Introduction

Top Summary Introduction Experimental Procedures Results Discussion References

2-Chloro-9-(2-deoxy-2-fluoro--D-arabinofuranosyl)adenine [Cl-F()-dAdo] is a promising new antitumor agent that has structural similarities to both 2-chloro-2'-deoxyadenosine (Cl-dAdo; cladribine) and 2-fluoro-9-(-D-arabinofuranosyl)adenine (F-araA or fludarabine), both of which are Food and Drug Administration-approved drugs for the treatment of certain types of cancer (Bonnet and Robins, 1993). In the past, Cl-F()-dAdo has been abbreviated as Cl-F-araA and CAFdA. However, for clarity of presentation, we have changed its abbreviation. In addition to its activity against hematologic malignancies (Carson et al., 1992; Waud et al., 1992), Cl-F()-dAdo is also active against solid tumors such as renal and colon carcinomas (W. R. Waud, personal communication). The mechanism of action of Cl-F()-dAdo has been extensively studied (Parker et al., 1991; Carson et al., 1992; Xie and Plunkett, 1995, 1996) and is similar to that of Cl-dAdo and F-araA (Parker et al., 1991; Plunkett and Saunders, 1991). These agents are metabolized via 2'-deoxycytidine (dCyd) kinase to their respective triphosphates, which are potent feedback inhibitors of ribonucleotide reductase and are used as substrates by DNA polymerases.

During the development of Cl-F()-dAdo, 2-chloro-9-(2-deoxy-2-fluoro--D-ribofuranosyl)adenine [Cl-F()-dAdo], and 2-chloro-9-(2-deoxy-2,2-difluoro--D-ribofuranosyl)adenine [Cl-diF()-dAdo] (Fig. 1) were also synthesized and evaluated for antitumor activity. In contrast to Cl-F()-dAdo, these two compounds were not potent inhibitors of cell growth and did not have significant selectivity in animal antitumor models (W. R. Waud and L. W. Hertel, personal communications). Cl-diF()-dAdo also has structural features similar to 2',2'-difluoro-dCyd (gemcitabine), another new antitumor nucleoside analog recently approved by the Food and Drug Administration for the treatment of pancreatic cancer (Plunkett et al., 1996). Because of the similarities in structure among Cl-F()-dAdo, Cl-F()-dAdo, and Cl-diF()-dAdo, it was important to our understanding of the mechanistic implications of specific structural variations to determine why these changes resulted in very different antitumor activities. Of particular interest was the use of this information for our detailed understanding of the biochemical activities of Cl-F()-dAdo that are important to its highly selective antitumor activity.

View larger version (21K): [in this window] [in a new window]   Fig. 1.   Structures of Cl-dAdo, Cl-F()-dAdo, Cl-F()-dAdo, and Cl-diF()-dAdo.

	Experimental Procedures

Top Summary Introduction Experimental Procedures Results Discussion References

Materials. Cl-F()-dAdo and Cl-F()-dAdo were synthesized in our laboratories (Secrist et al., 1988; Thomas et al., 1994), and Cl-diF()-dAdo was synthesized at Eli Lilly and Company (Indianapolis, IN) (Grindey and Hertel, 1991, 1995). [methyl-³H]Thymidine(dThd) (69 Ci/mmol), [5-³H]uridine (Urd) (20 Ci/mmol), [5-³H]cytidine (Cyd) (25 Ci/mmol), [5-³H]dCyd (26 Ci/mmol), and [4,5-³H]leucine (20 Ci/mmol) were obtained from Moravek Biochemicals (Brea, CA). CEM cells (American Type Culture Collection, Rockville, MD) were grown in RPMI 1640 medium (GIBCO BRL, Gaithersburg, MD) containing 10% fetal bovine serum (Atlanta Biologicals, Atlanta, GA), 1 mg/ml sodium bicarbonate, 10 units/ml penicillin, 10 µg/ml streptomycin, and 50 µg/ml gentamycin. Cells were routinely checked for the presence of mycoplasma and were discarded if contaminated. The Aquasil C18 and Partisil-10 strong anion exchange (SAX) HPLC columns were obtained from Keystone Scientific Inc. (Bellefonte, PA). Other compounds were of standard analytical grade.

