Numerous enzymes are involved in the activation and activity of nucleoside analogs, which is a primary reason why drug discovery within this class of compounds is a very empirical process. It is difficult, if not impossible, to design a compound that will behave as needed with each of the enzymes that must be involved in its activity. Therefore, the rational drug discovery strategy with these molecules is to generate numerous analogs that are based on a thorough understanding of the biological activity of the many synthesized compounds and then evaluate them in appropriate model systems to identify the agents with the desired activities.

A thorough explanation of the rationale for the synthesis of clofarabine can be found in a review by Secrist et al. [5]. In brief, nucleosides with halogens at the two position of the adenine base are known to be very poor substrates for adenosine deaminase [6] and therefore are known to be stable in plasma. These compounds (2-halo-2′-deoxyadenosine analogs) potently inhibit tumor cell growth. In addition, 2′-flouro nucleosides have been synthesized and are known to have biological activity [7]. 2′-Halogens on nucleoside analogs are also resistant to cleavage by acid and purine nucleoside phosphorylases, another feature that would contribute to their in vivo stability. Therefore, in a search for new anticancer agents, numerous 2′-halo (F, Cl, or Br) and 2-halo (F, Cl, or Br) dAdo analogs have been synthesized and evaluated for biological activity in both in vitro and in vivo tumor models [8]. Clofarabine was the most potent compound in in vitro cell-killing assays and demonstrated excellent activity against numerous human tumor xenografts in mice [9, 10].

Human cells will salvage natural nucleosides, if available, for nucleic acid synthesis in place of de novo synthesis, and cells express numerous transporters to aid the entry of these compounds. Both equilibrative transporters (hENT1, hENT2, and hENT3) and concentrative transporters (hCNT1, hCNT2, and hCNT3) are expressed on most human cells [11]. Because of the abundance of these transporters, natural nucleosides in the cell culture will equilibrate across the cell membrane in just a few seconds. To be active, all nucleoside analogs (with the possible exceptions of deoxycoformycin and forodesine) must be able to cross cellular membranes to interact with their cellular targets. Most nucleoside analogs that are useful in the treatment of various diseases are recognized by these transporters; therefore, they also equilibrate across cell membranes very quickly. Precise transport studies have been conducted with clofarabine, and it has been shown that it is transported into cells by both human equilibrative and concentrative nucleoside transporters [12]. Cells producing hENT1 exhibited the highest uptake of clofarabine (0.14 pmole/μl), followed by hCNT2 (0.025 pmole/μl), whereas uptake in cells producing hENT2 or hCNT1 was not higher than that in cells that were nucleoside transport deficient. In oocytes expressing recombinant transporters, the efficiency of transport by hCNT3 (1.3) was much greater than that of hENT1, hENT2, or hCNT2 (0.11, 0.20, or 0.15). Although not entirely consistent, these results indicate that clofarabine can enter cells by all known transporters except hCNT1, which is selective for pyrimidine nucleosides. In addition, at high concentrations, diffusion of clofarabine directly through the membrane is also possible [12].

Comparative studies have shown [12] that cladribine had the highest efficiency of transport with hENT1, with a V_max/K_m value of 1.8 versus 0.7 for clofarabine and 0.8 for fludarabine (9-β-D-[arabinofuranosyl]-2-F-adenine). The uptake and flux of clofarabine, cladribine, and fludarabine were very similar in human renal proximal tubule cells, which express equilibrative hENT1, hENT2, and hCNT3 [13]. These results indicate that although there are some differences in transport, all of these compounds are readily transported across human membranes.

