Search

Nucleoside analogues having a boronic acid in place of the 3-hydroxyl group of deoxyribose have been synthesized. The synthesis of 3'-dihydroxyboryl-2',3'-dideoxyribose was based on asymmetric homologation of boronic esters with (dihalomethyl)lithium, beginning from a (silyloxymethyl)boronic ester. A change of chiral director is required before introduction of the second stereocenter, and the direct displacement of (S,S)-1,2-dicyclohexyl-1,2-ethanediol by (1S,2S,3R,5S)-pinanediol was used for this purpose. Coupling of the pinanediol ester of the 1-acetoxy-3-dioxyboryl-5-tert-butylsilyloxy deoxyribose analogue with silylated pyrimidine bases was accomplished with trimethylsilyl bromide. The boronic acid nucleoside analogues were not cytotoxic toward Hep G2 (human hepatocarcinoma) cells. Decomposition occurred over a period of several hours at 37 °C, pH 7.4, with liberation of free pyrimidine base.

The occasion of this 25th anniversary issue encouraged us to reminisce about the important history of the discovery of the dideoxynucleoside analogues for the treatment of HIV/AIDS and to chronicle our thoughts about a particular exciting and rewarding period of our scientific careers. Following the identification of the anti-HIV activity of zidovudine (AZT), we participated in the urgent quest to discover optimal treatments of HIV infection and AIDS. A number of previously synthesized nucleoside analogues were comparatively evaluated, and stavudine (D4T) emerged as a promising candidate for development. Following clinical evaluation, D4T became a mainstay of the initial antiretroviral combination therapy, prolonging and saving numerous lives. It has only recently been supplanted by better-tolerated treatments. This article forms part of a special issue of Antiviral Research marking the 25th anniversary of antiretroviral drug discovery and development, vol. 85, issue 1, 2010., (Copyright 2009 Elsevier B.V. All rights reserved.)

Six boronated tetrapeptides with the carboxy moiety of phenylalanine replaced by dihydroxyboron were synthesized, and their activities against human immunodeficiency virus 1 (HIV-1) protease subsequently investigated. The sequences of these peptides were derived from HIV-1 protease substrates, which included the C-terminal part of the scissile bond (Phe-Pro) within the gag-pol polyprotein. Enzymatic studies showed that these compounds were competitive inhibitors of HIV-1 protease with K(i) values ranging from 5 to 18 microM when experiments were performed at high enzyme concentrations (above 5 x 10(-8) M); however, at low protease concentrations inhibition was due in part to an increase of the association constants of the protease subunits. Ac-Thr-Leu-Asn-PheB inhibited HIV-1 protease with a K(i) of 5 microM, whereas the non-boronated parental compound was inactive at concentrations up to 400 microM, which indicates the significance of boronation in enzyme inhibition. The boronated tetrapeptides were inhibitory to an HIV-1 protease variant that is resistant to several HIV-1 protease inhibitors. Finally, fluorescence analysis showed that the interactions between the boronated peptide Ac-Thr-Leu-Asn-PheB and HIV-1 protease resulted in a rapid decrease of fluorescence emission at 360 nm, which suggests the formation of a compound/enzyme complex. Boronated peptides may provide useful reagents for studying protease biochemistry and yield valuable information toward the development of protease dimerization inhibitors.

We previously reported that the racemic mixture and both enantiomers of gossypol inhibit the replication of human immunodeficiency virus-type 1 (HIV-1) (Lin et al., Antimicrob Agents Chemother 33: 2149-2151, 1989). The present study evaluates the activities of a variety of analogs of gossypol as well as a few non-gossypol analogs. Compounds 2, 3, 10, and 13 were slightly more inhibitory than (-)-gossypol to the replication of HIV-1 in cell culture. Compounds 4 and 8 were cytotoxic to human peripheral blood monocyte (PBM) cells, and compounds 2 and 3 were cytotoxic to Vero cells but not PBM cells. The effects of the two enantiomers of gossypol on the cell volume and migration of H9 cells through the cell cycle were evaluated during 72 hr of incubation. The (-)-enantiomer of gossypol was more toxic to H9 cells than the (+)-enantiomer of gossypol as evidenced by cell destruction. Prior to cell destruction, there appeared to be no significant effect on cell cycle distribution with either enantiomer.

