j631
`
`25
`Clofarabine: From Design to Approval
`John A. Secrist III, Jaideep V. Thottassery, and William B. Parker
`
`25.1
`Introduction
`
`Many different classes of compounds have been found to have utility in treating a
`variety of different types of cancers. These compounds act through a variety of
`different mechanisms, in many cases targeting metabolic differences between
`normal and cancerous cells, and more recently targeting cancer-specific targets and
`pathways. An examination of the drugs that have been approved by governments
`worldwide demonstrates that antimetabolites – compounds that affect the pathways
`leading to nucleic acids – represent a rich source of anticancer drugs. Almost all of
`those approved drugs are either nucleosides or compounds that are converted to
`nucleosides or nucleotides after administration to patients. In the United States, the
`list of FDA-approved antimetabolites of this type includes 5-fluorouracil (colorectal,
`breast, stomach, and pancreatic carcinomas), 6-thioguanine (acute non-lymphocytic
`leukemias), 1-b-D-arabinofuranosylcytosine [acute lymphocytic leukemia (ALL), and
`0
`acute myelocytic leukemia (AML)], 5-fluoro-2
`-deoxyuridine (metastatic colon cancer),
`fludarabine phosphate [chronic lymphocytic leukemia (CLL)], 2-deoxycoformycin
`(hairy cell leukemia), cladribine (hairy cell leukemia), gemcitabine [pancreatic cancer,
`non-small cell lung cancer (NSCLC)], capecitabine (metastatic colorectal and breast
`cancer), nelarabine (T-cell acute lymphoblastic leukemia and lymphoma), decitabine
`(myelodysplastic syndrome), and clofarabine (pediatric ALL). Both fludarabine phos-
`phate andclofarabine were discovered andpushed forwardpreclinically attheSouthern
`Research Institute, and both are products of our preclinical optimization process for
`nucleoside analogues.
`This chapter will focus on the development of clofarabine, and will present that
`development from our viewpoint as preclinical scientists. A recent review covering
`many aspects of the development of clofarabine is recommended to the reader for
`additional details [1].
`
`Modified Nucleosides: in Biochemistry, Biotechnology and Medicine. Edited by Piet Herdewijn
`Copyright Ó 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
`ISBN: 978-3-527-31820-9
`
`1
`
`GIL2006
`I-MAK, INC. V GILEAD PHARMASSET LLC
`IPR2018-00122
`
`

`

`632j 25 Clofarabine: From Design to Approval
`
`25.2
`Clofarabine: The Background
`
`Before examining the chronology of the discovery and development of clofar-
`abine, it is important first to identify its current and potential future uses, as well
`as the companies involved. Clofarabine was approved for the treatment of
`pediatric ALL in the US in December 2004, and in Europe in May 2006. It has
`been granted orphan drug status in both the US and Europe. Early efforts toward
`pushing the drug into clinical trials, beginning in 1992, were initiated at the M.
`D. Anderson Cancer Center (MDACC) in Houston, Texas, and included clin-
`icians Drs. J. Freireich, M. Keating and H. Kantarjian, as well as pharmacologists
`Drs. W. Plunkett and V. Gandhi. The initial licensing of the drug by Southern
`Research Institute was to the Eurobiotech Group in 1998, which utilized MDACC
`for the clinical trials. The lag time evident in the above dates is truly unfortunate
`from the standpoint of cancer patients who might have benefited from the drug.
`The difficulties in licensing clofarabine stemmed from two views prevalent in the
`pharmaceutical industry at the time: (i) clofarabine was just another fludarabine,
`and had little chance of making a mark on its own; and (ii) future cancer drugs
`needed to focus on solid tumors such as colorectal, breast, prostate and lung cancer,
`and nucleosides were of little interest in that regard. Moreover, even if they
`had some activity, the market size was too small to be of interest to the larger
`companies. The CEO of Eurobiotech (most recently called Bioenvision), Dr.
`Christopher Wood, understood the properties of fludarabine phosphate, and
`believed that clofarabine had properties that might make it significantly better,
`and on that basis he was eager to proceed. Similarly, the team at MDACC also
`believed that clofarabine had the potential to take its own place among anticancer
`treatments. The eventual approval of clofarabine was aided immeasurably by the
`commitment of these people.
`
`25.3
`The Beginnings
`
`During the early 1980s, two types of nucleoside were found to have very promising
`0
`selectivity in animal models. These compounds were the 2-halo-2
`-deoxyadenosines
`(1), with the halogen being fluorine, chlorine or bromine, and the corresponding b-D-
`0
`arabinofuranosyl analogues (2). The 2
`-deoxy compounds had been prepared in
`several different laboratories [2–4] and examined in various cell lines as potential
`0
`anticancer drugs. In our laboratories, we examined all three of the 2
`-deoxy com-
`pounds in a series of experiments in the then-standard L1210 mouse leukemia model
`system. Interestingly, all three compounds had excellent selectivity, and a summary
`of that previously unpublished data is shown in Table 25.1 [5]. Although it is easy to
`see that all three had some selectivity, the chloro and bromo analogues appeared to be
`the most promising, with some cures seen, and these results warranted further
`investigation.
`
`2
`
`

