Methylation of HSV-1 DNA as a mechanism of viral inhibition: studies of an analogue of methyldeoxycytidine: trifluoromethyldeoxycytidine (F3mdCyd).

Although several hypomethylating agents such as 5-azadeoxycytidine and 5-fluorodeoxycytidine have been shown to activate transcription after incorporation into viral or cellular DNA, agents which selectively affect the methylation status of virus-infected cells have not been described. Studies on the antiviral effect of the methyldeoxycytidine (mdCyd) analogue trifluoromethyldeoxycytidine (F3mdCyd) showed significant antiviral activity against herpes simplex virus type 1 (HSV-1). This analogue of both dCyd and dThd is selectively incorporated into the DNA of herpesvirus infected cells due to the unique specificity of the herpesvirus thymidine kinase (TK) because the HSV-1 TK is both a dCyd and dThd kinase. In contrast, the deoxycytidine kinase of uninfected cells preferentially phosphorylates dCyd and has a poor affinity for F3mdCyd. F3mdCyd hemisubstituted M13 DNA displayed the same properties as mdCyd-substituted M13 DNA with respect to cleavage by restriction enzymes, and acted as an efficient template for eukaryotic DNA methyltransferase (S-adenosyl-L-methionine DNA (cytosine-5) methyltransferase: EC 2.1.1.37). Using the persistently infected CEM cell model system, the extent of DNA methylation was shown to increase in a dose-related manner when HSV-1-infected CEM cells were treated with increasing concentrations of F3mdCyd. Higher levels of methylation correlated with significant decreases in HSV-1 titers. Isoschizomer analyses followed by Southern blotting and hybridization with genomic HSV-1 DNA showed that DNA from HSV-1-infected, analogue-treated Vero cells was resistant to cleavage by restriction enzymes at a time when productive virus was not present in culture. We infer from these results that the methylation-like properties of the incorporated F3mdCyd occur concomitantly with, and appear to be involved in, the mechanisms of the analogue's antiviral effect towards HSV-1.

Radiation, pool size and incorporation studies in mice with 5-chloro-2'-deoxycytidine.

Bolus doses of 5-chlorodeoxycytidine (CldC) administered with modulators of pyrimidine metabolism, followed by X-irradiation, resulted in a 2-fold dose increase effect against RIF-1 tumors in C3H mice. Pool size studies of the fate of [14C]-CldC in BDF1 mice bearing Sarcoma-180 tumors, which demonstrated the rapid formation of 5-chlorodeoxycytidylate (CldCMP), and incorporation of CldC as such in RIF-1 tumor DNA, indicate that CldC is a substrate for deoxycytidine kinase, as our past Km studies have shown. Our data indicate that 5-chlorodeoxyuridine triphosphate (CldUTP) accumulates from both the cytidine deaminase-thymidine kinase pathway, as well as from the deoxycytidine kinase-dCMP deaminase pathway, in tumor tissue. As shown in a previous study, tetrahydrouridine (H4U), a potent inhibitor of cytidine deaminase, can effectively inhibit the enzyme in the normal tissues of BDF1 mice. When H4U was administered with the modulators N-(phosphonacetyl)-L-aspartic acid (PALA) and 5-fluorodeoxycytidine (FdC), the levels of CldC-derived RNA and DNA directed metabolites increased in tumor and decreased in normal tissues compared to when CldC was administered alone. These modulators inhibit the de novo pathway of thymidine biosynthesis, lowering thymidine triphosphate (TTP) levels, which compete with CldUTP for incorporation into DNA. 5-Benzylacyclouridine (BAU), an inhibitor of uridine phosphorylase, was also utilized. DNA incorporation studies using C3H mice bearing RIF-1 tumors showed that the extent of incorporation of 5-chlorodeoxyuridine (CldU) into DNA correlates with the levels of cytidine and dCMP deaminases; this is encouraging in view of their high activity in many human malignancies and the low activities in normal tissues, including those undergoing active replication. Up to 3.9% replacement of thymidine by CldU took place in RIF-1 tumors, whereas incorporation into bone marrow was below our limit of detection. CldC did not result in photosensitization under conditions in cell culture in which radiosensitization to X rays was obtained. Thus, the combination of CldC with modulators of its metabolism has potential as a modality of selective radiosensitization for ultimate clinical use in a wider range of tumors than those of the brain.

Use of 5-trifluoromethyldeoxycytidine and tetrahydrouridine to circumvent catabolism and exploit high levels of cytidine deaminase in tumors to achieve DNA- and target-directed therapies.