Various 3'-deoxy-3'-C-(hydroxymethyl)-, 3'-deoxy-3'-C-(fluoromethyl)-, 3'-deoxy-3'-C-(azidomethyl)-, and 3'-deoxy-3'-C-(aminomethyl)-substituted nucleosides (total 12 compounds) have been synthesized and evaluated against L1210, P388, S-180, and CCRF-CEM cells and HSV-1, HSV-2, and HIV-1 in culture. Only 3'-deoxy-3'-C-(hydroxymethyl)thymidine (36) was found to show significant anticancer activity against L1210, P388, S-180, and CCRF-CEM cells with ED50 values of 50, 5, 10, and 1 microM, respectively. None of these compounds demonstrated significant antiviral activity against HSV-1, HSV-2, or HIV-1. These compounds were also evaluated against thymidine kinases derived from HSV-I (strain KOS), HSV-2 (strain 333), and mammalian (K562) cells. The thymidine kinase (HSV-1 strain KOS) was inhibited significantly by both 3'-deoxy-3'-C-(hydroxymethyl)- and 3'-deoxy-3'-C-(fluoromethyl)thymidine.

3'-Azido-2',3'-dideoxy-5-iodouridine (AzIdUrd) and 3'-azido-2',3'-dideoxy-5-bromouridine (AzBdUrd), previously shown to be potent and selective inhibitors of human immunodeficiency virus replication in vitro were minimally toxic to the uninfected human lymphoid cell line H9 (IC50 = 197 and 590 microM, respectively). Both compounds strongly inhibited the incorporation of [3H]thymidine but not [3H]deoxyadenosine into DNA, and we observed no significant inhibition of [3H]uridine incorporation into RNA or [3H]amino acid incorporation into protein. Exposure of H9 cells to AzIdUrd or AzBdUrd (100 microM, 24 hr) and pulse-labeling with [3H]thymidine resulted in approximately 80% reduction in levels of tritiated dTMP, dTDP, and dTTP relative to control. [125I]AzIdUrd was phosphorylated rapidly in H9 cells with the monophosphate accounting for over 90% of total soluble radioactivity. A relatively low but stable level of AzIdUTP was maintained over a 12-hr period. [125I]AzIdUrd was phosphorylated by a cell free extract of H9 cells at a rate approximately three times that of thymidine and its phosphorylation was inhibited by excess thymidine. AzIdUrd was found to be a competitive inhibitor of cytosolic thymidine kinase with a Ki of 2.63 microM and AzIdUMP a weak competitive inhibitor of thymidylate kinase with a Ki of 55.3 microM. Both AzIdUTP and AzBdUTP were potent competitive inhibitors of HIV-1 reverse transcriptase (Ki = 0.028 and 0.043 microM, respectively) and relatively poor inhibitors of H9 cell DNA polymerase alpha (Ki = 42.0 and 42.7 microM, respectively). Thus, the high therapeutic index of these compounds is due to the sensitivity of the viral reverse transcriptase, coupled with the relative insensitivity of the host cell DNA polymerase alpha.

