ICANCER RESEARCH 52, 533-539, February 1, 1992] Cellular Elimination of 2',2'-Difluorodeoxycytidine 5'-Triphosphate: A Mechanism of Self-Potentiation1
Volker Heinemann,2 Y¡-ZhengXu, Sherri Chubb, Alina Sen, Larry W. Hertel, Gerald B. Grindey, and William Plunkett1
Department of Medical Oncology, The University of Texas M. D. Anderson Cancer Center, Houston, Texas 77030 [V. H., Y. X., S. C., A. S., W. P.], and Lilly Research Laboratories, Indianapolis, Indiana 46285 (L. W. H., G. B. G.]
ABSTRACT less, clinical cellular pharmacology studies have demonstrated 2',2'-Difluorodeoxycytidine (dFdC, Gemcitabine) is a deoxycytidine that the dFdCTP:dCTP value reaches potentially inhibitory analogue which, after phosphorylation to the 5'-di- and 5'-triphosphate values during clinical trials (4, 10, 11). (b) dFdCTP is incor porated into DNA by DNA polymerases a and f, inhibiting (dFdCTP), induces inhibition of DNA synthesis and cell death. We examined the values for elimination kinetics of cellular dFdCTP and further elongation (9). (c) Once incorporated, dFdCMP resi found they were dependent on cellular concentration after incubation of dues in the terminal or penultimate positions of the DNA strand CCRF-CEM cells with dFdC and washing into drug-free medium. When inhibit the editing function of DNA polymerase <(9). This may the drug was washed out at low cellular dFdCTP levels (<50 n\\), fix damage caused by the incorporated analogue, (d) dFdCDP dFdCTP elimination was linear (t,: = 3.3 h), but it became biphasic at inhibits ribonucleotide reducÃase,blocking DNA synthesis by intracellular dFdCTP levels >100 JIM. Although the initial elimination decreasing the cellular concentrations of deoxynucleoside tri rate was similar at all concentrations, at higher concentrations the phosphates (12-14). terminal elimination rate increased with increasing cellular dFdCTP In several human (CEM and K562) and rodent (CHO) cell concentration, with a nearly complete inhibition of dFdCTP elimination at 300 pM. The deamination product 2',2'-difluorodeoxyuridine was the lines cellular elimination of high dFdCTP concentrations (>100 i/\i ) is biphasic, with a short initial half-life followed by a second predominant extracellular catabolite at low cellular dFdCTP concentra tions, whereas at high dFdCTP concentrations dFdC was the major phase of considerably slower degradation (7, 13, 15). This excretion product. The dCMP deaminase inhibitor 3,4,5,6-tetrahydro- biphasic elimination of dFdCTP differs from the linear kinetics deoxyuridine transformed the monophasic dFdCTP degradation seen at exhibited during elimination of the triphosphates of arabino- low dFdCTP levels into a biphasic process, whereas the deoxycytidine sylcytosine (16), arabinosyladenine (17), and arabinosyl-2-fluo- deaminase inhibitor 3,4,5,6-tetrahydrouridine had no effect on dFdCTP roadenine (18) in human leukemia cells after therapy. Further elimination. An in situ assay indicated that dCMP deaminase activity more, the slow terminal catabolism of dFdCTP contributes to was inhibited in whole cells, an action that was associated with a a greater dFdCTP area under the concentration x time curve decreased dCTP:dTTP value. In addition, dFdCTP inhibited partially in cells. The continued presence of the active nucleotides is purified dCMP deaminase with a 50% inhibitory concentration of 0.46 m\i. We conclude that dFdC-induced inhibition of dCMP deaminase associated with a prolonged inhibition of DNA synthesis and, thus, may contribute to greater cytotoxicity (7, 9). resulted in a decrease of dFdCTP catabolism, contributing to the concen tration-dependent elimination kinetics. This action constitutes a self- We investigated the extent to which dFdCTP elimination is potentiation of dFdC activity. affected in CEM cells by cellular concentrations of dFdCTP and its metabolites. We suggest here a mechanistic model that describes dFdC-mediated modulation of dCMP deaminase as a INTRODUCTION determinant of dFdCTP elimination and identifies dFdC as a dFdC4 (Gemcitabine) is a deoxycytidine analogue in which drug with self-potentiating activity. geminai fluorines replace both hydrogens of the 2' carbon atom (1). An unusually broad spectrum of activity in murine tumors MATERIALS AND METHODS (2) and human tumor xenografts (3) provided impetus for Chemicals. dFdC, dFdU, dFdCMP, and [5-14C]dFdC were synthe evaluation of its anticancer activity in clinical trials (4-6). dFdC sized by published procedures (1) at Lilly Research Laboratories. THU must be phosphorylated by deoxycytidine kinase to exhibit was generously provided by Dr. Ven Narayanan, Drug Synthesis and cytotoxic and therapeutic activities (7,8). Its major intracellular Chemistry Branch, National Cancer Institute, and dTHU was obtained metabolite is dFdCTP, although it remains in a constant ratio from Behring Diagnostics (La Jolla, CA). Deoxycytidine, dCMP, with the lesser metabolites, dFdCMP and dFdCDP (7). dCTP, and all other nucleosides and nucleotides were purchased from DNA synthesis is specifically inhibited by dFdC by several Sigma Chemical Co., Inc. (St. Louis, MO). separate mechanisms, (a) dFdCTP competes with dCTP as a Synthesis of dFdCTP. dFdCTP was synthesized from dFdCMP by a weak inhibitor of mammalian DNA polymerase (9). Neverthe- modification