`

`25.3 The Beginningsj633
`
`Over the next few years, the 2-chloro compound was examined in further detail
`through a collaboration between John Montgomery at Southern Research Institute
`and Dennis Carson at the University of California at San Diego, while the 2-bromo
`compound was further examined by Raymond Blakley and his colleagues at St. Jude
`Children’s Hospital. As events unfolded, the 2-chloro compound (cladribine) was
`eventually approved in 1992 by the FDA for the treatment of hairy cell leukemia.
`In the 2-halo-ara-A series, the fluoro, chloro, and bromo compounds were prepared
`at Southern Research Institute and elsewhere [2, 6–10], and all three had some
`activity [11, 12]. The data demonstrated that the fluoro compound was significantly
`better than the other two, and consequently it was carried forward, eventually being
`0
`approved in 1991 as the 5
`-phosphate (fludarabine phosphate), a prodrug form
`developed to aid solubility, for the treatment of CLL.
`During the first half of the 1980s, some preliminary data were acquired on these
`compounds which, when combined with previous information regarding the physi-
`cal properties of the two series, suggested that some improvements in the structures
`could be made that might have a significant effect on their potential clinical utility. At
`the time, it was of course not known whether any of these compounds would become
`approved, and the quest was to prepare compounds that would have enhanced
`properties that might either achieve approval if the earlier compounds did not, or
`might be next-generation compounds with more attractive properties than any earlier
`compounds that did achieve approval.
`
`Table 25.1 Response of intraperitoneally (i.p.) implanted L1210
`0
`-deoxyadenosines.
`leukemia to 2-halo-2
`
`Compound
`
`Optimal
`i.p. dose
`(LD10, mg
`
`1 dose1)a)
`kg
`
`Total dose
`1)
`(mg kg
`
`Median %
`ILSb) (dying
`mice only)
`
`Net log10
`cell killc)
`
`Tumor-free
`survivors/
`total
`
`F-dAdo
`25
`600
`Cl-dAdo
`20
`480
`Br-dAdo
`40
`960
`a) Treatment schedule was q3 h  8, Days 1, 5, and 9.
`b) Median day of death of tumored control mice (105 cells) was 8 days. ILS, increase in lifespan.
`c) Net log10 reduction in the tumor cell population between the beginning and the end of therapy,
`based on the median day of death of the mice that died.
`
`0/10
`5/10
`3/6
`
`þ118
`þ150
`þ125
`
`þ0.5
`þ2.9
`þ1.1
`
`3
`
`

