Cleavage of intracellular hepatitis C RNA in the virus core protein coding region by deoxyribozymes.

Hepatitis C virus (HCV) infection represents an important global health problem. Current antiviral therapeutics for HCV have proven inadequate in stemming the disease process. A novel therapeutic strategy involves the use of deoxyribozymes, also known as DNA enzymes or DNAzymes. These catalytic DNA molecules, designed to target and cleave specific RNA sequences, have shown promise in in vitro experimental models for various diseases and may serve as an alternative or adjunct to current HCV drug therapy. We designed and tested several deoxyribozymes that can bind and cleave highly conserved RNA sequences encoding the HCV core protein in in vitro systems. One of these deoxyribozymes reduced the level of our HCV RNA target by 32% and 48% after 24 h of cell exposure when tested in human hepatoma and epithelial cell lines, respectively. As this deoxyribozyme showed significant cleavage activity against HCV core protein target RNA in human cells, it may have potential as a therapeutic candidate for clinical trial in HCV infected patients.

Human cytidine deaminase as an ex vivo drug selectable marker in gene-modified primary bone marrow stromal cells.

Naturally occurring drug resistance genes of human origin can be exploited for selection of genetically engineered cells co-expressing a desired therapeutic transgene. Their non-immunogenicity in clinical applications would be a major asset. Human cytidine deaminase (hCD) is a chemoresistance gene that inactivates cytotoxic cytosine nucleoside analogs, such as cytosine arabinoside (Ara-C). The aim of this study was to establish if the hCD gene can serve as an ex vivo dominant selectable marker in engineered bone marrow stromal cells (MSCs). A bicistronic retrovector comprising the hCD cDNA and the green fluorescent protein (GFP) reporter gene was generated and used for transduction of A549 cells and primary murine MSCs. Analysis of transduced cells demonstrated stable integration of proviral DNA, more than 1000-fold increase in CD enzyme activity, and drug resistance to cytosine nucleoside analogs. In a mixture of transduced and untransduced MSCs, the percentage of retrovector-expressing cells could be increased to virtual purity (>99.5%) through in vitro drug selection with 1 microM Ara-C. Increased selective pressure with 2.5 microM Ara-C allowed for enrichment of a mixed population of MSCs expressing approximately six-fold higher levels of GFP and of CD activity when compared with unmanipulated engineered MSCs. Moreover, engraftment and endothelial differentiation of these in vitro selected and enriched gene-modified marrow stromal cells was demonstrated by Matrigel assay in vivo. In conclusion, these findings outline the potential of human CD as an ex vivo selection and enrichment marker of genetically engineered MSCs for transgenic cell therapy applications.

Potential of 5-aza-2'-deoxycytidine (Decitabine) a potent inhibitor of DNA methylation for therapy of advanced non-small cell lung cancer.

Although new agents and drug combinations have increased the response rate in advanced non-small cell lung cancer (NSCLC), long-term survivors are rare. There is an urgent need to develop new chemotherapeutic approaches for disease. In a previous pilot phase I-II study on 5-aza-2'-deoxycytidine (5-AZA-CdR) in patients with stage IV NSCLC, we observed several interesting responses, including one patient that was still alive (68 months) at the time of publication of our results. In the present report, we want to point out the long-term follow up of this patient, who survived 81 months, and discuss the interesting mechanism of action of 5-AZA-CdR that may have been responsible for this interesting response. 5-AZA-CdR is a potent inhibitor of DNA methylation. Recent progress in this field has shown that aberrant methylation of the promoter region of tumor suppressor genes inhibits their expression. This epigenetic event can contribute to tumorigenesis. Since 5-AZA-CdR can reactivate these genes by blocking DNA methylation, it has the potential to reverse tumorigenesis. This novel mode of action makes it an interesting agent to investigate for the chemotherapy of malignant disease, including lung cancer.

Selection of drug-resistant transduced cells with cytosine nucleoside analogs using the human cytidine deaminase gene.