of the procedure of Hoard and Ott (19). dFdCMP (34.3 mg, 0.1 nimnl) was converted into its pyridinium salt with the pyridi- Received 8/21/91; accepted 11/13/91. nium form of Dowex-50W X-8 cation exchange resin. The tributylam- The costs of publication of this article were defrayed in part by the payment monium salt was prepared by addition of tributylamine (2 equivalents), of page charges. This article must therefore be hereby marked advertisement in and the product was dried in vacuo. The residual gum was dissolved in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1Supported in part by Grant CH-130 from the American Cancer Society and anhydrous /V.jV-dimethylformamide (2 ml/0.1 mmol), and 1,1'-car- Grant CA28596 from the National Cancer Institute, Department of Health and bonyl-bis(imidazole) (1.6 mmol) was added. After 16 h of stirring at Human Services. : Present address: Department of Internal Medicine. Hematology/Oncology, room temperature under argon, methanol was added (0.035 ml/0.1 mmol), and the solution was stirred for IO min more. Tributylammon- Klinikum Gro¡ihadcrn,University of Munich, Munich. Germany. 3To whom requests for reprints should be addressed. ium pyrophosphate (5 equivalents, 0.5 mmol), prepared from the pyri 4 The abbreviations used are: dFdC, 2'.2'-difluorodeoxycytidine;CEM, CCRF- dinium salt by addition of 5 equivalents of tributylamine in N,N- CEM lymphoid cells; dFdCMP, dFdCDP, and dFdCTP, the 5'-mono-, dì-,and triphosphates of dFdC; dFdU and dFdUMP, 2'.2'-difluorodeoxyuridinc and its dimethylformamide (5 ml/0.1 mmol), was then added dropwise. The 5'-monophosphate; dNTP, deoxynucleoside triphosphate; HPLC, high-pressure reaction mixture was stirred vigorously for 16 h at room temperature liquid chromatography; THU, 3,4,5,6,-tetrahydrouridine; dTHU, 3,4,5,6-tetra- and under argon. The solution was then concentrated in vacuo to 1 ml. hydrodeoxyuridine; dCyd, deoxycytidine. dFdCTP was isolated by HPLC chromatography using a preparative 533
Downloaded from cancerres.aacrjournals.org on September 25, 2021. © 1992 American Association for Cancer Research. METABOLIC SELF-POTENTIATION OF dFdC aniónexchange Magnum-20 SAX column (Whatman, Inc.) using an of dUMP or dFdUMP produced/mg of protein/min using bovine serum isocratic How (10 ml/min) of 35% Buffer A (0.005 M NH4H2PO4, pH albumin as a standard for protein measurement. The final dCMP 2.8) and 65% Buffer B (0.75 M NH4H2PO4, pH 3.5). The nucleotide deaminase preparation had a specific activity of 850 nmol/mg/min. was adsorbed to activated charcoal, washed with H2O, and eluted with Assay of dCMP Deaminase Activity in Intact CEM Cells. After the ammoniacal ethanol, and the eluate was evaporated in vacuo. The indicated incubation with dFdC, CEM cells (1 to 3 x IO7) were dFdCTP yield was 87%, and the purity > 95% by HPLC analysis. incubated with 0.2 ^Ci of [14C]dCyd in 5 ml of cell culture medium Cell Line. The human T-lymphoblast cell line CCRF-CEM was without fetal bovine serum for 15 min in the presence of 5 ^g of obtained from the American Type Culture Collection (Rockville, MD) aphidicolin (23). Cells were quickly washed with ice-cold phosphate- and maintained in suspension culture in RPMI 1640 medium (GIBCO buffered saline and extracted with 0.4 N HC1O4. After neutralization Laboratories, Grand Island, NY) supplemented with 10% heat-inacti with KOH, portions of the soluble extracts were analyzed using a vated fetal bovine serum (GIBCO) at 37°Cina humidified atmosphere Partisil 10-SAX column with a flow rate of 1 ml/min: 0 to 10 min, containing 5% CO2. Periodic tests for Mycoplasma contamination, isocratic 100% Buffer A; 10 to 70 min, linear gradient from 100% conducted by the American Type Culture Collection, were consistently Buffer A to 100% Buffer C; 70 to 75 min, isocratic 100% Buffer C. negative. All experiments were performed with exponentially growing Radioactive deoxynucleotides were detected with a radioactive flow cells. Cell number and volume were determined by a Coulter Counter detector (Model A250; Packard Instrument Co., Meriden, CT). The equipped with a Model C-1000 particle size analyzer (Coulter Electron eluant was mixed with scintillation fluid (Flo-Scint, IV; Packard In ics, Hialeah, FL). The mean cell volume was 9.43 x 10"" liters/cell. strument Co) at a 1:3 ratio. The dCMP deaminase activity index was Nucleotide Extraction and Analysis. Cells were washed with ice-cold calculated using the following equation phosphate-buffered saline and collected by centrifugation, and the pellet [14C]dTTP (dpm) was extracted with 0.4 N HC1O4 as previously described (7, 8). The |['4C]dCTP (dpm) + [HC]dTTP (dpm)| nucleosides and nucleotides in the acid-soluble, neutralized cell extract were analyzed by HPLC using instruments from Waters Associates, Measurement of Cellular dCTP and dTTP Pools. Deoxyribonucleo- Inc. (Milford, MA). The system was equipped with two Model 6000A pumps, a Model 680 gradient programmer, and a Partisil 10-SAX side triphosphates were extracted from CEM cells with 0.4 N HC1O4, and the acid-insoluble material was removed by centrifugation. The aniónexchange column (250 x 4 mm) (Whatman, Inc., Clifton, NJ). supernatant was carefully monitored to pH 7 with KOH, and after The nucleotides were quantitated with a Model 440 UV