`

`634j 25 Clofarabine: From Design to Approval
`
`Thus, attention was focused on three properties while seeking to improve on the two
`seriespresentedabove.Itiseasiesttoconsiderthecharacteristicsthatmightbeimproved
`by focusing on the two compounds that were eventually approved, fludarabine and
`cladribine. In the case of fludarabine phosphate, when administered to an animal it is
`rapidly cleaved to fludarabine, which enters the cells and is further metabolized [13].
`0
`In the case of cladribine, it is well known that 2
`-deoxy compounds in the purine
`series are susceptible to chemical cleavage at low pH, and thus a loss of potency
`through hydrolytic cleavage is clearly an issue with cladribine. In addition, cleavage of
`the glycosidic bond by phosphorylases is another means of loss of potency. With both
`types of cleavage, 2-chloroadenine would be generated, which is a compound with
`only modest toxicity.
`For fludarabine, the chemical hydrolysis of the glycosidic bond is not a significant
`0
`problem as the presence of the 2
`-hydroxyl group provides significant stability,
`although there is a loss of potency through some phosphorylase cleavage [13–16].
`In the case of fludarabine, this process generates 2-fluoroadenine, which is a highly
`undesirable metabolite. This purine is readily metabolized up to 2-fluoro-ATP,
`which is an extremely toxic but unselective compound, and so its systemic generation
`could present a concern. In recent years, attention has been focused on utilizing
`gene therapy approaches to generate this toxin in tumor cells in a selective manner [17].
`The other mechanism of loss of potency that can occur with adenine derivatives is
`enzymic deamination, which is carried out by adenosine deaminase at the nucleoside
`level, and by AMP deaminase at the monophosphate level. The incorporation of a
`2-halogen into the adenine ring of a nucleoside confers significant resistance to
`deamination as compared to the parent adenine compounds [11, 12]. Many evalua-
`tions have been carried out examining the ability of the 2-haloadenine nucleosides to
`serve as substrates for adenosine deaminase, and although they are highly resistant to
`deamination, all have at least some substrate activity. The order of deamination is
`F > Cl > Br, and within that order the 2-fluoro compounds are significantly more
`susceptible to deamination than the other two. Thus, in the case of fludarabine there
`is a minor loss of potency through some deamination [13–16], a pathway that is not a
`significant problem with cladribine.
`The other key issue with regard to nucleoside analogues is their activation to an
`active metabolite which, in the vast majority of cases, is the nucleoside triphosphate
`(NTP). In general, nucleosides exert their effects on the biosynthetic pathway leading
`0
`to DNA, and thus, analogues of 2
`-deoxynucleosides are typically of more interest
`than the building blocks of RNA, although ribonucleoside analogues were also
`prepared. It was determined that the addition of a 2-halogen did not prevent the
`0
`phosphorylation of some 2
`-deoxynucleoside analogues, and many laboratories have
`determined that the major enzyme carrying out the initial phosphorylation is
`generally deoxycytidine kinase [18–20]. Another key observation was that nucleosides
`with an arabino configuration often were also substrates of deoxycytidine kinase.
`The other enzymes carrying out conversion of monophosphates to their di- and
`triphosphate metabolites were in general less discriminating, and the majority of
`0
`nucleosides that could be metabolized to the 5
`-monophosphate were converted at a
`meaningful rate to the triphosphate.
`
`4
`
`