Some studies have evaluated the importance of ABCG2 to the activity of clofarabine. This ABC transporter is known as the breast cancer resistance protein and transports various metabolites, including nucleoside monophosphates, out of cells. Cells transfected with DNA constructs, resulting in overexpression of human or mouse ABCG2, had significantly reduced transport of both clofarabine and cladribine metabolites [14]. These results were correlated with decreased levels of cellular metabolites of cladribine in ABCG2-expressing cells, which were much less sensitive to clofarabine and cladribine. In these studies, the ABCG2 transporter transported both cladribine monophosphate and cladribine itself. The results of a study by Nagai et al. [15] supported these conclusions and showed that ABCG2 primarily transports clofarabine. Their results indicate that the metabolism of clofarabine and its cytotoxicity are strongly determined by the interplay between dCyd kinase and ABCG2, and they suggest that inhibition of ABCG2 will improve antitumor activity in tumors in which dCyd kinase levels are low. The role of ABC transporters in nucleoside analog metabolism and activity has recently been reviewed [16].

The metabolism of clofarabine is schematically presented in Fig. 16.3. Once inside the cell, clofarabine is converted to clofarabine triphosphate [17]. Xie and Plunkett [18] showed that at low concentrations, the primary intracellular metabolite was clofarabine monophosphate, which was approximately twice the level of clofarabine triphosphate. At high concentrations of clofarabine, the monophosphate and triphosphate levels were similar. Clofarabine diphosphate levels were very low, indicating that phosphorylation of the monophosphate of clofarabine is the rate-limiting step in its activation to the triphosphate. This characteristic is similar to that of cladribine [19] and indicates that the monophosphate kinase does not prefer molecules with substitutions at the two position of a purine nucleoside as large as a Cl atom. In the case of most other deoxynucleoside analogs used for the treatment of cancer, the nucleoside kinase is usually the rate-limiting enzyme in their activation, and the triphosphate usually accounts for more than 90% of the intracellular metabolites.

Clofarabine triphosphate accumulates at high levels in CEM cells [17, 18]: treatment of cells with low concentrations of clofarabine (1–3 μM for 2–4 h) resulted in concentrations of clofarabine triphosphate that were 50–100 μM, which were similar to intracellular dATP concentrations in untreated cells. The accumulation of clofarabine triphosphate was dose dependent up to 10 μM. Similar intracellular concentrations of clofarabine triphosphate were achieved in all phases of the cell cycle (G₁, S, and G₂-M) [18]. Clinical studies have indicated that plasma clofarabine levels of 1.5 μM are achieved at the maximally tolerated dose of clofarabine in patients with acute leukemias after a 1-h infusion, and this treatment resulted in a clofarabine triphosphate concentration of approximately 19 μM in blast cells [20].

When clofarabine was removed from the medium of K562 and CEM cell cultures, clofarabine triphosphate had a half-life in cells of 1–3 h [18, 21]. However, in circulating leukemic cells, more than 50% of the clofarabine triphosphate was still present 24 h after clofarabine transfusion [20]. The metabolism of clofarabine and cladribine has been compared in ex vivo studies with CLL and AML cells that were obtained from patients [22]. In these studies, the half-life of clofarabine triphosphate was longer than that in CEM cells but similar to that of cladribine triphosphate (7.3 vs. 4.3 h, respectively). The enzymes involved in the degradation of nucleoside triphosphates in cells have not been studied in great detail. Triphosphate metabolites of nucleoside analogs do not readily cross cell membranes and are therefore trapped in the cells in which they were created, whereas nucleosides themselves will freely distribute across the cell membrane. Therefore, the antitumor activity of these agents can be extended well beyond the time that the drug circulates in the plasma because the active metabolite is maintained in tumor cells long after the drug has disappeared from the plasma. The long retention time of clofarabine triphosphate contributes to its lasting antitumor activity, even though the plasma half-life is fairly short [20, 23].

