Starting from readily available methyl 5‐methyloxazole‐4‐carboxylate (1) and 4‐methyl‐5‐oxazolylcar‐boxylic acid hydrazide (11) the title compounds were prepared. The reaction of compound 1 with hydrazine hydrate afforded the corresponding hydrazide 2. The reaction of compound 2 with formic acid yielded 1‐formyl‐2‐(5‐methyloxazole‐4‐carboxyl)hydrazine (3). Refluxing of the latter with phosphorus pentasulfide in xylene gave compound 5 in 62% yield. The reaction of compound 3 with phosphorus pentoxide afforded compound 4. Starting from hydrazide 11, compounds 13 and 14 were prepared similarly. Reaction of compound 2 with substituted isothiocyanate yielded compound 9 which was cyclized in basic medium to 4‐alkyl‐5‐(5‐methyl‐4‐oxazolyl)‐2,4‐dihydro‐3H‐1,2,4‐triazole‐3‐thione (10). The isomer 19 was prepared similarly. Methylation and subsequent oxidation of compound 19 gave compound 21. Reaction of the acid 7 with thiosemicarbazide in the presence of phosphorus oxychloride gave 2‐amino‐5‐(5‐methyl‐4‐oxazolyl)1,3,4‐thiadiazole (8). 2‐Amino‐5‐(4‐methyl‐5‐oxazolyl)‐1,3,4‐thiadiazole (17) was prepared from acyl chloride 15 by the usual method.

A group of 3'-O-nitro-2'-deoxyuridines, 3'-O-nitro-2'-deoxycytidines, and 5'-O-nitro-2'-deoxyuridines possessing a variety of substituents (H, Me, F, I) at the C-5 position were synthesized for evaluation as anticancer/antiviral agents that have the ability to concomitantly release cytotoxic nitric oxide (*NO). Although these compounds generally released a greater percent of *NO than the reference drug isosorbide dinitrate upon incubation in the presence of l-cysteine, or serum, their cytotoxicity (CC(50) = 10(-3) to 10(-6) M range) was comparable to 5-iodo-2'-deoxyuridine, but weaker than 5-fluoro-2'-deoxyuridine, against a variety of cancer cell lines. No differences in cytotoxicity against nontransfected (KBALB, 143B), and the corresponding transfected (KBALB-STK, 143B-LTK) cancer cell lines possessing the herpes simplex virus type 1 (HSV-1) thymidine kinase gene (TK(+)) were observed, indicating that expression of the viral TK enzyme did not provide a gene therapeutic effect. These nitrate esters were inactive antiviral agents except for 5-iodo-3'-O-nitro-2'-deoxyuridine that showed modest activity against HSV-1, HSV-2, and vaccinia virus.

Substrates for monitoring HSV1-tk gene expression include uracil and acycloguanosine derivatives. The most commonly used uracil derivative to monitor HSV1-tk gene transfer is 1-(2-fluoro-2-deoxy--D-arabinofuranosyl)-5-[*I]iodouracil (fialuridine; I*-FIAU), where the asterisk denotes any of the radioactive iodine isotopes that can be used. We have previously studied other nucleosides with imaging properties as good as or better than FIAU, including 1-(2-fluoro-2-deoxy--D-ribofuranosyl)-5-[*I]iodouracil (FIRU). The first aim of this study was to extend the biodistribution data of 123I-labelled FIRU. Secondly, we assessed the feasibility of detecting differences in HSV1-tk gene expression levels following adenoviral gene transfer in vivo with 123I-FIRU. 9L rat gliosarcoma cells were stably transfected with the HSV1-tk gene (9L-tk+). 123I-FIRU was prepared by radioiodination of 1-(2-fluoro-2-deoxy--D-ribofuranosyl)-5-tributylstannyl uracil (FTMRSU; precursor compound) and purified using an activated Sep-Pak column. Incubation of 9L-tk+ cells and the parental 9L cells with 123I-FIRU resulted in a 100-fold higher accumulation of radioactivity in the 9L-tk+ cells after an optimum incubation time of 4 h. NIH-bg-nu-xid mice were then inoculated subcutaneously with HSV1-tk (-) 9L cells or HSV1-tk (+) 9L-tk+ cells into both flanks. Biodistribution studies and gamma camera imaging were performed at 15 min and 1, 2, 4 and 24 h p.i. At 15 min, the tumour/muscle, tumour/blood and tumour/brain ratios were 5.2, 1.0 and 30.3 respectively. Rapid renal clearance of the tracer from the body resulted in increasing tumour/muscle, tumour/blood and tumour/brain ratios, reaching values of 32.2, 12.5 and 171.6 at 4 h p.i. A maximum specific activity of 22%ID/g tissue was reached in the 9L-tk+ tumours 4 h after 123I-FIRU injection. Two Ad5-based adenoviral vectors containing the HSV1-tk gene were constructed: a replication-incompetent vector with the transgene in the former E1 region, driven by a modified CMV promoter, and a novel replication-competent vector with the HSV1-tk gene in E3 driven by the natural E3 promoter. The human glioma cell lines U87MG and T98G were infected with a multiplicity of infection (m.o.i.) of 10. Forty-eight hours later the cells were incubated with 123I-FIRU and radioactivity was measured in a gamma counter. We found significantly higher levels of radioactivity in both cell lines following infection with the replication-competent vector (P<0.001). NIH-bg-nu-xid mice were then inoculated subcutaneously with U87MG cells. Tumours (approximately 1,000 mm3) were injected with 108 and 109 Infectious Units (I.U.) of either vector. After 48 h, the tracer was injected, followed by gamma camera imaging and direct measurement of radioactivity in the tumours at 4 h p.i. Images and direct measurements indicated increased uptake of tracer with higher I.U. and also demonstrated increased accumulation of tracer in the tumours treated with the replication-competent adenoviral vector (P=0.03). These results demonstrate that 123I...