The present study was undertaken to elucidate the mechanism of resistance to a relatively new nucleoside, 9-s-D-arabinofuranosylguanine (Ara-G), which is highly toxic to T-cell malignancies. In ord ...

The purine nucleoside 2-chlorodeoxyadenosine (CdA) is often used in leukemia therapy. Its efficacy, however, is compromised by the emergence of resistant cells. In the present study, 3 CdA-resistant cell lines were generated and characterized. Their ability to accumulate 2-chloroadenosine triphosphate (CdATP) varied, reflecting differences in activities of deoxycytidine kinase (dCK) and deoxyguanosine kinase (dGK). Nonetheless, the selected lines were uniformly resistant to CdA-induced apoptosis, as assessed by caspase activation and DNA fragmentation. In contrast, cytosols from resistant cells were capable of robust caspase activation when incubated in the presence of cytochrome c and dATP. Moreover, replacement of dATP with CdATP also resulted in caspase activation in the parental and some of the resistant cell lines. Strikingly, CdA-induced decreases in mitochondrial transmembrane potential and release of cytochrome c from mitochondria were observed in the parental cells but not in any resistant lines. The lack of cytochrome c release correlated with an increased ability of mitochondria from resistant cells to sequester free Ca2+. Consistent with this enhanced Ca2+buffering capacity, an early increase in cytosolic Ca2+after CdA treatment of parental cells but not resistant cells was detected. Furthermore, CdA-resistant cells were selectively cross-resistant to thapsigargin but not to staurosporine- or Fas-induced apoptosis. In addition, CdA-induced caspase-3 activation and DNA fragmentation were inhibited by the Ca2+ chelator BAPTA-AM in sensitive cells. Taken together, the data indicate that the mechanism of resistance to CdA may be dictated by changes in Ca2+-sensitive mitochondrial events.

We have previously reported that in a MOLT-4 leukemia cell line the acquired resistance to 9-β-d-arabinofuranosylguanine (Ara-G) is due to deficiency of the activating enzymes deoxyguanosine kinase and deoxycytidine kinase [Biochem. Biophys. Res. Commun. 293 (5) (2002) 1489]. In this study we investigated whether apoptotic pathways are affected in two human T-cell lymphoblastic MOLT-4 cell lines with acquired resistance to Ara-G. In contrast to the MOLT-4 wild type cells, Ara-G resistant cells displayed no increase in caspase-3 or caspase-9 activity, DNA fragmentation, cytochrome c release or a drop in the mitochondrial membrane potential (ΔΨmito) upon Ara-G treatment. A drop in the ΔΨmito was induced in wild type cells after treatment with tributyltin, an inducer of mitochondrial permeability transition, and with carbonyl cyanide m-chlorophenylhydrazone, an uncoupling agent that reduces the ΔΨmito, although not in Ara-G resistant cells. Ara-G resistant cells displayed higher levels of the anti-apoptotic protein Bcl-xL in immunoblots. A recent study indicates that Ara-G-induced apoptosis is mediated in part via the Fas pathway [Cancer Res. 43 (2047) (2002) 411]. When cells were treated with anti-Fas antibody, the wild type cell line exhibited increased caspase-3-like activity but the Ara-G resistant cells did not. Using FACS analysis and semi-quantitative PCR, 3–6-fold decreased protein levels and almost no detectable mRNA levels of Fas in the resistant cells were recorded. These data indicate that the inability to induce apoptosis via both the apoptosome pathway and the Fas pathway, due to increased levels of Bcl-xL and a lack of Fas, contributes to Ara-G resistance. This resistance to apoptosis in Ara-G resistant cells may serve to explain the overall resistance to a variety of anti-neoplastic drugs.