Hematopoietic toxicity produced by most anticancer drugs limits their potential for curative therapy. We have shown previously that the human cytidine deaminase (CD) gene can confer drug resistance in murine bone marrow cells (BMCs) to the nucleoside analog, cytosine arabinoside (ARA-C). In the present study, as the first objective we showed that the CD gene can also render drug resistance in BMCs to related analogs, 2',2'-difluorodeoxycytidine (dFdC) and 5-azadeoxycytidine (5-AZA-CdR). As a second objective, we investigated the potential of ex vivo selection with cytosine nucleoside analogs of CD-transduced BMC. The goal of this approach was to enrich the fraction of CD-transduced BMCs so as to increase the transgene expression and level of drug resistance before transplantation. This strategy may have the potential to circumvent the problem in clinical gene therapy of low level of gene transfer and adequate long-term gene expression. Using a bicistronic retroviral vector containing the CD and the green fluorescent protein (CDiGFP), we transduced murine L1210 leukemic cells. All three analogs, ARA-C, dFdC, and 5-AZA-CdR were demonstrated in vitro to enrich (>95%) the population of leukemic cells expressing the GFP transgene. However, with CD-transduced primary murine BMCs cultivated at high cell density we observed that in vitro selection with ARA-C was not possible due to release of CD into the culture medium at amounts that were sufficient to inactivate the analog. The CD-containing medium produced a chemoprotective effect on mock BMCs as shown by lack of significant growth inhibition in the presence of ARA-C. However, at low cell density in a cell mixture containing CD-transduced cells, the mock BMCs showed marked drug sensitivity to ARA-C as determined by clonogenic assay. Selection with ARA-C was shown to significantly increase the CD enzyme activity in transduced BMC. These results suggest that CD gene has the potential to be a good selectable marker and a possible tool for chemoprotection in cancer gene therapy.

Antineoplastic action of 5-aza-2'-deoxycytidine and histone deacetylase inhibitor and their effect on the expression of retinoic acid receptor beta and estrogen receptor alpha genes in breast carcinoma cells.

PURPOSE: During tumorigenesis several cancer-related genes can be silenced by aberrant methylation. In many cases these silenced genes can be reactivated by exposure to the DNA methylation inhibitor, 5-aza-2'-deoxycytidine (5-AZA-CdR). Histone acetylation also plays a role in the control of expression of some genes. The aim of this study was to determine the antineoplastic activities of 5-AZA-CdR and trichostatin A (TSA), either administered alone or in combination. in MDA-MB-231 breast carcinoma cells. The effects of these drugs (alone and in combination) on the expression of the tumor suppressor gene, retinoic acid receptor (RAR beta) and of the estrogen receptor alpha gene (ER alpha), whose expression is lost in the cell line used in the study, were also investigated. METHODS: MDA-MB-231 cells were treated with 5-AZA-CdR and TSA and the antitumor activity of these drugs was determined by clonogenic assay. Total RNA was extracted from the treated cells and RT-PCR was used to determine the effect of the treatment on the expression of RAR beta and ER alpha. Methylation-sensitive PCR analysis was used to confirm that lack of expression of both genes was due to hypermethylation of their promoter regions. A single nucleotide primer extension assay was also used to quantify the reduction in DNA methylation following drug treatment. RESULTS: Both 5-AZA-CdR and TSA alone showed significant antineoplastic activity. The combination of the two drugs was synergistic with respect to MDA-MB-231 cell kill. 5-AZA-CdR alone weakly activated the expression of both RAR beta and ER alpha. TSA alone only activated RAR beta, but not ER alpha. The combination of these agents appeared to produce a greater activation of both genes. CONCLUSIONS: The interesting interaction between 5-AZA-CdR and TSA in both cell kill and cancer-related gene reactivation provides a rationale for the use of inhibitors of DNA methylation and histone deacetylation in combination for the chemotherapy of breast cancer.

