Antiherpesviral activity and inhibitory action on cell growth of 5-alkenyl derivatives of 1-beta D-arabinofuranosyluracil.

Antiherpesviral activity of 5-vinyl-1-beta-D-arabinofuranosyluracil was as high as that of 1-beta-D-arabinofuranosylthymine, whereas the former was less inhibitory to cell growth than the latter. 5-Propenyl- and 5-butenyl-1-beta-D-arabinofuranosyluracil were less active than 5-vinyl-1-beta-D-arabinofuranosyluracil.

Several thymidine analogs have been developed which inhibit selectively herpes simplex virus (HSV) replication without any significant action against cell growth (3-8, 11, 12). The selectivity is reported to be attributed to the fact that these compounds can be phosphorylated by the virus-induced deoxypyrimidine nucleoside kinase but not by cellular thymidine kinase (1,2,13). In a previous paper (11), it was reported that among 5-alkyl derivatives of 1-fl-D-arabinofuranosyluracil (ara-U), 5-ethyl-ara-U, as well as 1-ft-D-arabinofuranosylthymine (ara-T), displayed a wide margin of safety. We have also synthesized 5-alkenyl derivatives of ara-U from 5-hydroxymethyl-ara-U (Ikeda et al., submitted for ptublication). In this communication, their inhibitory activities against both HSV replication and cell growth are described.
Human embryonic lung fibroblast (HEL-F) cells, kindly supplied by T. Kuwata, Chiba University, were used in this study. HSV type 1 (HSV-1) strain VR-3 and HSV type 2 (HSV-2) strain MS, kindly supplied by S. Yamazaki, National Institute of Health of Japan, and two isolates of HSV from patients with herpes keratitis (HSV WT-20 and HSV WT-34) (9), kindly supplied by T. Kurimura, Tottori University School of Medicine, were employed for antiviral experiments. Methods for cultivation of the cells and propagation of the viruses were described previously (10,11). Antiherpesviral activity of the compounds was determined by a modified virus-rating method as described previously (11).
The antiviral activity was also expressed as the minimal discernible virus inhibitory concentration (MIC) of the compound at which HSVinduced cytopathogenic effect was depressed more thah 50%. The inhibitory action on cell growth was determined after 4 days of incubation of exponentially growing HEL-F cells with various concentrations of the compound, and was expressed in terms of the 50% cell growthinhibitory dose (IID50) obtained graphically.
The antiherpesviral activity of 5-vinyl-ara-U was almost the same as that of ara-T (Table 1). Thus, 5-vinyl-ara-U is considered to be more effective than 5-ethyl-ara-U, 5-iododeoxyuridine, and arabinosyladenine which has been shown to be less active than ara-T (11). 5-Vinylara-U was highly active against HSV-2 as well as against HSV-1. On the other hand, 5-propenyl-ara-U exhibited moderate activity and 5butenyl-ara-U showed relatively low activity against HSV-1. Their activities against HSV-2 were much lower than those against HSV-1. Thus, the antiherpesviral activity of 5-alkenyl derivatives of ara-U is concluded to decrease with increasing chain length in the substituent just like the activity of 5-alkyl derivatives of ara-U (11). Activity of 5-vinyl-ara-U against two isolates of HSV was compared with that of ara-T (Table 1). 5-Vinyl-ara-U was somewhat more active against HSV WT-34 than was ara-T, although it exhibited relatively low activity against HSV WT-20 which was resistant to 5iododeoxyuridine (9) as well as to ara-T (11).
Influence of 5-alkenyl derivatives of ara-U on the growth of HEL-F cells was also tested ( Table  1). The cell growth was inhibited by both ara-T and 5-vinyl-ara-U at high concentrations only. Although the ID50 of ara-T was reported to be 150 ug/ml in a previous paper (11), it was shown to be 140, 75, and 85 ,ug/ml (average, 100 ,ug/ml) in three separate experiments. The ID50 of 5vinyl-ara-U was about sixfold that of ara-T. 5-Propenyland 5-butenyl-ara-U scarcely inhibited the cell growth even at a concentration as high as 1,000 pg/ml. The antiviral indexes (the ID50 against HEL-F cells divided by the MIC against HSV-1) of ara-T, 5-vinyl-ara-U, 5-propenyl-ara-U, and 5butenyl-ara-U were 100, 620, more than 100, and cND, Not determined. more than 31, respectively. In terms of the margin of safety, therefore, 5-vinyl-ara-U seems to be superior to ara-T. Cheng (2) and Cheng et al. (3) reported that although 5-viniyl-deoxyuridine markedly inhibited replication of HSV-1 and HSV-2, it also inhibited the growth of WI-38 cells, a human diploid cell line derived from embryonic lungs, and of other cell lines (CV-1 and L1210 cells) at a concentration of 5 ,M. On the other hand, 5vinyl-ara-U did not inhibit at all the growth of HEL-F cells even at a concentration of 50 jig/ml (185 ,uM). Probably 5-vinyl-ara-U may be much more resistant to cellular thymidine kinase than the corresponding derivative of deoxyuridine. Altematively, 5-vinyl-ara-U 5'-triphosphate, if 5-vinyl-ara-U would be phosphorylated intracellularly, could be much less inhibitory to cellular deoxyribonucleic acid polymerase than 5-vinyldeoxyuridine 5'-triphosphate. Similarly, 5-propyl-ara-U (11) was also recognized to be less active than 5-propyl-deoxyuridine (6). Muller et al. (14,15) reported that ara-T was phosphorylated not only by HSV-infected cells but also by noninfected cells, and it was a potent inhibitor of the growth of mouse leukemia cells L5178Y. However, as presented here and reported by others (1,5,8,11,13), inhibitory action of ara-T on cell growth is rather low in general. In addition, 5-vinyl-ara-U was less inhibitory to growth of HEL-F cells than ara-T, which was highly effective against HSV-induced encephalitis and had an extremely low toxicity for mice (10). Satisfactory chemotherapy with 5-vinyl-ara-U against HSV infection may be expected.