5-Trifluoromethyl-2'-deoxycytidine (F3methyl-dCyd), when coadministered with tetrahydrouridine (H4Urd), surpasses the efficacy of 5-trifluorothymidine and 5-trifluoromethyl-2'-deoxycytidine when administered alone as demonstrated with adenocarcinoma 755 and Lewis lung carcinoma as solid tumors implanted in C57BL X DBA/2 F1 mice. It appears that the reason for the heightened efficacy of F3methyl-dCyd, when coadministered with low concentrations of H4Urd, is decreased systemic deamination and subsequent systemic catabolism by pyrimidine nucleoside phosphorylases, which do not act on deoxycytidine and its analogues. Furthermore, the elevated levels of cytidine deaminase in these mouse tumors may result in selective conversion of F3methyl-dCyd to 5-trifluorothymidine at the tumor site. This suggests an approach to the treatment of human tumors possessing elevated levels of cytidine deaminase such as certain leukemias, bronchogenic carcinoma of the lung, adenocarcinomas of the colon and rectum, astrocytomas, and certain tumors which are refractory to chemotherapy with 1-beta-D-arabinofuranosylcytosine. In contrast to fluorinated pyrimidines in current use, F3methyl-dCyd + H4Urd potentially allows an exclusive DNA-, rather than both a DNA- and RNA-, directed approach. The major mechanism of the antitumor activity of F3methyl-dCyd appears to be via inhibition by 5-trifluorothymidine-5'-monophosphate of thymidylate synthetase, the target enzyme of fluoropyrimidine analogues in current use. However, the established and potential differences in the mode of action, anabolism, nature of incorporation into DNA, repair and cofactor requirements of F3methyl-dCyd and its anabolites, compared to that of the commonly utilized fluorinated pyrimidines, indicate that F3methyl-dCyd + H4Urd is a novel combination of agents. In comparative studies with Lewis lung carcinoma, F3methyl-dCyd (+ H4Urd) was shown to surpass the efficacies of 5-fluorouracil and 5-fluorodeoxyuridine and to be essentially equal in efficacy to 5-fluorodeoxycytidine (+ H4Urd). The optimum established protocol against Lewis lung carcinoma is F3methyl-dCyd, 175 mg/kg, + H4Urd, 25 mg/kg, once per day for 7 days. Studies utilizing high concentrations of H4Urd coadministered with F3methyl-dCyd indicate that the major pathway of tumor inhibition is via conversion of F3methyl-dCyd to 5-trifluorothymidine in view of the fact that tumor inhibition diminishes at doses of H4Urd which result in extensive (93%) inhibition of tumor cytidine deaminase.(ABSTRACT TRUNCATED AT 400 WORDS)

Capacity of deoxycytidine to selectively antagonize cytotoxicity of 5-halogenated analogs of deoxycytidine without loss of antiherpetic activity.

Enzyme kinetic studies from this laboratory (M. Dobersen and S. Greer, Biochemistry 17:920-928, 1978) suggested that deoxycytidine could antagonize the toxicity of 5-halogenated analogs of deoxycytidine without interfering with their antiviral activity. Antagonism by deoxycytidine of the toxicity of 5-chlorodeoxycytidine without impairing its anti-herpes simplex virus type 2 activity is demonstrated in the present studies. Tetrahydrouridine, an inhibitor of cytidine deaminase, was utilized. The high Km for deoxycytidine (0.6 mM) with respect to the herpes pyrimidine nucleoside kinase as compared with the low Km for 5-chlorodeoxycytidine (1.1 microM) accounts for the absence of antagonism of the antiviral activity. The high Km for 5-chlorodeoxycytidine (56 microM) as compared with the low Km of deoxycytidine (2 microM) with respect to mammalian deoxycytidine kinase accounts, in great part, for the antagonism of toxicity. In addition, antagonism of toxicity by deoxycytidine is the result of factors other than the kinetic parameters of nucleoside kinases, as indicated by its antagonism of the cytotoxicity of 5-chlorodeoxyuridine. This may be attributed to replenishment of low dCTP pools, diminished because of effector inhibition of ribonucleoside diphosphate reductase by Cl-dUTP. Resistance of the herpes-encoded enzymes to effector control may also play a role in the selective antagonism. Cell culture studies with high concentrations of tetrahydrouridine and 2'-deoxytetrahydrouridine suggest that competition by deoxycytidine for deaminases may not play a major role. The fact that deoxycytidine antagonizes the toxicity of chlorodeoxyuridine also argues against competition for the deaminases as a major reason for its effect. Limited studies with a topical herpes simplex virus type 2 infection system indicate heightened efficacy of 5-chlorodeoxycytidine (and tetrahydrouridine) when deoxycytidine is coadministered. The concepts of selective antagonism of a chemotherapeutic agent derived from these studies may be applied to other approaches that extent beyond viral chemotherapy.