9). In the absence of any 0 nucleotide (Fig. 9, Lane 1), an study demonstrate that dFdGuo is activated by the nucleoside kinase intense band at the pre-G site (23-mer) indicated that polymerization that phosphorylates dGuo: (a) a CHO cell line that is deficient in by pol a paused before the 0 site (site 23). Addition of increasing dCyd kinase was as sensitive to dFdGuo as were the wild-type cells; concentrations of dGTP (Fig. 9, Lanes 2â€”4)resulted in a proportional (b) exogenous dGuo spared CHO cells from the toxicity of dFdGuo decrease in the band at the pre-G site and an increase in the intensity whereas up to 10 mM dCyd did not have a sparing effect (Fig. 3); (c) of product bands at sites 29â€”31.When 10 p.MdFdGTP was available finally and more directly, dGuo kinase identified after DEAE-cellu to compete with 0.1 p.MdOlT (Fig. 9, Lane 5), dFdGTP incorporated lose column chromatography phosphorylated dFdGuo to its mono at the 0 site and caused a pause band at the post-G site (site 25). phosphate (Fig. 4). This activation was totally inhibited by dGTP, a Increasing concentrations of dGTP (0.25 and 0.5 p.M)in these reaction potent inhibitor of this enzyme (33). Additionally, fractions that mixtures, however, decreased the incorporation of dFdGTP which contained dCyd-phosphorylating activity did not convert dGuo or resulted in a decrease of band formation at the post-G site and an dFdGuo to their monophosphates. These data indicate that dGuo increase of the full-length products. Compared to the respective corn kinase is a separate enzyme and is not associated with dCyd kinase plete reactions, the product formation was 61, 66, and 72% at 0.1, (21, 33â€”35).Our previous studies using K562 cells are also consistent 0.25, and 0.5 p.M dGTP, respectively. These data suggested that the with this conclusion (36) and the observation that guanine nucleosides ratio of intracellular dFdGTP to dOlT would affect the extent of are not efficient substrates for dCyd kinase. In contrast, dGuo kinase inhibition of DNA synthesis. dGTP did not appear to compete as also phosphorylated other dGuo analogues, such as ara-0 (36). Under efficiently with ara-GTP (Fig. 9, Lanes 8â€”10)as it did with dFdGTP. chromatographic separation conditions similar to those used in the Hence the band at the 0 site caused by incorporation of ara-GMP and present study, 5'-nucleotidase activity eluted between dGuo kinase inhibition of further primer elongation was still visible at the highest and dCyd kinase (37). Because our assay conditions did not contain concentration of dOlT (Fig. 9, Lane 10). The presence of dGTP IMP, which is a phosphate donor for this enzyme (37, 38), the supported some primer extension to fullest extent (Fig. 9, Lanes identified peak for dFdGuo phosphorylation stongly appears to be 8â€”10); however, this was always less than 25% of a complete reaction related to the activity of dGuo kinase. However, it must be established compared with the reaction containing the same concentrations of whether, in the presence of IMP, dFdGuo may serve as a substrate for dGTP and no ara-GTP. These data could be explained partly by the phosphorylation by 5'-nucleotidase. relative affinities of pol a for these nucleotides as determined by the Identification of the kinase and determination of the apparent Km of kinetic values. this enzyme for dFdGuo (40 p.M) implied that its apparent affinity for 1522