Potential of ribozymes against deoxycytidine kinase to confer drug resistance to cytosine nucleoside analogs.

Hematopoietic toxicity is the dose-limiting side effect produced in cancer chemotherapy with deoxycytidine nucleoside analogs. Deletion of the deoxycytidine kinase (dCK), results in a drug resistance phenotype to these analogs. An interesting gene therapy strategy to confer drug resistance to cytosine nucleoside analogs would be to specifically inactivate the dCK in normal hematopoietic stem cell. In this study, we designed hammerhead ribozymes that can specifically cut and downregulate the murine dCK mRNA. Three different ribozymes were identified and shown to cleave in vitro the dCK RNA. After introduction of ribozyme cDNA into murine L1210 leukemic cells by retroviral transfer, two of the ribozymes showed some capacity in reducing dCK activity. However, analysis of transduced L1210 clones showed that the significant reduction in the dCK mRNA was not sufficient to confer drug resistance to cytosine arabinoside. Nevertheless, these results provide a new avenue of modulating the dCK enzyme activity and with improved modifications may have the potential for use in gene therapy to confer drug resistance to deoxycytidine analogs.

Coexpression of rat glutathione S-transferase A3 and human cytidine deaminase by a bicistronic retroviral vector confers in vitro resistance to nitrogen mustards and cytosine arabinoside in murine fibroblasts.

The transfer of drug resistance genes into hematopoietic cells is an experimental approach to protect patients from drug-induced myelosuppression. Because anti-cancer drugs are often administered in combination to increase their clinical efficacy, vectors that express two drug resistance genes are being developed to broaden the spectrum of chemoprotection. We have constructed a bicistronic vector, MFG/GST-IRES-CD (MFG/GIC) coexpressing rat glutathione S-transferase (GST) A3 isoform (rGST Yc1) and human cytidine deaminase (CD). Murine NIH 3T3 fibroblast cells transduced with this vector were evaluated for their resistance to nitrogen mustards and cytosine nucleoside analogs. GIC-transduced polyclonal cell populations (GIC cells) demonstrated marked increases in selenium-independent glutathione peroxidase (peroxidase) and CD activities, as well as increased resistance to melphalan (2.3-fold), chlorambucil (3.4-fold), and cytosine arabinoside (Ara-C) (8.1-fold). After selection with Ara-C, the peroxidase and CD activities of GIC cells were augmented 2.6- and 2.9-fold, respectively, in comparison with unselected cells, and the resistance to melphalan, chlorambucil, and Ara-C was further increased to 3.7-, 5.9-, and 53-fold, respectively. Melphalan selection of GIC cells likewise augmented their peroxidase (2.3-fold) and CD (1.9-fold) activities. GIC cells proliferated in the simultaneous presence of melphalan and Ara-C at drug concentrations that completely inhibited the growth of untransduced cells. The growth rate of unselected GIC cells exposed to the drug combination averaged 18% that of drug-free cultures. The growth rate of GIC cells exposed to the drug combination increased to 30% of controls after Ara-C selection and to 50% after melphalan selection. Our results suggest that retroviral transfer of MFG/GIC may be useful for chemoprotection against the toxicities of nitrogen mustards and cytosine nucleoside analogs.

Quantitation of inhibition of DNA methylation of the retinoic acid receptor beta gene by 5-Aza-2'-deoxycytidine in tumor cells using a single-nucleotide primer extension assay.