The poly ADP-ribosylation of proteins catalyzed by poly(ADP-ribose)polymerase (PARP) is involved in a number of important cellular metabolic activities. We evaluated various analogs of deoxythymidine and deoxyuridine as inhibitors of PARP. Most of these compounds have antiviral and/or anticancer activities. The structural requirements for these nucleoside analogs to be inhibitors of PARP were determined. The compounds evaluated had various substitutions on the 2-, 4- and/or 5-position of the pyrimidine ring, as well as on the 2'-, 3'- and/or 5'-position of the pentose moiety. Inhibition of PARP was strongly dependent on the size of the alkyl or halogen substituent on the 5-position of the pyrimidine ring. Whereas the 5-position of the pyrimidine ring could be varied, alteration of the 2- or 4-position drastically decreased the inhibition of PARP. Kinetic analysis was performed with concentrations of 1-10 microM NAD+. The Ki values for many compounds were five to seven times lower than the Ki for 3-aminobenzamide, a previously described potent inhibitor of PARP. Compounds with combined substituents at both the 5-position of the pyrimidine ring and the 3'- or 5'-position of deoxyribose generally were potent inhibitors of PARP, as for example 3'-amino-2', 3'-dideoxy-(E)-5-(2-bromovinyl)uridine (Ki = 0.7 microM), or 5'-azido-2',5'-dideoxy-5-ethyluridine (Ki = 0.8 microM). The 5-halogenated analogs had Ki values of 18, 35, 110 and greater than 1000 microM for 5-iodo-2'-deoxyuridine, 5-bromo-2'-deoxyuridine, 5-chloro-2'-deoxyuridine, and 5-fluoro-2'-deoxyuridine, respectively, and the 5-alkyl analogs had Ki values of 45, 2.2, 7, 16 and 180 microM for 5-methyl-2'-deoxyuridine, 5-ethyl-2'-deoxyuridine, 5-propyl-2'-deoxyuridine, 5-butyl-2'-deoxyuridine and 5-pentyl-2'-deoxyuridine, respectively. Two other compounds with substituents in the 5-position of the pyrimidine moiety also had potent activities: (E)-5-(2-bromovinyl)-2'-deoxyuridine (Ki = 6 microM) and 5-trifluoromethyl-2'-deoxyuridine (Ki = 1.6 microM). Compounds substituted in the 2'-, 3'- and/or 5'-position of the deoxyribose moiety were investigated and 5'-azido-5'-deoxythymidine, 5'-amino-5'-deoxythymidine, 3'-azido-3'-deoxythymidine and 3'-deoxythymidine (d2T) and Ki values of 12, 16, 18 and 30 microM, respectively.

Various 2'- and 3'-methylidene-substituted nucleoside analogues have been synthesized and evaluated as potential anticancer and/or antiviral agents. Among these compounds, 2'-deoxy-2'-methylidene-5-fluorocytidine (22) and 2'-deoxy-2'-methylidenecytidine (23) not only demonstrated potent anticancer activity in culture against murine L1210 and P388 leukemias, Sarcoma 180, and human CCRF-CEM lymphoblastic leukemia, producing ED50 values of 1.2 and 0.3 microM, 0.6 and 0.4 microM, 1.5 and 1.5 microM, and 0.05 and 0.03 microM, respectively, but also were active in mice against murine L1210 leukemia. Of all the tested drug dosage levels (25, 50, and 75 mg/kg, respectively) compound 23 had no toxic deaths and compound 22 yielded only one toxic death at the highest dosage level. On the contrary, in the same study, 1-beta-D-arabinofuranosylcytosine (ara-C) resulted in 2/5, 5/5, and 5/5 toxic deaths, respectively. Both compounds 22 and 23 have shown better anticancer activity than ara-C, yielding higher T/C x 100 values and some long-term survivors (greater than 60 days). In addition, compounds 22 and 23 were found to have, respectively, approximately 130 and 40 times lower binding affinity for cytidine/deoxycytidine deaminase derived from human KB cells compared to ara-C, suggesting that the two 2'-methylidene-substituted analogues may be more resistant to deamination. Cytoplasmic deoxycytidine kinase (dCK) was required for compounds 22 and 23 action. Furthermore, compounds 14, 22, 23, and 24 also have antiherpes simplex virus type 1 (HSV-1) and type 2 (HSV-2) activity in cell culture. In addition, the crystal structure of 2'-deoxy-2'-methylidenecytidine hydrochloride (23-HCl) was determined by X-ray crystallography.