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1995 American Association for Cancer Research. METABOLISM AND ACI1ONS OF DIFLUORODEOXYGUANOSINE this analogue is 3â€”7-foldlower compared to dGuo (Km values 6â€”16 elongation at the incorporation site. These results provide evidence p.M) (33â€”35, 39). Considering experience with other nucleosides, this for differences in the action of these drugs at the molecular level. information may be important in the design of clinical protocols. For The pattern of inhibition observed in the present study using syn example, based on its kinetic properties, dFdCyd would be considered thetic oligonucleotides was similar to that found using single strand an efficient substrate for dCyd kinase [Km dCyd = 1.5; dFdCyd = 4.6 M13 mpl9 (+) as the template for primer extension (data not shown). p.M (40), 0.23 p.M for dCyd and 0.73 p.M for dFdCyd when UTP is Furthermore, this pattern was shared with dFdCyd when pol a pun used as a phosphate donor (41)]; hence the accumulation of dFdCTP fled from human leukemia cells was used to extend a primer (5 1). The is saturated at a low concentration of exogenous dFdCyd (10â€”15p.M) observation that incorporation of either difluoro analogue by pol a both for cell lines and for leukemia cells in vitro (17, 42â€”44)and results in a strong pause in polymerization after incorporation of one during therapy (45). Based on the affinity of d0uo kinase, dFdGuo dNTP beyond the analogue offers a new insight into the mode of would be expected not to saturate dFdGTP accumulation at low action of these analogues at the molecular level. This pattern of concentrations. This was consistent with the observation that there inhibition required incorporation of the next correct dNTP. If the next was a dose-dependent increase in dFdGTP accumulation with up to 1 dNTP was eliminated from the reaction, pol a paused at the incorpo mM exogenous dFdGuo (Ref. 46 and data not shown). Hence a higher ration site (data not shown). The kinetics of incorporation of the next (>50 p.M) concentration of dFdGuo in plasma would be needed to normal deoxynucleotide in a dFdGMP-terminated DNA and further saturate the rate of dFdGTP accumulation. extension of this DNA require further investigation to determine Studies using Escherichia coli ribonucleotide reductase have dem whether this pattern of inhibition is dependent upon the identity of the onstrated that the 2'-fluorinated purine and pyrimidine nucleotide next nucleotide. analogues inactivate this enzyme (47). Additional studies with dFd In contrast to 2',2'-difluoro analogues, the arabinose substitution in CDP suggested that this analogue functions as a mechanism-based the carbohydrate acted in the in vitro assay system as a frank DNA irreversible inactivator (48). Because dFdGuo contains the identical terminator, consistent with other arabinosyl analogues such as ara-C geminal 2'-fluorine modification as does dFdCyd, dFdGDP would (51, 53, 54), ara-G (present study), ara-A (55), and 9-f3-D-arabino also be expected to affect ribonucleotide reductase by a similar furanosyl-2-fluoroadenine (3, 56). These data demonstrate that sub mechanism. Additionally, dFdGTP may act as an allosteric regulator stitution at the 2' position plays a major role in the substrate speci of this enzyme. The effect of dFdGuo on dNTP pools provided the ficity of DNA pol a. Comparison of the incorporation of an analogue evidence that the metabolites of dFdGuo inhibited this enzyme; the with deoxynibose sugar (2-chlorodeoxyadenosine), a nucleoside with major effect was a depletion of dATP followed by a lesser decrease in an arabinose carbohydrate and a base analogue (fludarabine), and the cellular dCTP concentration. dGTP was not affected, and dTTP in 2'-deoxy-2'-fluoroarabinosyl analogue of dAdo (2-chloro-2'-fluoro fact increased slightly (Table 1). arabinosyladenine) revealed that extension of DNA alter incorpora Although the biochemical basis for the actions of dFdGuo nucleo tion of the dAdo analogue was without a major stop site. Insertion of tides on ribonucleotide reductase remains to be elucidated, the differ both arabinosyl and 2'-fluoroarabinosyl dAdo analogues resulted in a ential effect of dFdGuo on dNTP pools may be explained by the major polymerization stop