The expression of several cancer-related genes has been reported to be silenced by DNA methylation of their promoter region. 5-Aza-2'-deoxycytidine (5-AZA-CdR), a potent and specific inhibitor of DNA methylation, can reactivate the in vitro expression of these genes. In future clinical trials in tumor therapy with 5-AZA-CdR a method to quantitate its inhibition of methylation of specific tumor suppressor genes would provide important data for the analysis of the therapeutic efficacy of this analogue. We have modified the methylation-sensitive single-nucleotide primer extension assay reported by Gonzalgo and Jones (Nucleic Acids Res. 25, 2529-2531, 1997). Genomic DNA was treated with bisulfite and a fragment of the promoter region of the human retinoic acid receptor beta (RARbeta) gene, a tumor suppressor gene, was amplified using seminested PCR. Using two different primers we quantitated the inhibition of methylation produced by 5-AZA-CdR at two specific CpG sites in the RARbeta promoter in a human colon and a breast carcinoma cell line. The results obtained with the modified assay show a precise and reproducible quantitation of inhibition of DNA methylation produced by 5-AZA-CdR in tumor cells.

DNA methylation and cancer.

The methylation of DNA is an epigenetic modification that can play an important role in the control of gene expression in mammalian cells. The enzyme involved in this process is DNA methyltransferase, which catalyzes the transfer of a methyl group from S-adenosyl-methionine to cytosine residues to form 5-methylcytosine, a modified base that is found mostly at CpG sites in the genome. The presence of methylated CpG islands in the promoter region of genes can suppress their expression. This process may be due to the presence of 5-methylcytosine that apparently interferes with the binding of transcription factors or other DNA-binding proteins to block transcription. In different types of tumors, aberrant or accidental methylation of CpG islands in the promoter region has been observed for many cancer-related genes resulting in the silencing of their expression. How this aberrant hypermethylation takes place is not known. The genes involved include tumor suppressor genes, genes that suppress metastasis and angiogenesis, and genes that repair DNA suggesting that epigenetics plays an important role in tumorigenesis. The potent and specific inhibitor of DNA methylation, 5-aza-2'-deoxycytidine (5-AZA-CdR) has been demonstrated to reactivate the expression most of these "malignancy" suppressor genes in human tumor cell lines. These genes may be interesting targets for chemotherapy with inhibitors of DNA methylation in patients with cancer and this may help clarify the importance of this epigenetic mechanism in tumorigenesis.

DNA methylation of retinoic acid receptor beta in breast cancer and possible therapeutic role of 5-aza-2'-deoxycytidine.

The retinoic acid receptor beta (RARbeta), a putative tumor suppressor gene, has been reported to be poorly expressed in breast cancer. In this report using the methylation-specific PCR reaction we observed DNA methylation in the promoter region of RARbeta in several primary breast tumors. DNA sequence analysis showed that the positions of 5-methylcytosine in the RARbeta promoter region was almost identical to that reported previously by our laboratory for human DLD-1 colon carcinoma cells (Anti-Cancer Drugs 1998; 9: 743). Several other cancer-related genes have been also reported to be silenced by DNA methylation, including the p16 tumor suppressor gene, E-cadherin, an invasion suppressor gene and the estrogen receptor gene in breast cancer cell lines. Since breast cancer cells have several potential target genes for the DNA methylation inhibitor, 5-aza-2'-deoxycytidine (5-Aza-CdR), we investigated the in vitro antineoplastic activity of this analog on the human breast cancer cell line MDA-MB-231. We report that 5-Aza-CdR is a potent growth inhibitor and a potent cytotoxic agent against the breast carcinoma cells. These results suggest that 5-Aza-CdR may be an interesting agent to investigate in patients with breast cancer resistant to conventional chemotherapy.

Gene amplification of human cytidine deaminase proviral cDNA and increased levels of its mRNA produces enhanced drug resistance to cytosine arabinoside in retroviral-transduced murine fibroblasts.