`

`25.4 The Next Generation of Compoundsj635
`
`The above-described information relates to the situation during the early to mid-
`1980s as ways were sought to improve on the activity of this class of potential
`anticancer nucleosides. The set of simple conclusions drawn from the above
`information can be summarized as follows:
`
`. A 2-halogen in an adenine ring analogue dramatically reduces deamination,
`but in general will allow phosphorylation, depending upon the carbohydrate
`attached.
`. A 2-chloro or 2-bromoadenine ring is more desirable than a 2-fluoro, based upon
`the high toxicity of any 2-fluoroadenine that may be generated, and also based upon
`its increased ability of 2-fluoroadenine-containing nucleosides to serve as sub-
`strates for adenosine deaminase.
`0
`in the arabino configuration – one that will significantly
`. A stabilizing group at C-2
`reduce both phosphorylase cleavage and hydrolytic cleavage of the glycosidic
`bond – is highly desirable.
`
`25.4
`The Next Generation of Compounds
`
`Over the years, a highly efficient system was developed for the rapid examination of
`new compounds in our anticancer drug discovery program, which was strongly
`supported by the US National Institutes of Health. Whenever a new compound had
`been prepared and properly characterized, it was submitted for an evaluation of its
`cytotoxicity in a small series of cancer cell lines, with generally six or seven such lines
`being derived from various types of human tumors. Typically, the results were
`available in a few weeks. In all of these cases, the corresponding human tumor
`xenograft mouse model was available if a compound exhibited significant cytotoxicity.
`The main challenge was to prepare sufficient material for evaluation in a mouse
`model, once it had been learned that such an examination was warranted based upon
`the cytotoxicity profile. When sufficient material became available for such an initial
`evaluation, the compound was submitted and generally placed into a test system
`within a month. In parallel, the mechanistic evaluation of compounds of potential
`interest was started in our biochemistry laboratories. Together, this research effort
`provided us with the basic information on the activation of new nucleosides to the
`various phosphorylated metabolites, their effects on DNA, RNA and protein synthe-
`sis, and also specific information on key enzymes. Feedback from both the in-vitro
`and in-vivo evaluations was thus rapidly available, and we were able quickly to adjust
`our target structure list based upon this iterative feedback. This simple system
`prevented us from spending too much time on the synthesis of series of compounds
`that did not show promise as anticancer drugs.
`By utilizing this efficient system, a variety of compounds was evaluated relatively
`rapidly. The major efforts revolved around carbohydrate modifications with the
`2-haloadenines as the bases, and on similar compounds with the nitrogen base
`somewhat altered.
`
`5
`
`

`

`636j 25 Clofarabine: From Design to Approval
`
`Examples of ring-altered compounds included 2-fluoro-8-azaadenine nucleosides
`such as 3 [21], and a ring-fluorinated analogue 4 of formycin A [22]. Unfortunately,
`however,thetwocompounds3and4hadcharacteristicsthatpreventedthemfrombeing
`of therapeutic use. For example, compound 3 was not significantly cytotoxic, presum-
`ably because it was not amenable to initial phosphorylation, while 4 was converted to
`the monophosphate, but not further converted to the di- and triphosphate levels.
`
`In the case of carbohydrate-modified nucleosides, the initial focus – as noted above
`0
`– was at the 2
`-position. It was felt that building in halogen atoms as well as certain
`0
`other groups with significant electronegativity at the 2
`-position would accomplish
`several goals. First, these compounds should impart significant hydrolytic stability of
`the glycosidic bond. Second, this alteration might well reduce the ability of these
`compounds to serve as substrates for phosphorylases. Therefore, attention was
`0
`redirected towards preparing compound series 5, which incorporated a 2
`-bromine,
`chlorine, fluorine, azido or amino group. Another related compound, 6, which also
`has anticancer activity, has been prepared by the Matsuda group and includes a
`0
`2
`-cyano group and the arabino configuration [23].
`
`Hence, compounds were prepared that incorporated all of these groups at the
`0
`2
`-position and also contained a 2-haloadenine moiety [24]. At that time, the synthetic
`0
`routes involved the displacement of a leaving group at C-2
`of a nucleoside with
`inversion from the ribo to the arabino configuration, as had been accomplished by
`Ueda [25]. However, the cytotoxicity results from this series of compounds, with the
`0
`exception of those incorporating a 2
`-fluorine, were not impressive [24], and thus
`0
`attention was re-focused on compounds with a 2
`-fluorine, in turn leading to the
`preparation of the series of compounds 7. The incorporation of a fluorine atom has
`always been attractive, because the size of the atom causes the least disruption, and is
`the closest to hydrogen. A fluorine also often significantly alters the biochemical
`
`6
`
`