at the incorporation site. These patterns following hypothetical scheme. Two independent allosteric domains were in accord with the kinetic values obtained for the incorporation exist on ribonucleotide reductase: one is activated by ATP and regu of the next normal nucleotide after insertion of the dAdo analogue lated by dATP as a global inhibitor of this domain (49); the second (56). Because the bases were also modified in these dAdo analogues, domain is governed by other dNTPs, which serve as effectors for this is not a true comparison of the effect of the carbohydrate modi substrate specificity. In this scheme, dGTP is a positive effector for fications; nonetheless it suggests that a single fluorine atom produced ADP reduction (50). The endogenous level of dOT? in CHO cells is behavior different from that of difluoro analogues. Crystal structure low (<6 p.M), especially compared to cellular dFdGTP concentrations analysis followed by molecular modeling of DNA containing (Table 1; Fig. 5). By virtue of its structural similarity with dOlT, monofluoro (either nibose or arabinose) or difluoro anomers may dFdGTP may compete with dOlT to bind to the allosteric site, hence identify the structure-function relationship of these congeners. eliminating the positive effect of dOlT on ADP reduction. Another In conclusion, dFdGuo is a new and potent antimetabolite that possibility is that cell lines of unrelated origins may respond differ shares features with its base analogue ara-G and its carbohydrate ently to an analogue. For example in CHO4 and HT-29 colon cancer congener dFdCyd. Resistance to the catabolic enzyme PNP and acti lines (44) treated with dFdCyd, dATP was lowered more than other vation by dGuo kinase are the properties of dFdGuo that parallel dNTPs. In contrast, in leukemia cells, the dCTP pool was dramatically ara-G. Dependence on dGuo kinase for phosphorylation distinguishes reduced by dFdCyd incubation (15). In accord with this observation, dFdGuo from dFdCyd and other nucleoside analogues that are anabo incubation of the human leukemia cell line CEM with dFdGuo de lized by dCyd kinase. Although different from dFdCyd in its metab creased dCFP and not dATP (46). These differences remain unex olism, the molecular action of dFdGTP on DNA synthesis is similar plained and the question of why there are tumor-specific differences to that of dFdCyd and hence seems to be a feature shared by difluoro and regulation of endogenous ribonucleotide reductase in whole cells analogues. This specific pattern of DNA synthesis inhibition places requires further investigation. difluoro analogues in the penultimate position, perhaps making it The present study demonstrated that dFdGuo incubation potently more difficult for proof-reading enzymes to rectify the error (51); this inhibited DNA synthesis. Similar to dFdCyd, dFdGuo affected masked nucleotide effect may be characteristic of the geminal fluorine DNA synthesis much more than RNA synthesis (51, 52). The substitution at the C-2 position. pattern of inhibition of DNA synthesis revealed that purified po1 a from CHO cells incorporated dFdGMP into the 0 sites of the elongating DNA strand but paused after incorporation of the next ACKNOWLEDGMENTS deoxynucleotide. In contrast, ara-OMP terminated DNA strand We thank Lidia Vogelsang for her assistance in preparing the manuscript 4 Unpublished data. and Walter Pagel for editing the manuscript. 1523