The aim of the thesis was to elucidate the mechanisms underlying resistance to nucleoside analogues used in the treatment of leukemias, with focus on cellular metabolism and induction of apoptosis. Cladribine (CdA), Clofarabine (CAFdA), Fludarabine (Fara-A) and Nelarabine (Ara-G) are nucleoside analogues with activity against various types of leukemias. CAFdA is a relatively new nucleoside analogue and we showed that CAFdA nucleotides were accumulated to a higher extent than CdA nucleotides in samples from patients with chronic lymphocytic leukemia (CLL) and acute myeloid leukemia (AML). CAFdA is more efficiently phosphorylated by deoxycytidine kinase (dCK) and CAFdA nucleotides are more slowly eliminated than CdA nucleotides. As earlier indicated in patients, there is an absence of cross-resistance between CdA and Fara-A. In an acute myeloid leukemia cell line the mechanism of resistance to CdA was a deficiency in dCK. Fara-A resistant cells had another contributing factor to resistance, the deoxynucleoside triphosphate pools being altered, indicating a mutation or altered regulation of ribonucleotide reductase (RR). Further studies in a lymphoid leukemia cell line supported these findings, and we also demonstrated that Fara-A resistant cells had increased RR activity and protein levels of the R2 subunit of RR. A real time quantitative PCR (RQ-PCR) method was established to measure mRNA levels of enzymes important in the metabolism of dCK, deoxyguanosine kinase (dGK) and high Km 5 ́-nucleotidase (5 ́-NT). The RQPCR method was compared to semi-quantitative PCR and enzyme activity measurements and tested with samples from pediatric patients with acute lymphocytic or myeloid leukemias, and was shown to be a convenient and reliable tool in the measurement of these enzymes. The major cause of resistance to CdA at the apoptotic level was a disturbed sensitivity to increased Ca levels in the cytosol. Increased Ca levels may induce changes in mitochondrial membrane potential (∆Ψmito). Accordingly, the increased Ca levels and the following drop in ∆Ψmito are important events for CdA-induced apoptosis. In another study we demonstrated that Ara-G-resistance was associated with perturbations in apoptotic events. Resistance to Ara-G correlated with upregulation of the anti-apoptotic protein Bcl-xL and downregulation of the Fas receptor. Thus, our data demonstrate that CAFdA is more effectively accumulated in samples from CLL and AML patients. Important for CdAand CAFdA-resistance is dCK and important for Fara-A-resistance in addition to dCK is RR. Apparent is also that abberations in apoptosis induced by nucleoside analogues may contribute to resistance. LIST OF PUBLICATIONS AND MANUSCRIPTS This thesis is based on the following papers: I. Lotfi, K., Månsson, E., Spasokoukotskaja, T., Pettersson, B., Liliemark, J., Peterson, C., Eriksson, S., and Albertioni, F. Biochemical pharmacology and resistance to 2-chloro-2'-arabino-fluoro-2'-deoxyadenosine, a novel analogue of cladribine in human leukemic cells, Clinical Cancer Research, Sep;5(9):2438-44, 1999. II. Månsson, E., Spasokoukotskaja, T., Sällström, J., Eriksson, S., and Albertioni, F. Molecular and biochemical mechanisms of fludarabine and cladribine resistance in a human promyelocytic cell line, Cancer Research, Dec 1;59(23):5956-63, 1999. III. Månsson, E., Liliemark, E., Söderhäll, S., Gustafsson, G., Eriksson, S., and Albertioni, F. Real-time quantitative PCR assays for deoxycytidine kinase, deoxyguanosine kinase and 5 ́-nucleotidase mRNA measurement in cell lines and in patients with leukemia, Leukemia, 16:386-392, 2002. IV. Chandra, J., Månsson, E., Gogvadze, V., Kaufmann, S. H., Albertioni, F., and Orrenius, S. Resistance of leukemic cells to 2-chlorodeoxyadenosine is due to a lack of calcium-dependent cytochrome c release, Blood, Jan 15;99(2):655-63, 2002. V. Månsson, E., Flordal, E., Liliemark, J., Spasokoukotskaja, T., Elford, H., Lagercrantz, S., Eriksson, S., and Albertioni, F. Downregulation of deoxycytidine kinase in human leukemic cell lines resistant to cladribine and clofarabine and increased ribonucleotide reductase activity contributes to fludarabine resistance, Biochemical Pharmacology, In Press, 2002. VI. Månsson, E., Stridh, H., and Albertioni, F. Resistance to mitochondrialand Fas-mediated apoptosis in human leukemic cells with acquired resistance to 9-β-D-arabinofuranosylguanosine, Biochemical and Biophysical Research Communications, Nov;298(3):338-44, 2002. The published papers are reprinted with permission from the copyright holders.