`

`25.4 The Next Generation of Compoundsj637
`
`properties of a molecule. Initially, the synthetic route was quite laborious because the
`fluorocarbohydrate precursor required a lengthy synthetic route. However, starting
`0
`in the 1970s the group of Fox and Watanabe was also preparing nucleosides with a 2
`-
`fluorine [26–28], though with a major focus on pyrimidine nucleosides as antiviral
`agents, and we were able to take advantage of their synthetic procedures. The initial
`route (see below) [29, 30] yielded predominantly the desired b nucleoside, but
`separation of the anomers was necessary. The first synthesis in this target series
`was completed during the mid-1980s. By utilizing different bases for the coupling, all
`three compounds 7 were prepared. Based upon the pursuit of one of the pyrimidine
`nucleosides of antiviral interest, an improved synthesis of the fluorocarbohydrate
`precursor was developed and published [31], and this route proved to be a very useful
`0
`advance. Quite recently, another important advance relative to the a/b ratio at C-1
`has been made by utilizing a three-solvent mix for the coupling reaction, which is
`carried out with 2-chloroadenine rather than 2,6-dichloropurine [32]. This improve-
`ment was made during the optimization of the synthesis of clofarabine (7) for
`manufacturing purposes.
`
`7
`
`

`

`638j 25 Clofarabine: From Design to Approval
`
`Table 25.2 Anticancer activity of clofarabine against human tumor xenografts in mice.
`
`Tumor
`
`COLO 205 colon
`DLD-1 colon
`HCC-2998 colon
`HCT-15 colon
`HCT-116 colon
`HT29 colon
`KM20L2 colon
`SW-620 colon
`A498 renal
`CAKI-1 renal
`RXF 393 renal
`SN12C renel
`
`Activitya)
`þ
`A594 lung
`
`NCI-H23 lung
`
`NCI-H322M lung
`þ
`NCI-H460 lung
`þ
`DU-145 prostate
`þ
`LNCAP prostate
`þþ
`PC-3 prostate
`þþþ
`HL-60 leukemia
`þ
`CCRF-CEM leukemia
`þþ
`K-562 leukemia
`–
`MOLT-4 leukemia
`
`AS283 lymphoma
`þþ
`RL lymphoma
`a) – inactive;  marginal; þ minimal; þ þ good (tumor regressions); þ þ þ excellent (cures).
`
`Tumor
`
`Activitya)
`þþ
`þ
`þþþ
`þþ
`þ
`þþþ
`þþ
`þþ
`þþ
`þþþ
`–
`þþ
`
`0
`-fluorine in the arabino
`All three of the 2-haloadenine nucleosides with a 2
`configuration were significantly cytotoxic in multiple cancer cell lines [29, 30], in
`contrast to the other compounds mentioned above, and also in contrast to the
`0
`0
`0
`2
`-fluoro compound with the ribo configuration or the 2
`,2
`-difluoro analogue [33].
`This result encouraged us to move in the direction of human tumor xenograft
`experiments, and also to embark upon biochemical studies to determine the
`compound’s mechanism of action. The initial animal data strongly suggested that
`the 2-chloro compound had the most activity, and it was therefore chosen for further
`experiments. Although the 2-fluoro compound was also of some interest, it was more
`difficult to synthesize, and even a small amount of glycosidic cleavage yielding 2-
`fluoroadenine (see above) would be undesirable.
`Mechanistic information presenting the unique profile of clofarabine is summa-
`rized in the following section. With regard to further animal studies, many different
`human tumor xenograft models were employed, and much of the data obtained have
`been presented in one report [34]. A concise summary of clofarabine activity in
`human tumor models in mice is provided in Table 25.2. The conclusion to be drawn
`from these animal studies was that clofarabine had significant and often curative
`activity in a number of systems across a broad spectrum of human tumor types. This
`type of profile is exactly what is required for a new drug to have a strong chance of
`success in clinical trials. It is known that there is no direct correlation between
`outstanding activity in a particular human tumor type in a xenograft model and
`clinical success with that agent in treating the same type of tumor. In our experience,
`however, robust, broad-spectrum activity against a variety of human tumors in these
`models is a good indication that a compound will find clinical utility. Clofarabine
`clearly had such a profile, and we were eager for it to have the opportunity to move to a
`clinical trial.
`
`8
`
`