detector and removal of KC1O4, samples were stored at -20°C until analysis. A a Model 730 data module. A Model 490 UV detector and a Model DNA polymerase assay using synthetic oligonucleotides as template 840 data module (Waters) were used for determinations of primers was applied to determine dCTP and dTTP pools as described deoxynucleotides. by Sherman and Fyfe (24). Cellular NTP and dFdCTP were separated by HPLC using a concave gradient (Curve 9) run over 30 min at a flow rate of 3 ml/min starting with 65% Buffer A and 35% Buffer B and ending at 100% Buffer B. RESULTS External standard quantitation was used to determine the amount of NTP detected at 280 nm of UV. The intracellular NTP concentration dFdCTP Accumulation. dFdCTP accumulated rapidly in was calculated by dividing the NTP amount by the number of cells CEM cells incubated with 0.01 to 10 ^M dFdC. As shown in analyzed and the mean cell volume. Nucleoside mono-, di, -and tri- Fig. 1, increasing the concentration of exogenous nucleoside phosphates were also separated by anión exchange HPLC. A linear proportionally increased the triphosphate. Cellular dFdCTP gradient from 100% Buffer A to 100% Buffer B was run over 40 min concentrations as great as 500 ^M accumulated after cells were at a rate of 2 ml/min. To determine cellular dNTP concentrations, HClO4-soluble, neutralized extracts from 2 x IO7cells were evaporated incubated with 10 MMdFdC for 2 h. to dryness in an Evapomix volume-reduction apparatus (Buchler In Cellular dFdCTP Elimination. The relationship between cell struments, Fort Lee, NJ), and the ribonucleotides were degraded by ular dFdCTP concentration and the rate of dFdCTP elimina tion was analyzed after a 2-h incubation of CEM cells with 0.1, periodate oxidation as previously described (20). dNTP and dFdCTP were separated on a Partisil-10 SAX column with a total run time of 0.3, 1.0, and 10 MMdFdC (Fig. 2). Average cellular dFdCTP 43 min. An isocratic elution with 75% Buffer A and 25% Buffer C concentrations of 37, 138, 346, and 525 MM accumulated, (buffer B adjusted to pH 3.7) was maintained for 20 min at a flow rate respectively. After the cells were washed into drug-free culture of 3 ml/min and was followed by a linear gradient to 21% Buffer A and medium, the patterns of dFdCTP elimination were determined. 69% Buffer C over 23 min. More than 90% of the dNTP was recovered Cellular elimination of dFdCTP was linear with a />/,of 3.0 h by this procedure, as determined by material balance measurements of after a 2-h incubation with 0.1 MMdFdC. In contrast, after radioactive standards. The pattern of dFdCTP elimination, correlation coefficients for goodness of fit (r), and rrt were determined by the ESTR1P computer 500 •¿ program (21). Determination of dFdC and dFdU in the Culture Medium. The culture medium was extracted with 0.4 N HC1O4, neutralized, and assayed by 400 - reverse-phase HPLC using a ¿iBondapakC,8 column (Waters Associ ates, Inc.). The nucleosides were separated by isocratic elution with 300 - 0.05 M ammonium acetate, pH 6.8. The column was then regenerated by washing with 50% methanol. 200 - Partial Purification and Assay of dCMP Deaminase Activity. Extracts of CEM cells were prepared by a previously described method (12). dCMP deaminase was further purified by HPLC through a molecular 100 - sieve column (Protein Pak Glass 300SW; Nihon Waters, Ltd., Japan). The flow rate was 0.8 ml/min, and the buffer contained 50 mM 4-(2- o- hydroxyethyl)-l-piperazineethanesulfonic acid, pH 7.5, 2 mM MgCl2, and 3 mM dithiothreitol. dCMP deaminase activity was measured by 0 0.0 1 0 1 I 10 the HPLC method reported by Fridland and Verhoef (22). The produc dFdC, liU tion of dUMP or dFdUMP was determined after 2, 4, and 6 min at 37°C;the initial reaction rate was calculated from the slope of the Fig. 1. dFdCTP accumulation in CEM cells. Cells were incubated with the indicated concentrations of dFdC for 2 h, when nucleotide pools were extracted regression line. The dCMP deaminase activity was expressed as nmol and analyzed for dFdCTP content. Points, mean; bars, SEM. 534
Downloaded from cancerres.aacrjournals.org on September 25, 2021. © 1992 American Association for Cancer Research. METABOLIC SELF-POTENTIATION OF dFdC
1000 whereas at low cellular dFdCTP levels, deamination predominated. The activity of deoxycytidine deaminase is very low in CEM cells (22); this suggests that the dCMP deaminase pathway is 1 00 mainly responsible for dFdCMP deamination. Additional evi dence was sought to reveal the role this enzyme plays in dFdCTP elimination. CEM cells were incubated for 2 h with 0.1 MM [l4C]dFdC, followed by washing into fresh medium 10 alone, with 100 MMdTHU or 100 MMTHU. After 4 h, the proportions of dFdC and dFdU in the medium containing control cells were 32% to 68%, respectively. This did not change in the culture to which the cytidine deaminase inhibitor, THU, 10 was added (33 to 67%, respectively). Excretion of dFdU was Hours largely inhibited, however, in cells incubated with dTHU (88% dFdC to 12% dFdU, respectively). It is worth noting that dTHU Fig. 2. dFdCTP elimination as a function of cellular dFdCTP concentration. CEM cells were incubated for 2 h with 0.1 (•),0.3 (A), 1.0 (»),or 10 (O) MM inhibits cytidine deaminase and, after intracellular phosphor- dl (K - After washing cells into drug-free medium, cellular concentrations of ylation to the monophosphate, it is a potent inhibitor of dCMP dFdCTP were measured at the indicated times. Correlation coefficients (r) for Tilting of elimination kinetics were the following: 0.1 MM,0.964; 0.3 /.