Measurement of DNA, RNA, and Protein Synthesis in Intact Cells. The effect of compounds on the incorporation of radiolabeled precursors ([5-³H]Urd, [methyl-³H]dThd, or [4,5-³H]leucine) into RNA, DNA, or protein was determined as described previously (Hershko et al., 1971; Bennett et al., 1978). The incorporation of Urd into RNA is determined by subtracting the incorporation of radiolabel into the alkali-stable/acid-precipitable fraction (DNA) from the total acid-precipitable fraction (DNA plus RNA). [5-³H]Urd is primarily incorporated into RNA but also can be incorporated into DNA as dCyd. The incorporation of dThd and leucine into acid-precipitable material is a measure of their incorporation into DNA and protein, respectively.

Extraction and Analysis of Acid-Soluble Nucleotide Pool. Cells were collected by centrifugation and resuspended in ice-cold 0.5 M perchloric acid. The samples were centrifuged at 12,000g for 20 min, and the supernatant fluid was removed and neutralized with 1 M potassium phosphate, pH 7.4, and 4 M KOH. KClO₄ was removed by centrifugation, and a portion of the supernatant fluid was injected onto a Partisil-10 SAX column. Elution of the nucleotides was accomplished with a 50-min linear gradient from 5 mM NH₄H₂PO₄ (pH 2.8) to 750 mM NH₄H₂PO₄ (pH 3.7) buffer with a flow rate of 2 ml/min. The nucleotides were detected by measurement of the UV absorbance at 260 nm.

Measurement of dCyd Kinase Activity. dCyd kinase was purified 24,000-fold from MOLT-4 cells to greater than 95% purity. The procedures for the isolation and for assay of nonlabeled nucleosides as substrates have been described elsewhere (Shewach et al., 1992). The Michaelis-Menten parameters were determined from linear double reciprocal plots of 1/velocity versus 1/concentration of the substrate. The best line was determined using linear regression from at least five data points, and the K_m and V_max values were determined from the x- and y-intercepts.

	Results

Top Summary Introduction Experimental Procedures Results Discussion References

Inhibition of CEM Cell Growth. Cl-F()-dAdo was approximately 50-fold more potent as an inhibitor of CEM cell growth than either Cl-F()-dAdo or Cl-diF()-dAdo (Fig. 2). The IC₅₀ values of Cl-F()-dAdo, Cl-F()-dAdo, and Cl-diF()-dAdo were 6 ± 1, 313 ± 47, and 317 ± 78 nM, respectively (mean ± S.D. from three experiments). The addition of dCyd completely protected against the cytotoxicity of these three compounds (data not shown), which suggested that these compounds were phosphorylated to active metabolites by dCyd kinase.

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Cytotoxicity of Cl-F()-dAdo, Cl-F()-dAdo, and Cl-diF()-dAdo to CEM cells. CEM cells were incubated with varying concentrations of compounds for 72 h, and the cell numbers were determined using a Coulter Counter. The results shown are from a representative experiment from a total of three that were done.

Phosphorylation in CEM Cells. Differential activation of these compounds could explain their differences in cytotoxicity to CEM cells; therefore, the level of conversion of each compound to its respective triphosphate analog was determined (Fig. 3). A 2-h incubation with 25 µM Cl-F()-dAdo, Cl-F()-dAdo, or Cl-diF()-dAdo resulted in 75 ± 13, 67 ± 14, or 76 ± 9 pmol of triphosphate analog formed/10⁶ cells (mean ± S.D. from three experiments). For each of the compounds, the accumulation of triphosphate increased during an 8-h incubation period and the amount of triphosphate formed were directly related to the concentration of compound in the culture medium (1-50 µM; data not shown). Because of the much smaller peaks formed at lower concentrations, GTP and the other ribonucleotides were removed from the extracts by boronate column chromatography (Shewach, 1992) before analysis by SAX HPLC. The conversion of these nucleoside analogs to their respective nucleoside triphosphates was inhibited by incubation with dCyd (data not shown), which again indicated that dCyd kinase was the enzyme primarily responsible for the metabolism of these agents in CEM cells.