3'[3-(2-Chloroethyl)-3'nitrosoureido]-3'-deoxythymidine (3'-CTNU), a chloroethylnitrosourea analog of thymidine, is a potent antineoplastic agent against murine leukemia L1210. In this study, we have examined the effects of 3'-CTNU on cellular oncogene (proto-oncogene) expression. We found that the expression of the c-myb proto-oncogene was dramatically enhanced in a concentration- and time-dependent manner by 3'-CTNU in murine leukemia L1210 cells, whereas the expression of the c-myc proto-oncogene was suppressed. The enhancement of c-myb gene expression was found to be cell type-specific and to involve an increase of the c-myb transcription rate rather than an alteration of c-myb gene structure or increased stability of c-myb mRNA. Further analysis demonstrated that the altered c-myb gene expression was largely due to the presence of 3'-amino-3'-deoxythymidine, a decomposition product of 3'-CNTU. The expression of five other proto-oncogenes was unaffected by 3'-CTNU treatment. Our study showed that an antineoplastic agent can increase or decrease the expression of proto-oncogenes.

The coadministration of thymidine (dThd) with either 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU) or 3'-[3-(2-chloroethyl)-3-nitrosoureido]-3'-deoxythymidine (3'-CTNU) to L1210-bearing mice significantly enhanced the antitumor activity of both nitrosoureas (T-S. Lin and W. H. Prusoff, Cancer Res., 47:394-397, 1987, and T-S. Lin, P.H. Fischer, J. C. Marsh, and W. H. Prusoff, Cancer Res., 42:1624-1629, 1982). As a possible mechanism for this observed enhancement, we have investigated the role of dThd as an inhibitor of poly(ADP-ribose) polymerase (ADPRP), an enzyme which is activated in response to DNA damage. Exposure of L1210 cells in culture to 50 microM BCNU resulted in a greater than 10-fold increase in ADPRP activity within 3-4 h. The polymerase activity increased with increasing BCNU concentration after a 4-h exposure, reaching apparent saturation at 50 microM BCNU. However, this activation was abolished by 2 mM dThd. Median inhibition of the ADPRP activity elicited by 30 and 75 microM BCNU occurred at 38 and 135 microM dThd, respectively. When BCNU was replaced by 3'-CTNU, no activation of ADPRP was observed, even at or above concentration of 3'-CTNU previously shown to cause DNA damage. 3'-Amino-3'-deoxythymidine, the principal hydrolysis product of 3'-CTNU, was found to be an inhibitor of BCNU-stimulated ADPRP activity with potency similar to dThd. Furthermore, intact 3'-CTNU was found to inhibit BCNU-stimulated ADPRP activity. Although 3'-CTNU should be capable of activating ADPRP by causing DNA damage, our results suggest that no net activation is observed due to inhibition by the various thymidine species present. Thus, inhibition of ADPRP by dThd following DNA damage by BCNU is consistent with the potentiation of antitumor activity previously reported. However, the observed potentiation of 3'-CTNU activity by dThd does not appear to result from such a mechanism.