Hematopoietic toxicity is one of the major problems that limits the effectiveness of many antineoplastic drugs. One approach to overcome this problem is to confer chemoresistance to the hematopoietic cells by gene transfer of drug resistance genes. Human cytidine deaminase (CD) inactivates the cytosine nucleoside analogues, such as cytosine arabinoside (ARA-C), by deamination. We have reported previously that retroviral-mediated gene transfer of CD conferred drug resistance to ARA-C in murine cells. One of the major problems in the use of these vectors is to obtain adequate and prolonged expression of the transferred gene to produce a therapeutic effect in the transduced cells. The objective of this investigation was to determine if it is possible to increase the expression of CD proviral DNA in transduced murine fibroblast cells. We observed that by the use of continuous exposure to increasing concentrations of ARA-C it was possible to enhance drug resistance in the transduced cells. This drug resistance was found to be associated with increases in CD enzyme activity and CD proviral mRNA and by amplification of the proviral CD gene.

Demethylation by 5-aza-2'-deoxycytidine of specific 5-methylcytosine sites in the promoter region of the retinoic acid receptor beta gene in human colon carcinoma cells.

The role of retinoic acid receptor beta(RARbeta), a putative tumor suppressor gene, in the development of colon malignancy still remains to be clarified. We reported previously that the expression of RARbeta in DLD-1 human colon adenocarcinoma cells was silenced by DNA methylation at the level of the promoter region (Anti-Cancer Drugs 1997; 8: 56). In addition, we observed that RARbeta expression could be activated by the hypomethylating action of 5-aza-2'-deoxycytidine (5-Aza-CdR). In this report we have identified, by sequencing of bisulfite-modified DNA of DLD-1 colon tumor cells, the specific 5-methylcytosine positions in the region of -46 to +251 bp from the transcription start site of RARbeta2. We observed that 5-Aza-CdR treatment demethylated these specific sites. Based on this sequence data, specific primers for the methylation-specific PCR (MSP) assay were designed to discriminate methylated from unmethylated CpG sites in the promoter region of RARbeta. This assay confirmed the changes in the methylation status of the RARbeta gene in DLD-1 colon tumor cells before and after treatment with 5-Aza-CdR. The methylation status of the promoter region of the RARbeta gene was also examined in primary human colon adenocarcinomas using the MSP assay. Six of the 14 colon tumor samples showed signs of hypermethylation of this gene. The MSP assay for RARbeta may be a useful tool to clarify the role of DNA methylation for this gene in colon tumorigenesis.

Coexpression of cytidine deaminase and mutant dihydrofolate reductase by a bicistronic retroviral vector confers resistance to cytosine arabinoside and methotrexate.

The transfer of a drug resistance gene into hematopoietic cells is an approach being investigated to overcome the problem of myelosuppression produced by anticancer drugs. Chemotherapeutic agents are often given in combination in order to increase their effectiveness. Consequently, there is an advantage in designing vectors for gene transfer that are capable of expressing two drug resistance genes. We have constructed a bicistronic retroviral vector, MFG-DHFR-IRES/CD, which contains the mutated human dihydrofolate reductase (DHFR) cDNA with a phenylalanine-to-serine substitution at codon 31 (F31S) and the human cytidine deaminase (CD) cDNA. Murine fibroblast and hematopoietic cells were transduced with this vector and evaluated for their resistance to methotrexate (MTX) and cytosine arabinoside (ARA-C). The transduced fibroblast cells showed high levels of resistance to MTX and to ARA-C as determined by a clonogenic assay. Using enzymatic assays, we observed a coordinate increase in resistance to MTX and DHFR enzyme activity following an ARA-C selection. In addition, MTX selection produced an increase in CD enzyme activity and ARA-C resistance. Murine hematopoietic cells transduced with the bicistronic vector also showed drug resistance to both MTX and ARA-C. Interestingly, the double-gene construct conferred an equivalent level of drug resistance compared with single-gene vectors bearing only CD or DHFR genes in the hematopoietic cells. These results demonstrate the potential of the MFG-DHFR-IRES/CD vector to confer drug resistance to both MTX and ARA-C and may have future application in chemoprotection of normal hematopoietic cells in patients with cancer.

Drug resistance to 5-aza-2'-deoxycytidine, 2',2'-difluorodeoxycytidine, and cytosine arabinoside conferred by retroviral-mediated transfer of human cytidine deaminase cDNA into murine cells.