`

`25.5 Mechanism of Action of Clofarabinej639
`
`25.5
`Mechanism of Action of Clofarabine
`
`Generally speaking, all nucleoside analogues used in the treatment of cancer,
`including clofarabine, have a similar mechanism of action, namely that they are
`0
`converted to their respective 5
`-triphosphates and inhibit DNA synthesis. However, it
`is clear from the varying clinical activities of these agents that subtle quantitative and
`qualitative differences in the metabolism of these agents and their interactions with
`target enzymes can have a profound impact on their antitumor activity. Thus, a precise
`understanding of the mechanism of action of each of these compounds is important in
`order to determine those biochemical actions which are most important to antitumor
`activity and might aid the rational design and development of new agents.
`
`25.5.1
`Transport and Metabolism to Active Metabolites
`
`Clofarabine is efficiently transported into cells [35] by both the human equilibrative
`nucleoside transporters (hENT1, hENT2) and the human concentrative nucleoside
`transporters (hCNT2, hCNT3). As noted earlier, the primary enzyme involved in the
`activation of clofarabine in tumor cells is deoxycytidine kinase [36–38]. Clofarabine is
`a very good substrate for this enzyme, with Km and Vmax values similar to those of
`deoxycytidine. The structure of deoxycytidine kinase with clofarabine in the active site
`has recently been determined [39]. The results indicated that the conformation of the
`enzyme/clofarabine complex was similar to structures of the pyrimidine-bound
`complexes, and that interactions between the 2-Cl group and its surrounding
`hydrophobic residues contributed to the high catalytic efficiency of deoxycytidine
`kinase with clofarabine. Clofarabine is also a good substrate for deoxyguanosine
`kinase [40], an enzyme that is expressed in mitochondria. The contribution of
`deoxyguanosine kinase to the phosphorylation of clofarabine in cells is low due to
`the much higher expression of deoxycytidine kinase activity in most cell types [41],
`although this may be important in cells that express low activities of deoxycytidine
`kinase.
`Similar intracellular concentrations of clofarabine-5
`-monophosphate (the product
`0
`-triphosp-
`of the reaction of clofarabine with deoxycytidine kinase) and clofarabine-5
`hate (clofarabine-TP) accumulate in tumor cells treated with clofarabine [37]. This
`indicates that phosphorylation of the monophosphate of clofarabine is the rate-
`limiting step in its activation to the triphosphate, and that the monophosphate kinase
`does not easily tolerate substitutions at the 2-position of a purine nucleoside as large
`as a chlorine atom. Nucleoside kinases are usually the rate-limiting enzymes in the
`activation of anticancer nucleoside analogues. Even though clofarabine is a relatively
`poor substrate for the monophosphate kinase, clofarabine-TP accumulates to high
`levels in cancer cells treated with clofarabine. Clinical studies have indicated that the
`concentration of clofarabine-TP in blasts cells is 10-fold greater than the plasma
`concentration achieved after a 1-h infusion [42].
`
`0
`
`9
`
`