REFERENCES 29. LePage, G. A., and Junga, I. G. Metabolism of purine nucleoside analogues. Cancer Rex., 25: 46â€”52,1965. 1. Cozarelli,N.R.ThemechanismofactionofinhibitorsofDNA synthesis.Annu.Rev. 30. Stoeckler, J. D., Bell, C. A., Parks, R. E., Chu, C. K., Fox, J. J., and Ikehara, M. Biochem., 46: 641â€”668,1977. q2')-substituted purine nucleoside analogs. interactions with adenosine deaminase 2. Kufe, D., Sprigs, D., Egan, E. M., and Munroe, D. Relationships among ara-CTP and purine nucleoside phosphorylase and formation of analog nucleotides. Biochem. pools, formation of (ara-C) DNA, and cytotoxicity of human leukemic cells. Blood, Pharmacol.,31:1723â€”1728,1982. 64: 54-58, 1984. 31. Shewach, D. S., Daddona, P. E., Ashcraft, E., and Mitchell, B. S. Metabolism and 3. Huang, P., Chubb, S., and Plunkett, W. Termination of DNA synthesis by 9-g3-D- selective cytotoxicity of9-Ã˜-r-arabinofursnosylguaninein human lymphoblasts.Can arabinofuranosyl-2-fiuoroadenine: A mechanism for cytotoxicity. J. Biol. Chem., cer Rex., 45: 1008â€”1014,1985. 265:16617â€”16625,1990. 32. Elion, G. B., Rideout, J. L., deMiranda, P., Collins, P., and Bauer, D. J. Biological 4. Keating, M. J., McCredie, K. B., Bodey, G. P., Smith, T. L, Gehan, E. A., and activities of some purine arabinosides. Ann. NY Acad. Sci., 255: 468â€”480,1975. Freireich, E. J Improved prospects for long-term survival in adults with acute 33. Barker, J., and Lewis, R. A. Deoxyguanosine kinase of neonatal mouse skin tissue. myelogenous leukemia. JAMA, 248: 2481â€”2486,1982. Biochim. Biophys. Acts, 658: 111â€”123,1981. 5. Robertson, L E., Hall, R., Keating, M. J., Estey, E., Kantaijian, H. M., McLaughlin, 34. Gower, W. R, Carr, M. C., and Ives, D. H. Deoxyguanosine kinase. Distinct molec P., Hagemeister, F. B., and Plunkett, W. High-dose cytosine arabinoside in chronic ular forms in mitochondria and cytosol. J. Biol. Chem., 254: 2180â€”2183,1979. lymphocytic leukemia: a clinical and pharmacologic analysis. Leukemia Lymphoma, 35. Wang, L., Karlsson, A., and Amer. E. S. J., and Eriksson, S. Substrate specificity of 10:43â€”48,1993. mitochondrial 2'-deoxyguanosine kinase. Efficient phosphorylation of 2-chlorodc 6. Keating, M. J., Kantarjian, H., Talpaz, M., Redman, J., Koller, C., Barlogie, B., oxyadenosine. J. Biol. Chem., 268: 22847â€”22852,1993. Velasquez, W., Freireich, E. 1, and McCredie, K. B. Fludarabine: a new agent with major activity against chronic lymphocytic leukemia. Blood, 74: 19â€”25,1989. 36. Gandhi, V., and Plunkett, W. Interaction of arabinosyl nucleotides in 1362 human 7. Whitely, R. J., Sooning, S. J., and Dolin, R. Adenosine arabinoside therapy of biopsy leukemia cells. Biochem. Pharmacol., 38: 3551â€”3558,1989. proved herpes simplex encephalitis. N. Engi. J. Med., 297: 289â€”294,1977. 37. Saunders, P. P., Spindler, C. D., Tan, M-T., Alvarez, E., and Robins, R. K. TiaZOfUrin 8. Brox, L. W., LePage, G. A., Hendler, S. S., and Shannahoff, D. H. Studies on the is phosphorylated by three enzymes from Chinese hamster ovary cells. Cancer Rex., growth inhibition and metabolism of 2'-deoxy-2'-fiuorocytidine in cultured human 50: 5269-5274, 1990. lymphoblasts. Cancer Res., 34: 1838â€”1842,1974. 38. Johnson, M. A., and Fridland, A. Phosphorylation of 2',3'-dideoxyinosine by 9. Jakeman, E. J., Tisdale, M., Russell, S., Leone, A., and Sweet, C. Efficacy of cytosolic 5'-nucleotidase of human lymphoid cells. Mol. Pharinacol., 36: 291â€” 2'-deoxy-2'-fluororibosides against influenza A and B viruses in ferrets. Antimicrob. 295,1989. Agents Chemother., 38: 1864â€”1867,1994. 39. Lewis, R. A., and Link, L. Phosphorylation of arabinosyl guanine by a mitochondrial 10. Burchenal, F. H., Leyland-Jones, B., Watanabe, K., Klein, R., Lopez, C., and Fox, enzyme of bovine liver. Biochem. Pharmacol., 38: 2001â€”2006,1989. J. j. In: Nucleosides, Nucleotides, and Their Biological Applications, pp. 47. New 40. Bouffard, D. Y., Laliberte, J., and Momparler, R. L Kinetic studies on 2',2'- York: Academic Press, 1983. difluorodeoxycytidine (gemcitabine) with purified human deoxycytidinc kinase and 11. Montgomery, J. A., Shortnacy, A. T., Carson, D. A., and Secrist, J. A., III. 9-(2- cytidine deaminase. Biochem. Pharmacol., 45: 1857â€”1861,1993. Deoxy-2-fluoro-@3-r-arabinofuranosyl)guanine:ametabolically stable cytotoxic ana 41. Shewach, D. S., Reynolds, K. K., and Hertel, L. Nucleotide specificity of human logue of 2'-deoxyguanosine. J. Med. Chem., 29: 2389â€”2392,1986. deoxycytidine kinase. Mol. Pharmacol., 42: 518â€”524,1992. 