\i. 0.99S; deaminase (25). These results are consistent with the notion 1.0 MM,0.946; 10 MM,0.970. Points, mean of three separate experiments. Standard that dCMP deaminase has a central role in dFdCTP elimina deviations, which were omitted for clarity, were less than 15% of the means. tion. Based on the foregoing data, we tested the hypothesis that deamination was the rate-limiting step in dFdCTP elimination incubation with 0.3,1.0, and 10 MMdFdC, dFdCTP elimination and that high concentrations of dFdC metabolites may either was best fit to biphasic kinetics; the initial tv, values of 0.6, 1.2, and 1.3 h, respectively, were not significantly different. The Table 1 Cellular distribution ofdFdC metabolites terminal tv, values of dFdCTP in these cultures, 5, 16, and 43 CEM cells were incubated for 2 h with either 10 MM[MC]dFdC or 0.1 MM[MC] h, respectively, increased as a function of cellular dFdCTP dFdC. The cellular dFdCTP levels were 464 MMand 38 MM,respectively. Cells concentration. Cellular integrity, as determined by Coulter vol were then washed into fresh medium and portions of each culture were extracted ume analysis, was maintained throughout the experiments. at the indicated times and analyzed for intracellular dFdC metabolites. dFdC Metabolites. The intracellular distribution of dFdC % of total metabolites metabolites in CEM cells was determined after a 2-h incubation Oh 2b 41, 6h 8h with either 0.1 MMor 10 MMdFdC (Table 1) to investigate the metabolitedFdCdFdUdFdCMPdFdUMPdFdCDPdFdCTP0.110 MMdFdC metabolic basis for the concentration-dependent elimination of dFdCTP. Similar percentages of dFdC metabolites were ob served with the two concentrations. dFdCTP was the major drug metabolite (about 80% of the total) followed by dFdUMP (about 15%), dFdCDP (about 2%), and dFdCMP (about 1%). The percentage of dFdCTP decreased with increasing time after MMdFdCmetabolitedFdCdFdUdFdCMPdFdUMPdFdCDPdFdCTP0.30.41.415.31.880.9ND"0.90.411.92.384.50.60.83.115.76.173.6ND1.41.413.92.980.50.61.44.217.74.971.2ND3.22.417.53.473.51.24.41.915.23.074.3ND4.93.417.42.471.83.62.216.03.076.3ND8.32.420.51.667.2 washout as did the ratio of dFdU to dFdUMP. Although dFdU rose approximately 10-fold in cells from each culture, it com prised less than 10% of the cellular dFdC metabolites at 8 h. However, dFdUMP levels nearly doubled to >20% of cellular dFdC metabolites as the percentage of dFdCTP decreased in 1ND, not detected. the cells with low initial dFdCTP concentrations. Excretion of dFdC and dFdU. To better characterize the end products of dFdCTP catabolism, we analyzed the culture me dium for dFdC and its metabolites after a 2-h incubation with 60 0.1 MMor 10 MM[14C]dFdC. dFdC and dFdU were the only radioactive compounds detected after washing cells into drug- 50 0 6 free medium. After incubation with 0.1 MMdFdC, dFdU accu 40 mulated in the medium to a substantially greater extent than did dFdC (Fig. 3A). The accumulation of dFdU was linear over 30 the initial 4 h (1.8 pmol/ml x h) and then reached a plateau, whereas the dFdC level detected in the medium remained at 20 essentially background levels after drug washout. In contrast, 10 accumulation of dFdC in the medium greatly exceeded that of dFdU after incubation with 10 MMdFdC (Fig. 3B). At 4 h after drug washout, when dFdC in the medium reached a maximum, the ratio of dFdU to dFdC was 21% to 79%. At that time dFdC 0246 2 4 6 accumulation reached a plateau, reflecting the low rate of Hours Hours dFdCTP elimination (Fig. 3). dFdU accumulation (3.1 pmol/ Fig. 3. Extracellular accumulation of dFdU and dFdC as catabolites of dFdCTP ml x h) remained linear over 8 h after drug washout. Thus, at elimination. CEM cells were exposed to 0.1 MM(14C|dFdC (A) or to 10 MM[UC] dFdC (B) for 2 h and washed into drug-free medium. At the indicated times, high cellular dFdCTP concentrations, dephosphorylation and dFdU (O) and dFdC (•)concentrations in the medium were determined. Points. excretion of dFdC appeared to be the dominant catabolic route, mean of two experiments; bars, SEM. 535
Downloaded from cancerres.aacrjournals.org on September 25, 2021. © 1992 American Association for Cancer Research. METABOLIC SELF-POTENTIATION OF dFdC
100 ËE*»0Ec¿'»_<"oC£700600 \ 00
10 80 5004003002001000AXx°"o'I/Q
60
--^'/' 40 /*T*
20
0.1 *, O 510I5 20 25 30 2 4 6 8 10 20 40 60 80 100 Hours dCTP, ¡tu dTTP, /IM Fig. 4. Effect of dTHU on dFdCTP elimination. After a 2-h exposure to 0.1 MM,dFdC, CEM cells were washed free from the drug and were reincubated with Fig. 6. Effect of dCTP and dTTP on dCMP deaminase activity. The effects of 100 ¡IMdTHU (O) or without dTHU (•).The cellular dFdCTP concentrations the indicated concentrations of dCTP (A) and (A) were determined on the were determined at the indicated time points. Points, mean of three separate deamination of dCMP (O) or dFdCMP (•)bypartially purified dCMP deaminase. experiments; bars, SEM. Occasionally a SEM was less than the size of the symbol. Points, mean of three determinations; bars, SEM.