`

`640j 25 Clofarabine: From Design to Approval
`
`Phosphorylated metabolites of nucleoside analogues do not readily cross cell
`membranes, and are therefore trapped in the cell in which they were created; this
`contrasts with the nucleosides themselves, which will freely distribute across the
`cell membrane. Consequently, the antitumor activity of these agents can extend
`well beyond the time that the parent drug circulates in the plasma, because the
`active metabolite is maintained in tumor cells long after the drug has disappeared
`from the plasma. The initial half-life for the clearance of clofarabine-TP from
`cultured tumor cells (CEM) is approximately 2 h [33, 37]. In samples taken from
`patients treated with clofarabine, more than 50% of the clofarabine-TP was still
`present in circulating leukemic cells 24 h after the completion of transfusion [42].
`The long retention time of clofarabine-TP is believed to be a major contributory
`factor to the high activity demonstrated by clofarabine in solid tumor xenografts in
`mice [34].
`
`25.5.2
`Inhibition of DNA Synthesis
`
`As with other anticancer nucleoside analogues, the primary activity of clofarabine
`that is responsible for its antitumor activity is the inhibition of DNA synthesis [33, 36].
`RNA and protein synthesis are inhibited by clofarabine only at high concentrations.
`Clofarabine-TP is a potent inhibitor of ribonucleotide reductase [36, 43], a critical
`enzyme involved in the de-novo synthesis of deoxynucleotides [44]. The activity of
`ribonucleotide reductase in cells is tightly controlled by the natural deoxynucleoside
`triphosphates to ensure that the cell has all of the deoxynucleotides needed for DNA
`synthesis, and in the correct concentrations. dATP is a potent regulator of ribonu-
`cleotide reductase activity, and inhibits the reduction of ADP, UDP, and CDP [45].
`The effect of clofarabine on intracellular nucleotide pools suggest that clofarabine-TP
`interacts with ribonucleotide reductase in the allosteric binding site as an analogue of
`dATP. The inhibition of ribonucleotide reductase activity results in decreases in the
`levels of deoxynucleotide triphosphates, which are required for the synthesis of DNA.
`Clearly, decreasing the natural deoxynucleotide pools is sufficient to cause an
`inhibition of DNA synthesis; this is also the mechanism of action of hydroxyurea,
`another useful anticancer agent.
`Clofarabine is readily incorporated into DNA, although a small amount of drug is
`also detected in RNA [37]. At low concentrations, most of the clofarabine detected in
`DNA is incorporated into internal sites in the DNA of CEM cells, which indicates that
`the elongation of DNA synthesis is not prevented by the drug’s incorporation into
`DNA. The immediate inhibition of DNA synthesis in cells treated with clofarabine
`(even when most of the clofarabine is in internal positions in the DNA), however,
`indicates that inhibition of the DNA replication complex is the primary action of
`clofarabine that results in the death of the tumor cell. The disruption of DNA function
`due to the incorporation of clofarabine into daughter strands is of secondary
`importance to the activity of clofarabine.
`Clofarabine-TP is a good substrate and inhibitor of DNA polymerases a and e, two
`important enzymes involved in the replication of chromosomal DNA [36, 37].
`
`10
`
`

`

`25.5 Mechanism of Action of Clofarabinej641
`
`Clofarabine-TP is utilized by DNA polymerase a, with Km and Vmax values similar to
`those of dATP (the natural substrate). Once clofarabine-TP is incorporated into the
`growing chain, however, the ability of the polymerase to add new nucleotides is
`significantly less than after dATP incorporation. The inhibition of DNA polymerase a
`by clofarabine-TP is similar to that of fludarabine-TP [36]. The effect of clofarabine-TP
`on DNA polymerase e activity was similar to that seen with DNA polymerase a, except
`that clofarabine-TP more effectively inhibited chain elongation by DNA polymerase
`e [43]. This enhanced inhibition of chain elongation resulted in a significant
`inhibition of DNA synthesis at clofarabine-TP concentrations that were only 3% of
`the dATP levels used in the experiment.
`0
`The removal of clofarabine from the 3
`-end of DNA chains is an important part of
`the mechanism of action of nucleoside analogues that has not yet been evaluated. If
`0
`clofarabine can be quickly removed from the 3
`-terminus of the DNA chain, then
`DNA synthesis could continue normally; however, if it is not removed, then it will
`have a lasting effect on the ability of the DNA polymerase to elongate the DNA chain.
`The incorporation of two or more clofarabine residues sequentially in the DNA [36]
`may be harder to repair than single incorporations, and may represent a greater block
`to DNA synthesis. The continued inhibition of DNA synthesis is another attribute of
`nucleoside analogues that is a major contributor to their mechanisms of action. A
`short exposure of tumor cells to circulating drug may have lasting effects on the ability
`of the tumor cell to replicate its DNA.
`The results of the above-mentioned studies concluded that DNA synthesis is
`inhibited in cells treated with clofarabine by two distinct, but complementary, actions:
`(i) the inhibition of ribonucleotide reductase; and (ii) the inhibition of DNA
`polymerase activities. The potent inhibition of ribonucleotide reductase activity by
`clofarabine-TP enhances its inhibition of the replicative DNA polymerases by
`decreasing the intracellular concentration of the natural substrate, dATP, which in
`turn competes with clofarabine-TP for use as a substrate by these enzymes. With
`respect to the inhibition of ribonucleotide reductase and DNA polymerases, clofa-
`rabine combines the features of cladribine (potent inhibition of ribonucleotide
`reductase) and fludarabine (potent inhibition of DNA polymerase) into one molecule.
`
`25.5.3
`Induction of Apoptosis
`
`The inhibition of DNA synthesis by clofarabine is responsible for the induction of the
`apoptotic response in replicating cells. Inhibition of DNA replication in cells
`normally leads to the “turning on” of replication checkpoint pathways. Stalled
`replication forks can threaten DNA replication fidelity, and cells respond to replica-
`tion blocks by triggering checkpoint pathways that monitor replication fork progres-
`sion [46]. This monitoring of DNA synthesis operates normally during low-intensity
`replication stress, and is required for tumor cells to resume cell-cycle progression.
`However, during chronic or high-intensity replication stress, stalled forks do not
`restart after removal of the stressor, and this results in irreversible S-phase arrest,
`possibly mitotic catastrophe, and cell death.
`
`11
`
`