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Phosphorylation of Cl-F()-dAdo, Cl-F()-dAdo, and Cl-diF()-dAdo in CEM cells. CEM cells were incubated with no drugs (a) or 25 µM Cl-F()-dAdo (b), Cl-F()-dAdo (c), or Cl-diF()-dAdo (d) for 2 h. The acid-soluble fraction was collected and analyzed by SAX HPLC as described in Materials and Methods. This is a representative experiment from a total of three that were done.

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Half-life for disappearance of Cl-F()-dATP, Cl-F()-dATP, and Cl-diF()-dATP. CEM cells were treated with 25 µM Cl-F()-dAdo, Cl-F()-dAdo, or Cl-diF()-dAdo. After 2 h, the cells were collected by centrifugation and resuspended in fresh media that contained none of the above compounds. At 0, 1, 2, 4, and 6 h after resuspension, a sample of the cell extract was collected, and the amount of analog triphosphate was determined as described in Materials and Methods. The results shown are from a representative experiment from a total of three that were done.

Substrate Activity with Purified dCyd Kinase. The Michaelis-Menten constants for these three compounds were determined with pure dCyd kinase isolated from Molt-4 cells (another cell line derived from human T cells) (Table 1). Even though the K_m values for the dAdo analogs were much higher than the K_m value for dCyd, their maximum velocities were also much higher, so the relative efficiencies were closer to that for dCyd. Of the three dAdo analogs, Cl-F()-dAdo was the best substrate, but it was only 2-fold better than Cl-diF()-dAdo and 7-fold better than Cl-F()-dAdo. The K_m value for Cl-F()-dAdo was 10-fold greater than the K_m value for Cl-F()-dAdo or Cl-diF()-dAdo. These differences in the substrate parameters between Cl-F()-dAdo and Cl-diF()-dAdo were not sufficient to account for the differences in the cytotoxicities of these two agents. However, the 7-fold difference in the preference of dCyd kinase for Cl-F()-dAdo versus Cl-F()-dAdo could help explain some of the differences in potency of these two compounds.

View this table: [in this window] [in a new window]   TABLE 1 Substrate characteristics of Cl-F()-dAdo, Cl-F()-dAdo, and Cl-diF()-dAdo with dCyd kinase isolated from MOLT-4 cells

Inhibition of dCyd Kinase Activity in Crude CEM Cell Extracts. The concentrations of Cl-F()-dAdo, Cl-F()-dAdo, and Cl-diF()-dAdo required to inhibit the phosphorylation in crude CEM extracts of 1 µM dCyd by 50% were 110 ± 23, 225 ± 72, and 105 ± 22 µM, respectively (mean ± S.D. from three experiments, data not shown). This result indicated that the affinity of these compounds for dCyd kinase in CEM cells were similar and supported our conclusion that the differences in the activation of the compounds were not responsible for the differences in cytotoxicity.

Inhibition of Macromolecular Synthesis. Cl-F()-dAdo was 10- to 20-fold more potent in its ability to inhibit the incorporation of dThd into DNA than either Cl-F()-dAdo or Cl-diF()-dAdo (Fig. 5, IC₅₀ = 0.05 versus approximately 1 µM). Furthermore, both Cl-F()-dAdo and Cl-diF()-dAdo inhibited the incorporation of Urd into RNA and leucine into protein at concentrations that were similar to those required to inhibit dThd incorporation into DNA. At 1, 2, and 4 µM, Cl-F()-dAdo, RNA, and protein syntheses were inhibited to a similar degree as that seen with Cl-F()-dAdo and Cl-diF()-dAdo (data not shown).