Various 3'-deoxy pyrimidine nucleoside analogues have been synthesized for evaluation as potential anticancer and antiviral agents. Among these compounds, 1-(3-deoxy-beta-D-threo-pentofuranosyl)cytosine (10, 3'-deoxy-ara-C) and 3'-deoxycytidine (22) had significant anticancer activity against CCRF-CEM, L1210, P388, and S-180 cancer cell lines in vitro, producing ED50 values of 2, 10, 5, and 34 microM, respectively, for 3'-deoxy-ara-C (10); and 25, 5, 2.5, and 15 microM, respectively, for 3'-deoxycytidine (22). Thus, 3'-deoxy-ara-C (10) was 12.5 times more active against CCRF-CEM cells than 3'-deoxycytidine (22). The 2'-O-acetyl, 5'-O-acetyl, and 2',5'-di-O-acetyl derivatives of 3'-deoxy-ara-C (10), compounds 34, 31, and 30, demonstrated anticancer activity in the same range as 3'-deoxy-ara-C (10) against CCRF-CEM, L1210, P388, and S-180 cells. The 5'-O-acetyl derivative (31) had significantly greater activity against CCRF-CEM with an ED50 value of 0.4, but this compound also showed similar activity, as did 3'-deoxy-ara-C, against L1210, P388, and S-180 with ED50 values of 3, 3, and 13 microM, respectively. 3'-Deoxy-ara-C was also evaluated in vitro against HSV-2, HCMV, and GPCMV viruses and was found to be not very active with respective IC50 values of 110, 220, and 1000 microM. The single-crystal structure of 3'-deoxy-ara-C (10) was determined by X-ray crystallography. There are two molecules of the nucleoside and one molecule of water in the asymmetric unit. The sugar moieties of the two nucleoside molecules adopt different conformations. In molecule A, the ring pucker is C3'-endo with P = 18.7 degrees and tau m = 37.3 degrees, while the CH2OH side chain is gauche+. In molecule B, the ring pucker is C2'-endo with P = 156.8 degrees and tau m = 37.8 degrees and the side chain is trans.

Herpes simplex virus type-1 (HSV-1) was grown in the presence of 5-iodo-2'-deoxyuridine (IdUrd), and the virion-DNA was isolated by isopycnic centrifugation in CsCl. Irradiation of IdUrd-containing HSV DNA with either 302 nm or 254 nm ultraviolet (UV) light introduced strand breakage into the DNA in a dose-dependent manner when analyzed by alkaline sucrose density gradient sedimentation. Irradiation of unsubstituted HSV DNA under similar conditions produced little strand breakage. These observations are in agreement with the proposed mechanism for photochemical generation of strand breakage in 5-halo-2'-deoxyuridine-containing DNA. Irradiation of IdUrd-substituted virions followed by analysis of the isolated DNA indicated less strand breakage than irradiation of isolated IdUrd-substituted DNA under equivalent conditions. The dosage of irradiation required to introduce DNA strand breakage in IdUrd-substituted virions was equivalent to that employed to affect greater than 99% loss of infectious virus activity in both control and IdUrd-containing virions. It is suggested that the relative UV insensitivity of IdUrd-substituted HSV may be due to the microstructure environment of the substituted HSV DNA which may favor recombination of the photochemically formed halogen-uracil radical pairs.

3'-Deoxythymidin-2'-ene (d4T) is a potent and selective inhibitor of human immunodeficiency virus replication in a variety of human cell types and is currently undergoing phase I clinical trials for the treatment of acquired immunodeficiency syndrome. As part of our ongoing studies of the cellular pharmacology of d4T, and in light of recent reports in which such nucleoside analogs as 3'-azido-3'-deoxythymidine (AZT) and 2',3'-dideoxyadenosine were shown to permeate cells by the unusual mechanism of nonfacilitated diffusion, we have investigated the uptake of d4T in the human lymphocyte cell line H9. Several lines of evidence suggest that d4T permeation of H9 cells occurs by nonfacilitated diffusion; 1) [3H]d4T influx was linear for the first 10 sec and was nonconcentrative, reaching equilibrium with the extracellular drug concentration in 2-3 min, 2) the initial rates of influx were a linear function of concentration over the range from 1 microM to 5 mM, with no sign of uptake by a saturable mechanism, and 3) the uptake of [3H]d4T was insensitive to the nucleoside transport inhibitors nitrobenzylthioinosine and dipyridamole, as well as a large molar excess of AZT, thymidine, or adenosine. The octanol/water partition coefficient of d4T was 0.179, intermediate between those of thymidine and AZT. Thus, d4T does not appear to be a substrate for the nucleoside transport system responsible for the uptake of physiological nucleosides as well as most nucleoside analogs, and it enters the cell by nonfacilitated diffusion.