12. Hertel, L. W., Kroin, J. S., Misner, J. W., and Tustin, J. M. Synthesis of 2-deoxy 42. Grunewald, R., Kantarjian, H., Keating, M. J., Abbruzzese, J., Tarassoff, P., and 2,2-difiuoro-D-riboseand 2-deoxy-2,2-difluoro-D-ribofuranosylnucleosides. J. Org. Plunkett, W. Pharmacologically directed design of the dose rate and schedule of Chem., 53: 2406â€”2409,1988. 2',2'-difluorodeoxycytidine (gemcitabine) administration in leukemia. Cancer Rex., 13. Heinemann, V., Hertel, L. W., Grindey, G. B., and Plunkett, W. Comparison of the 50: 6823â€”6826,1990. cellular pharmacokinetics and toxicity of 2',2'-difluorodeoxycytidine and 1-(3-r- 43. Grunewald, R., Abbruzzese, J. L, Tarassoff, P., and Plunkett, W. Saturation of arabinofuranosylcytosine. Cancer Res., 48: 4024-4031, 1988. 2',2'-difluorodeoxycytidine 5'-triphosphate accumulation by mononuclear cells dur 14. Heinemann, V., Xu, Y-Z., Chubb, S., Sen. A., Hertel, L W., Grindey, 0. B., and ing a phase I trial of gemcitabine. Cancer Chemother. Pharmacol., 27: 258â€”262, Plunkett, W. Inhibition of ribonucleotide reduction in CCRF-CEM cells by 2',2'- 1991. difluorodeoxycytidine. Mol. Pharmacol., 38: 567â€”572,1990. 44. Shewach, D. S., Hahn, T. M., Chang, E., Hertel, L W., and Lawrence, T. S. 15. Gandhi, V., and Plunkett, W. Modulatory activity of 2',2'-difluorodeoxycytidine on Metabolism of 2',2'-difluoro-2'deoxycytidine and radiation sensitization of human the phosphorylation and cytotoxicity of arabinosyl nucleosides. Cancer Res., 50: colon carcinoma cells. Cancer Rex., 54: 3218â€”3223,1994. 3675â€”3680,1990. 45. Grunewald, R., Kantarjian, H., Du, M., Faucher, K., Tarasoff, P., and Plunkett, W. 16. Plunkett, W., Huang, P., Xu, Y-Z., Heinemann, V., Grunewald, R., and Gandhi, V. Gemcitabine in leukemia: a phase I clinical, plasma, and cellular pharmacology study. Gemcitabine: metabolism, mechanisms of action, and self-potentiation. Seniin. On J. Clin. Oncol., 10: 406â€”413,1992. col., in press, 1995. 46. Gandhi, V., Mineishi, S., Huang, P., Yang, Y., Chubb, S., Chapman, A. J., Nowak, 17. Kaye, S. B. Gemcitabine: current status of Phase I and II trials (Editorial). J. Chin. B. J., Hertel, L W., and Plunkett, W. Difluorodeoxyguanosine: cytotoxicity, metab Oncol., 12: 1527â€”1531,1994. olism, and actions on DNA synthesis in human leukemia cells. Scmin. OncoL, in 18. Hertel, L. W., and Ternansky, R. J. Fluorine-containing antiviral and anticancer press, 1995. compounds. In: R. Filler et a!. (eds.), Organofluorine Compounds in Medicinal 47. Stubbe, J. A., and Kozarich, J. W. Fluoride, pyrophosphate, and base release from Chemistry and Biomedical Applications, pp. 23â€”71.Amsterdam:Elsevier Science 2'-deoxy-2'-fluoronucleoside-diphosphate reductase. J. Biol. Chem., 255: Publishers B. V., 1993. 5511â€”5513,1980. 19. Hertel, L. W., Grossman, C. S., Kroin, J. S., Mineishi, S., Chubb, S., Nowak, B., and 48. Baker, C. H., Banzon, J., Bollinger, J. M., Stubbe, J., Samano, V., Robins, M. J., Plunkett, W. Synthesis, cytotoxicity and metabolism of the 2',2'-difluoroanalogs of Lippert, B., Jarvi, E., and Resvick, R. 2'-deoxy-2'-methylenecytidine and 2'-deoxy deoxyadenosine (dFdA) and deoxyguanosine (dFdG). Nucleoside Nucleotides, 8: 2',2'-difluorocytidine 5-diphosphates: potent mechanism-based inhibitors of ribonu 951-955,1989. cleotide reductase. J. Med. Chem., 34: 1879â€”1884,1991. 20. Harris, B. A., Saunders, P. P., and Plunkett, W. Metabolism of 9-@-o-xylofuranosy 49. Reichard, P. Interactions between deoxyribonucleotide and DNA synthesis. Annu. ladenine by the Chinese hamster ovary cell. Mol. Pharmacol., 20: 200â€”205,1981. Rev. Biochem., 57: 349â€”374,1988. 