View larger version (15K): [in this window] [in a new window]   Fig. 5.   Effect of Cl-F()-dAdo, Cl-F()-dAdo, and Cl-diF()-dAdo on DNA, RNA, and protein synthesis. CEM cells were incubated with varying concentrations of Cl-F()-dAdo, Cl-F()-dAdo, or Cl-diF()-dAdo. Radiolabeled precursors of DNA ([methyl-3H]dThd), RNA ([5-3H]Urd), or protein ([4,5-3H]leucine) were added to the cell cultures 30 min hours after the addition of the compounds, and cell samples were taken 1, 2, 3, or 4 h after the addition of radiolabel to determine their incorporation into RNA, DNA, or protein as described in Materials and Methods. This experiment was repeated one time with similar results.

Inhibition of [³H]Cyd Incorporation into dCyd Nucleotides. Because the inhibition of DNA synthesis by Cl-F()-dAdo results from the inhibition of both ribonucleotide reductase and DNA polymerases (Parker et al., 1991), the effect of these compounds on the incorporation of [³H]Cyd into DNA was determined in an attempt to evaluate the inhibition of ribonucleotide reductase by these agents (Fig. 6). The concentrations of Cl-F()-dAdo, Cl-F()-dAdo, and Cl-diF()-dAdo that were required to inhibit the incorporation of Cyd into DNA by 50% were 0.26 ± 0.1, 3.2 ± 1.8, and 9.2 ± 2.5 µM, respectively (mean ± S.D. from three experiments). Cl-F()-dAdo was approximately 10-fold more potent than Cl-F()-dAdo and 30-fold more potent than Cl-diF()-dAdo in its ability to inhibit ribonucleotide reductase.

View larger version (17K): [in this window] [in a new window]   Fig. 6.   Inhibition of [3H]Cyd incorporation into DNA. CEM cells were incubated with 10 nM [3H]Cyd plus various concentrations of the nucleoside analogs. After 10 min, the incorporation of [3H]Cyd into the alkali-stable/acid-insoluble fraction (DNA) was determined as described in Materials and Methods. The results shown are from a representative experiment from a total of three that were done.

	Discussion

Top Summary Introduction Experimental Procedures Results Discussion References

The inhibition of ribonucleotide reductase and DNA polymerase activities is believed to be responsible for the antitumor activity of Cl-F()-dAdo (Parker et al., 1991; Carson et al., 1992; Xie and Plunkett, 1995, 1996). Cl-F()-dAdo is phosphorylated in human cells to Cl-F()-dATP, which potently inhibits ribonucleotide reductase as an allosteric regulator. Cl-F()-dATP is also used as a substrate by DNA polymerase , and its incorporation into the 3' end of the growing DNA chain causes disruption of further DNA synthesis. Because Cl-F()-dATP competes with dATP for incorporation into DNA, the decline in dNTP levels (including dATP) due to the inhibition of ribonucleotide reductase potentiates the activity of Cl-F()-dATP against DNA polymerases. The mechanism of action of Cl-F()-dAdo is very similar to that of Cl-dAdo. Cl-dAdo is phosphorylated to Cl-dATP, which is also a potent inhibitor of both ribonucleotide reductase and DNA polymerases (Parker et al., 1991). One difference between these compounds is that the incorporation of Cl-F()-dAdo into the 3' end of a DNA chain results in a greater disruption of DNA elongation than does the incorporation of Cl-dAdo (Parker et al., 1991). The incorporation of Cl-dAdo into DNA has been shown to disrupt various activities that involve DNA (Hentosh and McCastlain, 1991; Hentosh and Grippo, 1994a, 1994b; Hentosh and Tibudan, 1995). Cl-dATP has recently been shown to substitute for dATP in the activation of caspase-3 in a cell-free system (Leoni et al., 1998), and it is possible that these compounds could also substitute for dATP in the activation of this apoptotic pathway. However, the importance of this action of Cl-dAdo and related compounds to their ability to kill cells has yet to be determined.