The effect of 3'-deoxythymidin-2'-ene (d4T) on the metabolism of exogenously supplied radiolabeled nucleosides was investigated. Following a 24-hr exposure to 250 microM d4T, we observed no significant effect on the incorporation of [3H]thymidine or [3H]deoxycytidine into DNA. In contrast, the amounts of [3H]uridine, [3H]deoxyuridine, and [3H]cytidine were significantly lower than those incorporated by control cultures. d4T had no significant effect on the incorporation of [3H]uridine or [3H]cytidine into RNA, or the incorporation of 3H-labeled amino acids into protein. In d4T-treated cells the relative proportions of [3H]dTMP, [3H]dTDP, and [3H]dTTP formed did not change but their absolute concentrations were increased. d4T significantly reduced the level of [3H]dUMP, and a parallel decrease in [3H]dTMP derived from [3H]dUMP was also evident. d4T increased the amounts of labeled deoxycytidine metabolites formed, with increased dCMP levels the most prominent. In a cell-free extract, [3H]d4T was phosphorylated at a rate of 1.6 pmol/30 min. Increasing concentrations of both thymidine and deoxyuridine inhibited the phosphorylation of [3H]d4T with IC50 values of 5.7 and 35 microM respectively. d4T was found to be a weak substrate for purified H9 cytosolic thymidine kinase (Km = 138 microM) and a weak competitive inhibitor of thymidine and deoxyuridine phosphorylation by this enzyme (Ki = 1.37 and 0.33 mM respectively).

The effect of 5-iodo-2'-deoxyuridine monophosphate (IdUMP), various 5-halogenated-5'-azido-2', 5' -dideoxyuridine derivatives, 2'-deoxy-6-azauridine (AzdUrd), and its halogenated analogs on the ultraviolet sensitization of Escherichia coli thymidylate kinase has been investigated. Only those compounds iodinated in position 5 enhance the rate of ultraviolet inactivation of this enzyme. However, 5'-azido nucleosides with iodo, bromo, chloro, or fluoro substituents in position 5 neither protect nor sensitize thymidylate kinase to ultraviolet inactivation. Thymidine 5'-monophosphate partially protects the enzyme against ultraviolet inactivation either in the presence or absence of ultraviolet-sensitizing iodinated analogs. Magnesium ion does not enhance the ultraviolet inactivation of thymidylate kinase by 5-iodinated nucleoside analogs. The kinatic data support an active site-directed enhancement of the enzyme to ultraviolet inactivation by 5-iodo-2'-deoxyuridine monophosphate, since the concentration of IdUMP required to attain 50% maximal enhancement is 0.24 mM which is in good agreement with its Ki of 0.18 mM. When either [125I]IdUMP or [2-14C]IdUMP was irradiated with the enzyme, both radioactivities were associated with the enzyme, however only with the 14C analog was the amount bound at half-saturation essentially equal to the amount required to inactivate the enzyme by 50%. These data support the hypothesis that the active entity in the enhancement by IdUMP of thymidylate kinase inactivation during ultraviolet irradiation is the uridylate free radical which is formed photochemically from IdUMP. Photochemical studies of 6-azauracil (AzUra), 2'-deoxy-6-azauridine, and 5-iodo-2'-deoxy-6-azauridine (IAzdUrd) were performed. Photolysis of IAzdUrd in the presence of a hydrogen donor yields AzdUrd which upon further photolysis yields the photohydrate. The photohydrate of AzdUrd when incubated in the dark at pH 5.2 is 90% converted back to AzdUrd, whereas the photohydrate of AzUra is only partially (20%) converted to AzUra. The rate of deiodination of IAzdUrd is 2.1-fold greater than that of IdUMP. Although the Ki of IdUMP and IAzdUrd is similar, the increased photosensitivity of the aza analog accounts for the much greater enhancement of ultraviolet inactivation of thymidylate kinase. The ability of a compound to enhance the ultraviolet inactivation of deoxythymidylate kinase is correlated with the potential of the compound to produce a free radical rather than a photohydrate when the enzyme-substrate analog complex is irradiated.