0.1 to 10 MMdFdC before being washed into drug-free medium. 0.70.60.50.40.30.20 0.70.60.50.40.30.20.1 dCMP deaminase activity in situ was measured in indicated •¿ /^.^__— time periods after drug washout (Fig. 5A). dCMP deaminase -oA- was inhibited in a dose-dependent manner. After washing into drug-free medium, the enzyme activity remained suppressed in l^^*• cells treated with 1 and 10 MMdFdC, but recovered in cells ^-o^r'A?*^'D treated with 0.3 and 0.1 MMdFdC. The dCTP and dTTP pools "i^• were also measured in these cells. The dCTP:dTTP value in control cells was 0.38, whereas this decreased to 0.15, 0.11, \a—G_D. 0.08, and 0.04 after the 2-h incubation with 0.1, 0.3, 1.0, and -D-Ö 10 MMdFdC, respectively (Fig. 5B). As with inhibited dCMP D^ia^Ë)02468Hours—¿ D—¿ 1A•— deaminase activity, recovery of the cellular dCTP:dTTP values 0B/^*/-0 after dFdC washout was dose dependent. 02468 Substrate Specificity of dCMP Deaminase. The ability of Hours dCMP deaminase to deaminate dFdCMP, dCMP, and 1-/3-D- Fig. S. Inhibition of dCMP deaminase by dFdC after drug washout. CEM cells arabinofuranosylcytosine monophosphate was assayed in par were preincubated with 0 (O), 0.1 (•),0.3(A), 1.0 (X), or 10.0 (D) MMdFdC for tially purified extracts of CEM cells. The apparent Km values 2 h. Cells were then washed into drug-free medium, and the dCMP deaminase were 0.138 ±0.003, 0.061 ±0.003, and 1.062 ±0.483 mM, activity index (A) and dCTP:dTTP values (B) were measured. Points mean of two separate experiments. respectively. The apparent Fmaxvalues were 6.1 ±0.17, 27.1 ± 2.4, and 2.0 ±0.6 Mmol/mg/min, respectively, yielding respec tive substrate efficiencies ( Kmax:/Tm)of44,444, and 2. Thus, the directly or indirectly inhibit dCMP deaminase, decreasing ex effectiveness of dCMP deaminase-mediated deamination of cretion of dFdU into the medium. dFdCMP appears to be 10-fold less than that of dCMP but 22- dCMP Deaminase as a Determinant of dFdCTP Elimination. fold greater than l-/3-D-arabinofuranosylcytosine monophos- To evaluate the mechanism of the concentration-dependent phate deamination. elimination characteristics of dFdCTP, we used dTHU to in Modulation^ dFdCMP Deamination by dCTP and dTTP. It hibit dCMP deaminase activity in intact cells. After incubation is well established that the activity of dCMP deaminase from with 0. l MMdFdC for 2 h, the cells were washed into drug-free other sources (26, 27) on dCMP is regulated by dCTP and medium. As shown previously, the cellular dFdCTP concentra dTTP. A similar relationship appeared to hold for the deami tion reached about SO ¿
Downloaded from cancerres.aacrjournals.org on September 25, 2021. © 1992 American Association for Cancer Research. METABOLIC SELF-POTENTIATION OF dFdC deamination was dependent on the dFdCTP concentration, we sought to determine whether dFdCTP might directly inhibit 100 dCMP deaminase activity. The preparation of dCMP deami- nase from CEM cells was inhibited by dFdCTP (Fig. 7). In the 80 presence of 5 ^M dCTP, deamination of dCMP was inhibited with 50% inhibitory concentration of 0.46 mM dFdCTP. As shown in Fig. 1, this cellular concentration is exceeded after incubation with 10 UMdFdC for 2 h. 40
DISCUSSION 20 Cellular dFdCTP elimination is a concentration-dependent process with biphasic kinetics at cellular levels of 100 pM 0.0 0.5 1.0 dFdCTP or greater (Fig. 2). Lower dFdCTP concentrations (50 dFdCTP. mM MM)were eliminated with linear kinetics after cells were washed Fig. 7. Inhibition of dCMP deamination by dFdCTP. dCMP deamination was into fresh medium. Although the initial tVlwas apparently not measured in partially purified cell extracts by determining the amount of dUMP influenced by cellular concentrations of dFdCTP, the terminal formed. The experiment was performed with 61 /IM dCMP as a substrate for ivi increased with increasing cellular dFdCTP concentration; dCMP deaminase in the presence of 5 JIMdCTP. Points, mean of three determi nations; bars, SEM. elimination was almost completely inhibited at 300 /¿M dFdCTP. Comparable dFdCTP elimination kinetics values were found in K562 (13), Chinese hamster ovary (7), and HL- allosteric regulation by dCTP and dTTP (26, 27). dCTP acts 60 cells (data not shown). The concentration-dependent prolon as an allosteric activator of dCMP deaminase-mediated gation of the terminal !