The relative levels of the deoxycytidine kinase (dCK), deoxyguanosine kinase (dGK), and the 5′-nucleotidase (5′-NT) are of importance for the effect of many nucleoside analogues used in the treatment of hematological malignancies. To elucidate dCK, dGK and 5′-NT gene expressions in cell lines and in samples from patients with leukemia, we have established a real-time quantitative PCR (RQ-PCR) method. From the available dCK, dGK and 5′-NT cDNA sequences we designed specific primers and fluorogenic probes for the respective genes. The mRNA of dCK, dGK and 5′-NT was also measured by semi-quantitative RT-PCR, the enzyme activities by a radioactive substrate-based technique and Western blot was used to measure the amount of dCK and dGK protein. A MOLT-4 wild-type and its 9-β- D -arabinofuranosylguanine (Ara-G)-resistant subline was used for the methods comparisons and the RQ-PCR assay was used in 35 samples from pediatric patients with ALL and AML. The results from RQ-PCR for the cell lines were in agreement with the semi-quantitative RT-PCR. The mRNA expression for dCK, dGK and 5′-NT (expressed as the ratio of the respective gene and the reference gene) in pediatric ALL and AML patients showed a large interindividual variability from 0.06 to 2.34, non-detectable to 0.06 and 0.04 to 0.30, respectively. These results show that the quantitative evaluation by RQ-PCR is a valuable tool in the determination of dCK, dGK and 5’-NT mRNA levels in cell lines and in clinical samples which were expressed at various levels. This rapid, convenient and specific method is suitable for further studies of these genes in clinical samples.

Mitochondria require deoxyribonucleoside triphosphates for the synthesis of their DNA and one of the enzymes responsible for the initial phosphorylation of purine deoxyribonucleoside is deoxyguanosine kinase (dGK; EC 2.7.1.113). Recent studies have suggested that dGK in addition to deoxycytidine kinase phosphorylates several anti-cancer agents, such as 9-β-d-arabinofuranosylguanine (Ara-G), cladribine (CdA), and fludarabine. There appear to coexist several mRNA fragments of dGK. In the present study we found 10 fragments, the longest fragment had 834bp, and represented the entire open reading frame of dGK (780bp). The nine additional fragments detected ranged from 807 to 269bp. All the fragments were found to contain the specific mitochondria translocation signal sequence. Expression of these fragments in Escherichia coli demonstrated that only the full-length dGK resulted in a protein that could phosphorylate CdA and Ara-G. Given the difficulty to measure the full-length dGK, these data are of value for studying the mRNA gene expression of dGK in cell lines and in leukemic cells from patients.

The pyrimidine analogue cytosine arabinoside (AraC) is one of the most effective drugs used in the treatment of acute leukaemia. Overexpression of the multidrug resistance (MDR-1) gene and its product, P-glycoprotein (P-gp), is associated with cellular resistance to drugs, such as anthracyclines and vinca alkaloids. This resistance can be reversed by cyclosporine analogues or verapamil (ver). We investigated the in vitro cross-resistance to AraC in a doxorubicin-resistant HL60 cell line, with an elevated expression of the MDR-1 gene. The resistant clone showed an eightfold increased resistance to AraC and a two- to fourfold resistance to the other analogues, as measured by cytotoxicity test. There was no significant increase in the activity of 5'-nucleotidase or in the amount of deoxyribonucleotide pools between cell lines. We could, however, detect a reduction in deoxycytidine kinase (dCK) activity (30%, P = 0.021, using deoxycytidine as substrate) and the level of AraC triphosphates was significantly reduced in the resistant cells (70%, P = 0.009). When the cells were exposed to cyclosporin A (CsA) or the cyclosporine analogue PSC 833 (PSC) in combination with AraC, there was more extensive apoptosis, as measured by formation of oligonucleosomal DNA fragmentation and caspase-3-like activity, than with exposure to AraC alone. We also found an increased retention of AraC in the resistant cells when incubated with AraC in combination with CsA. Ver in combination with AraC, failed to increase apoptosis for the resistant cell line. Our data suggests that the resistance to AraC for the P-gp-expressing cells is a result of a reduction of dCK activity and an increase in efflux, the latter possibly depending on P-gp. A combination of CsA or PSC with AraC may improve the effect of AraC in vivo.