The (-) enantiomer of gossypol but not the (+) enantiomer had good antiviral activity in peripheral blood mononuclear cells against human immunodeficiency virus type 1 at a concentration more than 20-fold lower than that required for cytotoxicity; however, in H9 cells the (-) enantiomer, although more potent as an antiviral agent, was more cytotoxic.

The efficiency of 5-iodo-5'-amino-2'5'-dideoxyuridine (AIU) in the therapy of experimental herpes keratitis in rabbits has been examined. Virus infections were established bilaterally in 40 animals using herpes simplex, type 1 (NIH strain 11124). Twenty-four hours after infection the rabbits were divided into five matched groups of eight and each group was treated, double-blind, with topical drugs at four-hour intervals for a total of 72 hours. The solutions instilled were: (1) saline; (2) IdUrd, 1 mg. per milliliter; (3) AIU, 1 mg. per milliliter; (4) AIU, 4 mg. per milliter; and (5) AIU, 8 mg. per milliliter. Each eye was examined daily for 12 days and graded independently by two ophthalmologists. Although IdUrd and AIU (8 mg. per milliliter) were effective therapeutically, IdUrd had a greater effect. The AIU at 1 and 4 mg. per milliliter were less active, but showed more rapid healing than the saline control. Viral recovery studies are consistent with the clinical observations. A second independent experiment, similar to that described above, gave essentially identical results. Although less potent than IdUrd, AIU does provide effective therapy for herpes keratitis.

The antiviral effect of the nucleoside analog 5-iodo-5'-amino-2',5'-dideoxyuridine (AIU) was tested with three isolates of varicella-zoster virus (VZV). AIU concentrations of 10 to 800 muM (3.5 to 288 mug/ml) reduced the number of plaques produced by VZV-infected cells and cell-free VZV from approximately 30 to 95%. Smaller plaque size was also observed in the presence of AIU. AIU was less effective than arabinofuranosylthymine in reducing VZV-induced plaques since as little as 5 mug of arabinofuranosylthymine per ml completely blocked plaque formation by cell-free VZV. Toxicity assays with human diploid embryo fibroblast cells were also carried out. Drug concentrations as high as 800 muM were not toxic to human diploid embryo fibroblast cells as determined by radiolabeling of cell deoxyribonucleic acid, ribonucleic acid, and protein.

Radioactive 5'-amino-5'-deoxythymidine (5'-AdThd), an inhibitor of herpes simplex virus, was synthesized by direct amination of monotosylated thymidine with concentrated ammonium hydroxide. Incubation of 5'-AdThd with purified herpes simplex virus type 1-encoded pyrimidine deoxyribonucleoside kinase produced the diphosphate derivative. The monphosphate derivative of 5'-AdThd was not detected as a reaction product of this enzymatic phosphorylation. A purified mixture of nonviral thymidine kinase and thymidylate kinase derived from uninfected Vero cells was unable to phosphorylate 5'-AdThd under identical conditions. The rate of hydrolysis of 5'-AdThd diphosphate increased as the pH of the reaction mixture decreased, with a shoulder region appearing between pH 3 and 5. The rate of hydrolysis was markedly increased below pH 3. 5'-AdThd, but not the monophosphate derivative, was detected in the hydrolysis mixture. The hydrolysis of 5'-AdThd diphosphate pH 3 also yielded some thymine in addition to 5'-AdThd.

Nonviable, 5-iodouracil (IUra)-substituted bacteriophage T4td8 can be multiplicity reactivated. The data indicate that two nonviable, IUra-substituted T4td8 phage can complement each other intracellularly to produce viable progeny. Phage particles in lysates of T4td8-infected Escherichia coli BT(-), prepared in the presence of varying mole fractions of IUra plus thymine, were examined by infecting with low and high dilutions of lysate. The yields of multiplicity reactivable particles were identical, regardless of the mole fractions of IUra present in the growth media. However, the yields of viable phage, measured at low multiplicities of infection, decreased with increasing mole fraction of IUra. The results are consistent with the hypothesis that the lethal effect of IUra is a consequence of its incorporation into DNA. Further, the IUra-induced lesion cannot involve genetic damage that shuts off expression at a single region of the genome.