•isassociated with an increased area dFdCMP deamination, whereas dTTP inhibits this reaction under the concentration x time curve of cellular dFdCTP and (Fig. 6); the degree of inhibition was dependent on the thus may enhance cytotoxicity (7, 9). dCTP:dTMP value. The importance of these considerations is Accumulation of intermediary metabolites might indicate evident in view of dFdC-mediated perturbations of cellular which specific step was responsible for the differential elimi dNTPs (Fig. 5B). dFdCDP is a potent inhibitor of ribonucleo- nation kinetics of dFdCTP. The cellular distribution of accu tide reducÃase(12-14). dFdC decreased dCTP to a significantly mulated dFdC metabolites was not much different among cell greater extent than dTTP, thus lowering the value of cultures incubated 2 h with 0.1 and 10 UM dFdC (Table 1). dCTP:dTTP which probably contributed to dFdC-mediated However, 8 h after washing into fresh medium, the amount of inhibition of dCMP deaminase. The degree of dCTP:dTTP dFdUMP doubled in cells containing low concentrations of decrease and the rate of recovery from the decrease were cor dFdCTP, whereas dFdUMP in cells incubated with 10 ^M dFdC related with the cellular concentration of dFdCTP and its did not increase substantially. dFdUMP was a major cellular metabolites, (b) dFdCTP may consequently inhibit its own product of dFdCTP catabolism under either condition. Because catabolism by a concentration-dependent modulation of dCMP dFdU is a poor substrate for phosphorylation to dFdUMP (13), deaminase. dFdCTP inhibited partially purified dCMP deami it may be assumed that dFdUMP arose from the action of nase with a 50% inhibitory concentration of 0.46 mM. Intracel dCMP deaminase. lular dFdCTP concentrations close to this 50% inhibitory con To pursue this possibility, the extracellular medium was centration value were reached after a 2-h incubation with 10 analyzed for the catabolic end products dFdC and dFdU. Ex MMdFdC (Fig. 2). It may therefore be assumed that dFdCMP cretion of nucleosides into the culture medium after drug wash deamination is partly inhibited in the cell by high dFdCTP out initially occurs in a primarily unidirectional fashion. Thus, concentrations. reequilibration of intracellular and extracellular nucleosides is We tested the hypothesis that dFdC mediates dCMP deami initially not likely to confound the analysis. Although dFdC is nase inhibition with regard to the kinetics of dFdCTP elimi a good substrate for dCyd deaminase (13), the activity of this nation using THU as a selective inhibitor of dCyd deaminase. enzyme in CEM cells is extremely low (22). Extracellular dTHU, which also inhibits dCyd deaminase, is phosphorylated accumulation of dFdU is therefore a result of dCMP deaminase to dTHUMP, which is an inhibitor of dCMP deaminase (25). activity and can be suppressed by dTHU. Because dFdU was dTHU-mediated inhibition of dCMP deaminase and dCyd de not significantly phosphorylated in CEM cells, extracellular aminase changed linear dFdCTP elimination to biphasic elim dFdU may be regarded as an end product of dFdCTP catabo ination kinetics similar to that seen at high dFdCTP concentra lism arising from the activity of dCMP deaminase. tions (Fig. 4). This result could not be obtained by THU, which At low intracellular dFdCTP concentrations (Fig. 3/1), dFdU solely inhibits dCyd deaminase. Accordingly, we hypothesized was the predominant metabolite in the extracellular medium. that deflection from linear dFdCTP elimination was primarily However, at high cellular dFdCTP concentrations (Fig. 3Ä), due to inhibition of dCMP deaminase. Inhibition of dFdCMP relatively low levels of dFdU were excreted; the major extracel deamination would decrease dFdCTP catabolism, contributing lular metabolite was dFdC. It appears that, at high dFdCTP to the change in the elimination kinetics of dFdCTP. concentrations, the activity of dCMP deaminase was signifi Direct measurement of dCMP deaminase activity in whole cantly less important than dephosphorylation of dFdCMP to cells confirmed that the enzyme was inhibited by preincubation dFdC. Thus, dFdC accumulated extracellularly as the predom with dFdC (Fig. 6A). The inhibition was dependent on exoge inant product. These findings may be due to diminished dCMP nous dFdC concentration. Higher intracellular dFdCTP con deaminase activity at high cellular concentrations of dFdCTP. centrations apparently resulted in stronger inhibition of dCMP Two mechanisms of dFdC-mediated inhibition of dCMP deaminase activity. This was most likely the combined effect of deaminase were observed, (a) dCMP deaminase is subject to direct inhibition of the enzyme by dFdCTP and an indirect 537
Downloaded from cancerres.aacrjournals.org on September 25, 2021. © 1992 American Association for Cancer Research. METABOLIC SELF-POTENTIATION OF dFdC
ACKNOWLEDGMENTS The authors are grateful to Maureen Goode for valuable editorial assistance in the preparation of this manuscript.