Cross‐resistance between different classes of anti‐neoplastic agents can jeopardize successful combination cancer chemotherapy. In this study, we observed an unexpected cross‐resistance between the podophyllotoxine derivative etoposide (VP) and the nucleoside analogue cladribine (CdA) in CCRF‐CEM cells developed for resistance to VP. The resistant cells also displayed 14‐ and twofold resistance to cytarabine (ara‐C) and gemcitabine respectively. Closer analysis of these cells showed that they contained lower amounts of topoisomerase (topo) IIα ( P < 0·001) and β protein ( P < 0·026), formed substantially lower amounts of the topo II–DNA complex, and had a markedly decreased level of Fas (CD95/APO‐1)‐ligand mRNA expression. Interestingly, Fas expression in the resistant cells did not differ from that in the parental cell line. No differences were observed in the accumulation/efflux of daunorubicin or in the gene expressions of P‐glycoprotein, multidrug resistance‐associated protein and the lung resistance‐related protein. The activity of deoxycytidine kinase (dCK), responsible for activation of CdA and ara‐C, was the same for resistant and wild‐type cells. However, there was an increase in the activity of the cytosolic 5′‐nucleotidases (5′‐NT), responsible for deactivation of nucleotides, amounting to 206% ( P < 0·001) for the high K m and 134% ( P < 0·331) for the low K m 5′‐NT in resistant cells. The high K m 5′‐NT is probably responsible for the decreased amount of the active metabolite CdA 5′‐triphosphate [40% decreased ( P < 0·045)], as well as for other purine ribonucleosides and deoxyribonucleosides triphosphates in the resistant cells. In contrast, a significantly higher deoxycytidine triphosphate (dCTP) level (167%, P < 0·001) was observed in the resistant cells. Thus, this study suggests that the major cause of resistance to the nucleoside analogues CdA and ara‐C in cells selected for resistance to VP is a result of metabolic alterations producing increased activity of 5′‐NT and higher dCTP levels. Furthermore, these results indicate that there is a common factor in the regulation of nucleotide‐degrading enzymes and DNA topoisomerases, which may be altered in cross‐resistant cells.

The objective of the present study was to investigate the biochemical pharmacology of 2-chloro-2'-arabino-fluoro-2'-deoxyadenosine (CAFdA)--a fluorinated analogue of cladribine [2-chloro-2'-deoxyadenosine, Leustatin (CdA)] with improved acid and metabolic stability--in human leukemic cell lines and in mononuclear cells isolated from patients with chronic lymphocytic leukemia (CLL) and acute myelocytic leukemia (AML). We have also made and characterized two cell lines that are not sensitive to the growth inhibitory and cytotoxic effects of CAFdA. Incubation of cells isolated from the blood of CLL and AML patients with various concentrations of CdA or of CAFdA accumulated CdA and CAFdA nucleotides in a dose-dependent manner. A significantly higher rate of phosphorylation to monophosphates was observed for CAFdA than for CdA in cells from CLL patients (n = 14; P = 0.04). The differences in the phosphorylation were even more pronounced for the respective triphosphates in both CLL (n = 14; P = 0.001) and AML (n = 4; P = 0.04) cells. Retention of CAFdA 5'-triphosphate (CAFdATP) was also longer than that for CdA 5'-triphosphate (CdATP) in cells from leukemic patients. The relative efficacy of CAFdA as a substrate for purified recombinant deoxycytidine kinase (dCK), the key enzyme in the activation of nucleoside analogues, was very high and exceeded that of CdA as well as the natural substrate, deoxycytidine, by a factor of 2 and 8, respectively. The Km for CAFdA with dCK was also lower than that for CdA, as measured in crude extracts from the human acute lymphoblastic leukemia cell line CCRF-CEM and the promyelocytic leukemia cell line HL60. Acquired resistance to CAFdA in HL60 and in CCRF-CEM cell lines was directly correlated to the decreased activity of the nucleoside phosphorylating enzyme, dCK. Resistant cells also showed a considerable degree of cross-resistance to analogues that were activated by dCK. These observations demonstrated that dCK phosphorylates CAFdA more efficiently than CdA. Furthermore, CAFdATP is apparently more stable than CdATP and the mechanisms of resistance to CAFdA are similar to those leading to CdA resistance. These results encourage studies on the clinical effect of CAFdA in lymphoproliferative diseases.