PURPOSE: The hematopoietic toxicity produced by the cytosine nucleoside analogs is a critical problem that limits their effectiveness in cancer therapy. One strategy to prevent this dose-limiting toxicity would be to insert a gene for drug resistance to these analogs into normal bone marrow cells. Cytidine (CR) deaminase can deaminate and thus inactivate 5-aza-2'-deoxycytidine (5-AZA-CdR), 2',2'-difluorodeoxycytidine (dFdC) and cytosine arabinoside (ARA-C). The aim of this study was to determine if gene transfer of CR deaminase into murine fibroblast cells confers drug resistance to these cytosine nucleoside analogs and if this resistance can be prevented by the CR deaminase inhibitor, 3,4,5,6-tetrahydrouridine (THU). METHODS: NIH 3T3 murine fibroblast cells were transduced with retroviral particles containing the human CR deaminase cDNA. Assays measuring CR deaminase activity as well as the inhibitory action of 5-AZA-CdR, dFdC and ARA-C on colony formation, were performed in the presence of different concentrations of THU. RESULTS: Retroviral-mediated transfer of the CR deaminase gene into 3T3 fibroblasts produced a considerable increase in CR deaminase activity. The transduced cells also showed significant drug resistance to 5-AZA-CdR, dFdC and ARA-C, as demonstrated by a clonogenic assay. This drug resistance phenotype and elevated CR deaminase activity were reversed by THU. CONCLUSIONS: These findings indicate that the CR deaminase gene can potentially be used in cancer gene therapy for protecting normal cells against the cytotoxic actions of different cytosine nucleoside analogs. In addition, the CR deaminase-transduced cells can be used as a model for screening different CR deaminase inhibitors in an intact cellular system.

Effect of 5-aza-2'-deoxycytidine and vitamin D3 analogs on growth and differentiation of human myeloid leukemic cells.

PURPOSE: The object of this study was to investigate the antineoplastic action of 5-aza-2'-deoxycytidine (5-AZA) in combination with vitamin D analogs on HL-60 and NB-4 myeloid leukemic cells. The vitamin D analogs chosen for this investigation were 1,25-(OH)2-23-yne-cholecalciferol (23-D) and 1,25-(OH)2-delta16-23-yne-cholecalciferol (16-23-D) since they have the potential to be used clinically owing to their minimal action on calcium metabolism. METHODS: HL-60 and NB-4 leukemic cells were incubated with different concentrations of 5-AZA and either 23-D or 16-23-D and their antineoplastic action determined by inhibition of DNA synthesis and growth, induction of differentiation and colony assay. RESULTS: 5-AZA in combination with either vitamin D analog produced a greater growth inhibition and induction of differentiation than either agent alone. For HL-60 leukemic cells the combination of 5-AZA with either analog produced a synergistic loss of clonogenicity. These effects on clonogenicity correlated with the effects of the combination on inhibition of growth and DNA synthesis. CONCLUSION: These results suggest that vitamin D analogs may enhance the antileukemic action of 5-AZA and that it may be interesting to test these agents in combination in patients with myeloid leukemia.

Retroviral transfer and long-term expression of human cytidine deaminase cDNA in hematopoietic cells following transplantation in mice.