REFERENCES
1. Hertel, L. W., Kroin, J. S., Misner, J. W., and Justin, J. M. Synthesis of 2- deoxy-2,2-difluoro-D-ribose and 2-deoxy-2,2-difluoro-D-ribofuranosyl nucle- osides. J. Org. Chem., 53: 2406-2409, 1988. 2. Hertel, L. W., Boder, G. B., Kroin, J. S., Rinzel, S. M., Poore, G. A., Todd, G. C., and Grindey, G. B. Evaluation of the antitumor activity of a (2',2'- dif!uoro-2'-deoxycytidine). Cancer Res., 50: 4417-4422, 1990. dFdC dFdU 3. Braakhuis, B. J., van Dongen. A. M. S., Vermorken, J. B., and Snow, G. B. Preclinical in vivo activity of 2',2'-difluorodeoxycytidine (Gemcitabine) Fig. 8. Scheme of cellular metabolism and actions of dFdC and its nucleotides. Dashed lines indicate inhibitory actions. against head and neck center. Cancer Res., 51: 211-214, 1991. 4. Abbruzzese, J. L., Grunewald, R., Weeks, E. A., Gravel, D., Adams, T., Nowak, B., Mineishi, S., Tarassoff. P., Satterlee, W., Raber, M. N., and action brought about by the decrease of the dCTP:dTTP value. Plunkett, W. A Phase I clinical, plasma, and cellular pharmacology study of Gemcitabine. J. Clin. Oncol., 9:491-498, 1991. Like most of the dFdC metabolites, dCMP deaminase activity 5. Anderson, H., Lund, B., Hansen, H. H., Walling, J., and Thacher, N. Phase remained relatively unchanged up to 8 h after washing the cells II study of Gemcitabine in non-small cell lung cancer (NSCLC). Proc. Am. treated with 10 UMdFdC into fresh medium. Soc. Clin. Oncol., /0:247, 1991. A summary of the cellular metabolism of dFdC and the 6. Tanis, B., Clavel, M., Guastalla, J., Osborne. D., Quik, T., McDaniel, G., Pinedo, H., and Vermorken, J. Phase I study of Gemcitabine (difluorodeox- actions of its nucleotides is presented in Fig. 8. The present ycytidine: dFdC; LY188011) administered in a two-weekly schedule. Proc. studies will be discussed with the view that the pharmacody- Am. Assoc. Cancer Res., 31: 207, 1990. namic actions of dFdC nucleotides potentiate the cytotoxicity 7. Heinemann. V., Hertel, L. W., Grindey, G. B.. and Plunkett, W. Comparison of the drug, (a) The major cellular metabolite, dFdCTP, com of the cellular pharmacokinetics and toxicity of 2',2'-difluorodeoxycytidine and 1-0-D-arabinofuranosylcytosine. Cancer Res., 48: 4024-4031, 1988. petes with dCTP for incorporation into DNA by DNA poly- 8. Gandhi, V., and Plunkett, W. Modulatory activity of 2',2'-difluorodeoxycy- merases (9) (Fig. 8, Reaction 1). Although dFdCMP is not an tidine on the phosphorylation and cytotoxicity of arabinosyl nucleosides. absolute DNA chain terminator for DNA polymerase a or e, Cancer Res., 50: 3675-3680, 1990. its incorporation causes a pause in the replication rates of these 9. Huang, P., Chubb, S., Hertel. L. W., Grindey, G. B., and Plunkett, W. Action enzymes. Once incorporated, it is not removed by the 3' —¿ »5' of 2',2'-difluorodeoxycytidine on DNA synthesis. Cancer Res., 51: 6110- 6117, 1991. proof-reading exonuclease of DNA polymerase e, and its incor 10. Grunewald, R., Kantarjian, H., Keating, M. J., Abbruzzese, J., Tarassoff, P., poration is strongly correlated with loss of clonogenicity (9). and Plunkett, W. Pharmacologically directed design of the dose rate and schedule of 2',2'-difluorodeoxycytidine (Gemcitabine) administration in leu (b) dNTP pools decrease in cells treated with dFdC (12, 13), kemia. Cancer Res., 50: 6823-6826, 1990. an action that appears to be due to inhibition of ribonucleoside 11. Grunewald, R., Abbruzzese. J. L., Tarassoff, P., and Plunkett. W. Saturation diphosphate reducÃaseby dFdCDP (12, 14) (Fig. 8, Reaction of 2',2'-difluorodeoxycytidine 5'-triphosphate accumulation by mononuclear 2). This action decreases the cellular concentration of dCTP cells during a Phase I trial of Gemcitabine. Cancer Chemother. Pharmacol., and thus, its ratio with dFdCTP. As demonstrated in purified 27:258-262, 1991. 12. Heinemann, V., Xu, Y-Z., Chubb, S., Sen, A., Hertel, L. W., Grindey, G. B., enzyme systems (9), this is likely to increase the incorporation and Plunkett, W. Inhibition of ribonucleolide reduction in CCRF-CEM cells of dFdCTP into DNA and to potentiate its cytotoxicity. (c) by 2',2'-difluorodeoxycytidine. Mol. Pharmacol., 38: 567-572, 1990. Deoxycytidine kinase is required for dFdC phosphorylation (7, 13. Plunkett, W., Gandhi, V., Chubb, S.. Nowak. B., Heinemann, V., Mineishi, S., Sen, A., Hertel, L. W., and Grindey, G. B. 2',2'-Difluorodeoxycytidine 8) and is rate limiting to the accumulation of dFdCTP (8, 10, metabolism and mechanism of action in human leukemia cells. Nucleosides 11) (Fig. 8, Reaction 3). Because the activity of deoxycytidine & Nucleotides, 8: 775-785, 1989. kinase is regulated by dCTP (28, 29), it is likely that the 14. Baker, C. H., Banzon. J., Bollinger, J. M., Stubbe. J., Samano, V., Robins, dFdCDP-induced decrease in dNTP pools results in activation M. J., Lippert, B., Jarvi, E., and Resvick, R. 2'-Deoxy-2'-methylenecytidine and 2'-deoxy-2',2'-difluorocytidine 5'-diphosphates: potent mechanism- of the enzyme and a more rapid phosphorylation of dFdC. The based inhibitors of ribonucleotide reducÃase.J. Med. Chem.. 34:1879-1884, resulting increased dFdCTP accumulation and incorporation 1991. into DNA would be viewed as another action that potentiates 15. Heinemann, V., Hertel, L. W., Grindey. G. B., and Plunkett, W. Cellular the cytotoxicity of dFdC. (d) Deamination of dFdCMP by elimination of 2',2'-difluorodeoxycytidine 5'-triphosphate (dFdCTP). Proc. dCMP deaminase requires activation by dCTP (Fig. 8, Reaction Am. Assoc. Cancer Res., 29: 504, 1988. 