2F-Adenine arabinoside (fludarabine, Fara-A) and 2-chloro-2'-deoxyadenosine (cladribine, CdA) are nucleoside analogues with antineoplastic activity in vitro and in vivo. Lack of clinical resistance between CdA and Fara-A has been demonstrated in patients with chronic lymphocytic leukemia (G. Juliusson et al., N. Engl. J. Med., 327: 1056-1061, 1992). To clarify the differences in mechanism of resistance to CdA and Fara-A in vitro, we developed two stable, resistant cell lines, HL60/CdA and HL60/ Fara-A, by exposure to increasing concentrations of analogues over a period of 8 months. Resistant cells tolerated >8,000 and 5-fold higher concentrations of CdA and Fara-A, respectively. The specific activity of the nucleoside phosphorylating enzyme (using deoxycytidine as substrate) in cell extracts from HL60/CdA and HL60/Fara-A mutants was about 10 and 60%, respectively, compared with the parental cell line. Western blot analysis using a polyclonal antibody showed no detectable deoxycytidine kinase (dCK) protein in CdA-resistant cells, whereas in Fara-A-resistant cells, it was at the same level as in the parental cells. The mitochondrial enzyme deoxyguanosine kinase was not altered in resistant cell lines. The HL60/CdA cells showed cross-resistance to 2-chloro-2'-arabino-fluoro-2'-deoxyadenosine, Fara-A, arabinofuranosyl cytosine, difluorodeoxyguanosine, and difluorodeoxycytidine toxicity, most likely because of the decreased phosphorylation of these analogues by dCK. Using real-time quantitative PCR, the mRNA levels of dCK and cytosolic 5'-nucleotidase (5'-NT), a major nucleoside dephosphorylating enzyme, were measured. It was shown that the dCK mRNA levels in both CdA- and Fara-A resistant cells were decreased in parallel with the activity. The expression of 5'-NT mRNA was not significantly elevated in CdA- and Fara-A resistant cells, as compared with the parental cells. Ribonucleotide reductase maintains a balanced supply of deoxynucleotide triphosphate pools in the cell and may also be a major cellular target for CdA and Fara-A nucleotides. Except for the deoxycytidine triphosphate level, the intracellular deoxynucleotide triphosphate pools were significantly higher in Fara-A-resistant cells compared with the parental cell line. This might be a consequence of mutation or altered regulation of ribonucleotide reductase activity and may explain the 2-5-fold cross-resistance to several nucleoside analogues observed with HL60/Fara-A cells. It is likely that the resistance for CdA was mainly attributable to a dCK deficiency, and Fara-A-resistant cells might have another contributing factor to the resistance beyond the dCK deficiency.

Cyclosporine A (CyA) treatment of 4 patients with severe aplastic anemia, who were ineligible for bone marrow transplantation, was carried out for periods of between 12 weeks to 20 months. A normalization of Hb and bone marrow, together with a marked improvement in WBC and platelet counts, were observed in only one of these four patients. The remission was maintained for 20 months under continuous treatment. A relapse occurred only when the patient himself interrupted treatment. No serious side effects were observed with CyA doses of 4–10 mg/kg/daily and blood concentrations of 200–400 ng/ml. No significant changes in T helper/T suppressor ratios were noted during the course of CyA treatment.