The chemotherapeutic effectiveness of cytosine nucleoside analogues used in cancer therapy is limited by their dose-dependent myelosuppression. A way to overcome this problem would be to insert the drug-resistance gene, cytidine deaminase (CD), into normal hematopoietic cells. CD catalyzes the deamination and pharmacological inactivation of cytosine nucleoside analogues, such as cytosine arabinoside (Ara-C). The objective of this study was to determine if we could obtain long-term persistence and expression of proviral CD in hematopoietic cells following transplantation of CD-transduced bone marrow cells in mice. Murine hematopoietic cells were transduced with an MFG retroviral vector containing CD cDNA and transplanted into lethally irradiated mice. The recipient mice were administered three courses of 10-15 h i.v. infusions of Ara-C (75-110 mg/kg). Blood, marrow and spleen samples were obtained and analyzed for CD proviral DNA by PCR, CD activity by enzyme assay, and drug resistance to Ara-C by clonogenic assay. We detected the presence of the CD proviral DNA in most of the samples examined. Approximately 1 year after transplantation several mice showed increased expression of CD activity in these tissues and some mice displayed signs of Ara-C resistance. These data demonstrate that persistent in vivo expression of proviral CD can be achieved in transduced hematopoietic cells and indicate some potential of this gene for chemoprotection to improve the efficacy of cytosine nucleoside analogues in cancer therapy.

Pilot phase I-II study on 5-aza-2'-deoxycytidine (Decitabine) in patients with metastatic lung cancer.

5-Aza-2'-deoxycytidine (5-AZA-CdR, Decitabine) is a nucleoside analog and an active drug for the therapy of acute leukemia. The incorporation of 5-AZA-CdR into DNA blocks DNA methylation and can result in the activation of specific genes, such as tumor suppressor genes. This novel mechanism of action of 5-AZA-CdR stimulated our interest in its potential for cancer therapy in patients with lung cancer. Using a colony assay we observed that 5-AZA-CdR showed a potent antineoplastic effect against two human lung carcinoma cell lines. The objective of this preliminary phase I-II study was to evaluate the toxicity and clinical efficacy of 5-AZA-CdR in patients with stage IV non-small cell lung carcinoma. There were 15 patients that entered the clinical study. For nine assessable patients that received 5-AZA-CdR by a single 8 h i.v. infusion of 200-660 mg/m2 for one or more cycles, the median survival duration was 6.7 months, with three patients surviving more than 15 months. The steady-state plasma concentration of 5-AZA-CdR during the infusion was estimated in some patients and was in the same range that produced activation of a tumor suppressor gene in human lung tumor cell lines as reported by other investigators. The major side effect of 5-AZA-CdR was hematopoietic toxicity which required a 5-6 week recovery period before the next cycle of therapy. This study suggests that 5-AZA-CdR may have some clinical activity against metastatic lung carcinoma using this type of dose schedule.

Pharmacological approach for optimization of the dose schedule of 5-Aza-2'-deoxycytidine (Decitabine) for the therapy of leukemia.

5-Aza-2'-deoxycytidine (5-Aza-CdR; Decitabine) is an active antineoplastic agent in patients with leukemia. Since 5-Aza-CdR is an S phase specific agent and has a short plasma half-life, its antileukemic activity is dose schedule-dependent. Leukemia patients who are candidates for 5-Aza-CdR therapy following relapse after therapy with cytosine arabinoside are at greater risk for the problem of drug resistance since these cytosine nucleoside analogues are metabolized by the same enzymes. Due to its unique mechanism of action of demethylating DNA, 5-Aza-CdR has the potential to activate tumor (growth) suppressor and differentiation genes that have been accidentally silenced by DNA methylation in leukemic cells. All these factors should be taken into account in the design of the optimal dose schedule of this analogue. The optimal dose schedule of 5-Aza-CdR should be based on the kinetic parameters of deoxycytidine kinase, its pharmacokinetics, its effects on DNA methylation and the cell cycle parameters of the leukemic cells and the normal hematopoietic stem cells. Since granulocytopenia is the major toxic effect produced by 5-Aza-CdR, the use of hematopoietic growth factors to shorten the duration of leukopenia should be investigated. Another approach which we are investigating is to use the methods of gene therapy to insert the cytidine deaminase gene into normal hematopoietic progenitor cells so as to make them drug resistant to 5-Aza-CdR. The use of other agents that can induce the differentiation of leukemic cells in combination with 5-Aza-CdR may have the potential to increase the clinical effectiveness of this analogue for the therapy of leukemia.