In the current work, we compared the biochemical pharmacology of Cl-F()-dAdo with that of two compounds that are very similar in structure but much less potent as inhibitors of CEM cell growth and do not exhibit selective antitumor activity in mice. We determined that the reason for the differences in potency in CEM cells was primarily due to differences in the inhibition of ribonucleotide reductase activity by these compounds rather than to differences in their activation. Differences in the interaction of the 5'-triphosphate analogs of these compounds with DNA polymerases could also contribute to the differences seen in the potencies of these agents. However, the fact that most of the differences in cytotoxicity in CEM cells could be explained by the differences in the inhibition of ribonucleotide reductase activity suggested that the DNA-directed activities of these compounds (direct inhibition of DNA elongation due to their incorporation into DNA) are of secondary importance in the inhibition of CEM cell growth. The antitumor activity of hydroxyurea, an inhibitor of ribonucleotide reductase, indicates that the depression of dNTP levels alone is sufficient to result in antitumor activity.

Cl-F()-dAdo was a poorer substrate for dCyd kinase isolated from MOLT-4 cells than was Cl-F()-dAdo. There were no large differences in either the accumulation or degradation of Cl-F()-dATP and Cl-F()-dATP in intact CEM cells, and there was only a 2-fold difference in the inhibition of dCyd kinase activity in crude CEM cell extracts. The reason for these small discrepancies between the purified enzyme and studies with intact cells or crude extracts is not known. We conclude from all of these studies that decreased metabolism of Cl-F()-dAdo may play a small role in its low potency in CEM cells but that the primary reason for the relatively low potency of Cl-F()-dAdo is still due to its poor inhibition of ribonucleotide reductase activity.

2',2'-diF-dCyd (gemcitabine) also inhibits ribonucleotide reductase activity due to the interaction of 2',2'-diF-dCDP with the substrate active site, instead of allosteric inhibition by 2',2'-diF-dCTP (Heinemann et al., 1990). 2',2'-diF-dCDP is approximately 100-fold less potent than either Cl-F()-dATP or Cl-dATP in its ability to inhibit ribonucleotide reductase activity (Heinemann et al., 1990; Parker et al., 1991). These studies suggest that Cl-diF()-dAdo may inhibit ribonucleotide reductase activity as a diphosphate analog instead of the triphosphate analog. However, it is important to point out that dATP is an important negative regulator of ribonucleotide reductase activity, whereas dCTP is not (Nutter and Cheng, 1984).

The results of this work indicated that the cytotoxicity of Cl-F()-dAdo in CEM cells is primarily due to its inhibition of DNA synthesis, whereas Cl-F()-dAdo and Cl-diF()-dAdo inhibited RNA and protein synthesis at concentrations similar to those required to inhibit DNA synthesis. Because nonreplicating, as well as replicating cells, synthesize RNA and proteins, these results suggest that Cl-F()-dAdo and Cl-diF()-dAdo would be toxic to nonreplicating host cells, which could explain why these two agents have poor antitumor activity in vivo. The enzyme or enzymes inhibited by these agents (or one of their metabolites) that results in the inhibition of RNA and protein synthesis are not known. Xie and Plunkett (1995) have found that Cl-F()-dAdo can be incorporated into RNA at approximately 1% of the rate of its incorporation into DNA, which suggests that the incorporation of these molecules into RNA could be responsible for the inhibition of RNA and protein syntheses. It is also possible that these agents interfere with some ATP-requiring enzyme. The concentrations of compound required to inhibit protein and RNA synthesis were similar for all three compounds. The primary difference in the mechanism of action of these molecules is the ability of Cl-F()-dAdo to inhibit DNA synthesis at concentrations much lower than those required to inhibit either protein or RNA synthesis.