Cytoplasmic 5′-nucleotidase cN-II is one of the seven 5′-nucleotidases found in mammals. This enzyme catalyzes the removal of the 5′-phosphate from purine 5′-monophosphates and is involved in the maintenance of cellular nucleotide pools. Therefore, it can also degrade analogs of purine nucleoside monophosphates and could theoretically interfere with the activation of clofarabine. In cell culture experiments with CEM and RL cells, inhibition of this enzyme did not affect the cytotoxicity of clofarabine and had only modest effects on the induction of apoptosis [24], indicating that it has only a modest effect on the metabolism of clofarabine when clofarabine is constantly present in the cell culture medium. The results of these studies suggest that the monophosphate pool maintained by the dCyd kinase/5′-nucleotidase activity ratio in the presence of excess extracellular clofarabine is sufficient to maintain cytotoxic clofarabine triphosphate levels in the cell. However, it is possible that such inhibitors would have a more dramatic effect on metabolism after the removal of clofarabine from the external environment, which happens in vivo. As clofarabine triphosphate is degraded to the clofarabine monophosphate, inhibition of this enzyme could prevent further degradation, resulting in re-phosphorylation by the monophosphate kinase and an increase in the intracellular half-life of clofarabine triphosphate.

As expected for a dAdo analog, clofarabine was incorporated into DNA, but not RNA, indicating that it is a DNA-directed drug [18]. At low concentrations, clofarabine is mostly incorporated into internal positions in the DNA (>80%), suggesting that DNA synthesis is not inhibited by the incorporation of clofarabine monophosphate into the growing DNA strand. However, at high concentrations (1 or 10 μM), much more clofarabine is found at terminal positions in DNA. At 10 μM, 65% of the clofarabine in DNA is found at terminal positions, which suggests that DNA synthesis is inhibited at the site of incorporation.

The primary enzyme involved in the activation of clofarabine in tumor cells is dCyd kinase [17, 18, 22]. Clofarabine is a very good substrate for this enzyme, with K_m and V_max values that are similar to those of dCyd [21] but much better than those of dAdo (the catalytic efficiency of dAdo is only 0.3% that of dCyd). Clofarabine was modestly better substrate for dCyd kinase than was cladribine. Its catalytic efficiency was three to four times that of cladribine, with a V_max that was ten times greater. The X-ray crystal structure of dCyd kinase with clofarabine in the active site has been determined [25]. The results of these studies indicate that the conformation of the enzyme/clofarabine complex was similar to structures of the pyrimidine-bound complexes and that the interactions between the 2-Cl group and its surrounding hydrophobic residues contribute to the high catalytic efficiency of dCyd kinase with clofarabine.

Decreased dCyd kinase activity has been shown to result in resistance to clofarabine in various cell culture systems [22, 26–29]. Decreased histone acetylation [28], but not methylation of the dCyd kinase gene [26, 28, 29], was associated with decreased dCyd kinase expression. In HL-60 cells, the decreased activation of clofarabine was associated with decreased expression of dCyd kinase, deoxyguanosine kinase, and hENT1, hENT2, and hCNT3 genes [30]. Studies have shown that there are genetic variations in the expression of dCyd kinase activity in patient populations and suggest that these variations are responsible for the variable rates of activation of clofarabine and other nucleoside analogs [31].

Treatment of cells with clofarabine and other nucleoside analogs can enhance dCyd kinase activity in many, but not all, cell lines [32–34]. Others have shown that inhibition of DNA synthesis can enhance dCyd kinase activity [35, 36] by phosphorylating dCyd kinase at serine 74 [35, 37, 38]. The enhancement of dCyd kinase activity by clofarabine can be exploited to improve antitumor activity by combining clofarabine with other nucleoside analogs that are activated by dCyd kinase [33, 34, 39–41].

Clofarabine is also a good substrate for deoxyguanosine kinase [42], an enzyme that is expressed in mitochondria. The catalytic efficiency of clofarabine with this enzyme is similar to that of both deoxyguanosine and cladribine, although both the K_m and V_max values for the two deoxyadenosine analogs are 10–20 times greater than are those for deoxyguanosine. The contribution of deoxyguanosine kinase to the phosphorylation of clofarabine in cells is low due to the much higher expression of dCyd kinase activity in most cell types [43, 44], but it could be important in cells that express low amounts of dCyd kinase. Resistance to clofarabine in CEM cells was correlated with decreased dCyd kinase activity but not with decreased deoxyguanosine kinase [22, 27].

The mechanism of action of clofarabine resembles that of other dAdo analogs and is schematically presented in Fig. 16.3 . Below, we describe the primary actions of this nucleoside analog.