`

`642j 25 Clofarabine: From Design to Approval
`
`It has been found that replication stress induced by the treatment of CEM cells with
`clofarabine leads to Chk1 phosphorylation, Chk1 down-regulation, concomitant
`apoptosis, and cell death (unpublished data). Consistent with our findings on Chk1
`activation, it has been found that Cdc25A is down-regulated upon low-dose clofa-
`rabine treatment in CEM cells, a consequence of its phosphorylation by Chk1, which
`is accompanied by accumulation of cells in G1/S and S [47]. Chk1 and its upstream
`activator, ATR, phosphorylate a host of substrates involved in the control of the firing
`of additional replication origins. In addition, it is known that phosphorylation of
`Chk1 on Ser345 triggers its ubiquitylation, resulting in the proteasomal degradation
`of Chk1 [48].
`The induction of apoptosis in response to replication stress can also be mediated by
`the tumor suppressor/transcription factor p53. Recently, it was demonstrated that
`0
`two deoxycytidine analogues – gemcitabine and 4
`-thio-araC – can induce the
`stabilization of p73, a p53 paralogue in both p53 null cancer cells and in cells with
`wild-type p53 [49]. The p73 protein shares significant sequence similarity with p53,
`and can transactivate proapoptotic p53 target genes such as Bax and PUMA. In turn,
`these proteins can induce the release of cytochrome c and other apoptogenic
`molecules from the mitochondrial outer membrane. It has been found that the
`incubation of CEM cells (in which p53 is inactive) with clofarabine also robustly
`induced p73 levels (unpublished data).
`
`25.5.4
`Activity against Non-Proliferating Cancer Cells
`
`Clofarabine is also active against non-replicating cells [50], and has been shown to
`interfere with mitochondrial integrity and function in primary CLL cells [51], causing
`the release of cytochrome c and apoptotic inducing factor (AIF-1). This has led to the
`suggestion that clofarabine induces cell death by initiating the apoptotic cascade in
`these cells, although the precise mechanism for such activity has not yet been
`elucidated. Other adenine nucleosides such as cladribine and fludarabine have been
`shown to induce the intrinsic apoptotic pathway, which causes the release of
`apoptogenic molecules, such as cytochrome c into the cytosol; this leads to the
`cleavage and activation of caspase 9 (an initiator caspase) and subsequently caspase 3
`(an executioner caspase) [52, 53]. The present authors and others have shown that
`clofarabine induces the depleti