16. Plunkett, W., lacoboni. S., Estey, E., Danhauser, L., Liliemark, J. O., and 4); the lowering of the dCTP pool is associated with a decrease Keating, M. J. Pharmacologically directed ara-C therapy for refractory in the enzyme activity. Because this is a major pathway of dFdC leukemia. Semin. Oncol., 12 (Suppl. 3): 20-30, 1985. nucleotide elimination, inhibition of dCMP deaminase by the 17. Shewach, D. S., and Plunkett. W. Cellular retention of 9-^-D-arabinofura- nosyladenine 5'-triphosphate and the pattern of recovery of DNA synthesis action of dFdCDP on ribonucleotide diphosphate reducÃase in Chinese hamster ovary cells. Cancer Res., 46: 1581-1584. 1986. would prolong the retention of dFdC nucleotides, an action 18. Danhauser, L.. Plunkett. W., Keating, M., and Cabanillas, F. 9-0-D-Arabi- associated with increased cytotoxicity. (e) Finally, at relatively nofuranosyl-2-fluoroadenine 5'-monophosphate pharmacokinetics in plasma high concentrations, dFdCTP is a direct inhibitor of dFdCMP and tumor cells of patients with relapsed leukemia and lymphoma. Cancer deamination by dCMP deaminase (Fig. 8, Reaction 5). Com Chemother. Pharmacol., 18: 145-152, 1986. 19. Hoard, D. E., and Ott, D. G. Conversion of mono- and oligodeoxynucleotides bined with the dFdCDP-induced reduction in the dCTP pools, to S'-triphosphates. J. Am. Chem. Soc., 87: 1785-1788, 1965. this action may cause the concentration-dependent elimination 20. Heinemann. V., and Plunkett, W. Modulation of deoxynucleotide metabo kinetics of dFdCTP in CEM cells. The relatively prolonged lism by the deoxycytidylate deaminase inhibitor 3,4,5,6-tetrahydrodeoxyuri- retention of dFdC nucleotides, which has recently been observed dine. Biochem. Pharmacol., 38: 4115-4121. 1989. 21. Brown, R. D., and Manno, J. E. ESTRIP, a BASIC computer program for in circulating leukemia cells during dFdC therapy (30), may be obtaining initial polyexponential parameter estimates. J. Pharm. Sci., 67: the pharmacological basis for the activity of this drug in exper 1687-1691. 1978. imental systems (2, 3) and in clinical trials (4-6). 22. Fridland, A., and Verhoef, V. Mechanism of ara-CTP catabolism in human 538
Downloaded from cancerres.aacrjournals.org on September 25, 2021. © 1992 American Association for Cancer Research. METABOLIC SELF-POTENTIATION OF dFdC
leukemic cells and effect of deaminase inhibitors on this process. Semin. 27 Maley, F., and Maley, G. F. The regulatory influence of allosteric effectors Oncol., /¿(Suppl 1): 262-268. 1987. on deoxycytidylate deaminase. Curr. Top. Cell. Regul.. 5: 177-228, 1972. 23. Xu, Y. Z., and Plunkett. W. Regulation of dCM P deaminase in intact CCRF- 28 Datta, N. S., Shewach, D. S., Mitchell, B. S., and Fox, I. H. Kinetic properties CEM cells. Proc. Am. Assoc. Cancer Res., 31:405. 1990. and inhibition of human T-lymphoblast deoxycytidine kinase. J. Biol. < IK-MI.. 24. Sherman, P. A., and Fyfe, J. A. Enzymatic assay for deoxyribonucleoside 264: 9359-9364, 1989. triphosphates using synthetic oligonucleotides as template primers. Anal. 29. Kim. M-Y.. and Ivés,D.H. Human deoxycytidine kinase: kinetic mechanism Biochem., 180: 222-226, 1989. and end product regulation. Biochemistry. 28:9043-9047. 1989. 25. Maley, F.. and Maley, G. F. Tetrahydrouridylate: a potent inhibitor of deoxycytidine deaminase. Arch. Biochem. Biophys., 144: 723-729. 1971. 30. Grunewald. R.. Kantarjian. H.. Du, M., Faucher. K., Tarassoff, P., and Plunkett, W. Gemcitabine(2',2'-difluorodeoxycytidine)in leukemia: a Phase 26. Rossi, M., Momparler, R. L., Nucci, R., and Scarano, E. Studies on analogues of isosteric and allosteric ligands of deoxycytidylate deaminase. Biochemistry'. I clinical, plasma, and cellular pharmacology study. J. Clin. Oncol., in press, 9:2539-2543, 1970. 1992.
539
Downloaded from cancerres.aacrjournals.org on September 25, 2021. © 1992 American Association for Cancer Research. Cellular Elimination of 2′,2′-Difluorodeoxycytidine 5′ -Triphosphate: A Mechanism of Self-Potentiation
Volker Heinemann, Yi-Zheng Xu, Sherri Chubb, et al.
Cancer Res 1992;52:533-539.
Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/52/3/533
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/52/3/533. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.
Downloaded from cancerres.aacrjournals.org on September 25, 2021. © 1992 American Association for Cancer Research.