21. Saunders, P. P., and I@ai,M. M. Nucleoside kinase activities of Chinese hamster ovary 50. Eriksson, S., Gudas, L J., Clift, S. M., Caras, I. W., Ullman, B., and Martin, D. W., cells. Biochim. Biophys. Acts, 761: 135â€”141,1983. Jr. Evidence for genetically independent allosteric regulatory domains of the protein 22. Plunkett, W., Hug, V., Keating, M. J., and Chubb, S. Quantitation of 1-@-o-arabino Ml subunit of mouse ribonucleotide reductase. J. Biol. Chem., 256: 10193â€”10197, furanosylcytosine 5-triphosphate in the leukemic cells from bone marrow and pe 1081. ripheral blood of patients receiving 1-@-D-arabinofuranosyIcytosine therapy. Cancer 51. Huang, P., Chubb, S., Hertel, L. W., Grindey, G. B., and Plunkett, W. Action of Rca., 40: 588â€”591,1980. 23. Sherman, P. A., and Fyfe, F. A. Enzymaticassayfordeoxynudeosidetriphosphatesusing 2',2'-difluorodeoxycytidine on DNA synthesis. Cancer Rex., 51: 6110â€”6117, 1991. syntheticoligonucleotidesastemplateprimen.Anal.Biochem.,1&'J:222-226,1989. 52. Ruiz, V. W. T., Haperen, V., Veerman, G., Vermorken, J. B., and Peters, G. J. 24. Goodman, M. F., Creighton, S., Bloom, L B., and Petruska, J. Biochemical basis of 2,2'-difluoro-deoxycytidine (gemcitabine) incorporation into RNA and DNA of DNA replication fidelity. Crit. Rev. Biochem. Mol. Biol., 28: 83â€”126,1993. tumour cell lines. Biochem. Pharmacol., 46: 762â€”766,1993. 25. Huang, P., Farquhar, D., and Plunkett, W. Selective action of 3'-azido-3'-deoxythy 53. Townsend, A. J., and Cheng, Y-C. Sequence-specific effects of ara-5-aza-CIP and midine 5'-triphosphate on viral reverse transcriptases and human DNA polymerases. ara-C1'P on DNA synthesis by purified human DNA polymerases in vitro: visualiza J. Biol. Chem., 265: 11914â€”11918,1990. tion of chain elongation on a defined template. Mol. PharmacoL,32: 330â€”339,1987. 26. Chou, J., and Chou, T. Dose-Effect Analysis with Microcomputers, pp. 5â€”18.Am 54. Mikita, T., and Beardsley, 0. P. Functional consequences of the arabinosylcytosine sterdam, the Netherlands: Elsevier Science Publishers, 1985. structural lesion in DNA. Biochemistry, 27: 4698â€”4705,1988. 27. Parks, R. E., Jr., Stoeckler, J. D., Cambor, C., Savarese, T. M., Crabtree, G. W., and 55. Ohno, Y., Spriggs, D., Matsukage, A., Ohno, T., and Kufe, D. Sequence-specific Chu, S-H. Fume nucleoside phosphorylase and 5'-methylthioadenosine phosphoryl inhibition of DNA strand elongation by incorporation of [email protected] ase: targets of chemotherapy. In: A. C. Sartorelli, J. S. Laxo, and J. R. Bertino (cdx.), caine. Cancer Res., 49: 2077â€”2081,1989. Molecular Actions and Targets for Cancer Chemotherapeutic Agents, pp. 229â€”252. 56. Parker, W. B., Shaddix, S. C., Chang, C-H., White, E. L, Rose, L M., Brockman, New York: Academic Press, 1981. R. W., Shortnacy, A. T., Montgomery, J. A., Secrist, J. A., ill, and Bennett, L L, Jr. 28. Heinemann, V., Xu, Y-Z., Chubb, S., Sen, A., Hertel, L W., Grindey, G. B., and Effects of 2-cMoro-9-(2-deoxy-2-fluoro-@-D-arabinofuranosyl)adenineonK562 eel Plunkett, W. Cellular elimination of 2',2'-difiuorodeoxycytidine 5-triphosphate: a lular metabolism and the inhibition of human ribonucleotide reductase and DNA mechanism of self-potentiation. Cancer Res., 52: 533â€”539,1992. polymerases by its 5'-triphosphate. Cancer Res., 51: 2386â€”2394,1991.

1524

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1995 American Association for Cancer Research. Cytotoxicity, Metabolism, and Mechanisms of Action of 2′,2′ -Difluorodeoxyguanosine in Chinese Hamster Ovary Cells

Varsha Gandhi, Shin Mineishi, Peng Huang, et al.

Cancer Res 1995;55:1517-1524.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/55/7/1517

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/55/7/1517. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.