We developed a new model of oral HSV-1 infection in mice. After oral inoculation, 100 percent of mice developed the clinical lesions at the inoculated area and latent HVS infection in their trigeminal ganglia without mortality. The antiherpetic efficacy of AIdUrd, an agent specifically activated by herpesvirus-encoded enzyme, has been evaluated in this animal model. Early topical or systemic treatment of AIdUrd notably reduced the development of clinical lesions and the virus content in the inoculated lips. However, the establishment of latent HSV infection in the sensory ganglia was not influenced by AIdUrd treatment.

Various 3'-azido, 3'-amino, 2',3'-unsaturated, 2',3'-dideoxy, and 5-substituted analogues of pyrimidine deoxyribonucleosides have been prepared and tested against Moloney-murine leukemia virus (M-MULV), a mammalian T-lymphotropic retrovirus in vitro. Among these compounds, the 3'-azido analogues of thymidine, 2'-deoxy-5-bromouridine, and 2'-deoxy-5-iodouridine, the 2',3'-unsaturated analogue of thymidine and and 2'-deoxycytidine, and 2',3'-dideoxycytidine were found to be most active, with ED50 values of 0.02, 1.5, 3.0, 2.5, 3.7, and 4.0 microM, respectively. These active compounds were nontoxic to the host SC-1 cells up to 100 microM concentration. The 3'-azido analogues of thymidine and 2'-deoxy-5-bromouridine were also tested in vitro against HTLV-III/LAV/AAV ("AIDS" virus) and found to be significantly active, with ED50 values of 0.23 and 2.3 microM, respectively. The structure-activity relationships are discussed.

Several platinum(II) complexes of 3',5'-diamino-3',5'-dideoxythymidine (compound 1), 5'-amino-5'-deoxythymidine (compound 2), and 3'-amino-3'-deoxythymidine (compound 3) and the respective 2'-deoxyuridine amino nucleoside complexes, 4-6, have been synthesized. Whereas compounds 1, 2, and 4-6 had no inhibitory effect on the replication of murine L1210 cells in cell culture, compound 3 [(3'-AdThd)2PtCl2] inhibited these cells with an ED50 of 0.8 microM. Incubation of L1210 cells with 10-20 microM compound 3 for 2 h produced less than 18% inhibition of RNA, DNA, or protein synthesis, which is of questionable significance. However a 16-h incubation resulted in an increased uptake of labeled thymidine into DNA (77%), labeled uridine into RNA (17%), and labeled amino acids into protein (100%). These unexpected results indicate that inhibition of macromolecules may not be involved in the inhibition of the replication of L1210 cells. The increased incorporation of labeled metabolites into macromolecules may be related to the increase in cell volume after a 2-h incubation of L1210 cells with compound 3 plus a marked increase after 2 h in the proportion of cells in their S phase. Compound 3 appears to delay the progression of cells through their cell cycle. A marked inhibitory effect on the transport of methionine or aminoisobutyric acid into L1210 cells was found with compound 3, which was slightly greater than that produced with cisplatin. Compound 3 had a dose-dependent effect on the survival of mice bearing the L1210 ascites neoplasm, with a T/C X 100 of 175 at a dose of 320 mg/kg. Investigation of the kinetics of decomposition in aqueous systems demonstrated that the primary UV-absorbing decomposition product is 3'-amino-3'-deoxythymidine and that only a limited amount of the compound is formed (less than 8%). Although 3'-amino-3'-deoxythymidine could account for a part of the inhibition of the replication of L1210 cells in culture, it cannot account for the inhibition of amino acid transport by compound 3, the platinum complex of 3'-amino-3'-deoxythymidine. Compound 3 has been shown to limit part of the amino acid uptake into L1210 cells in a similar manner to cisplatin.

5-Iodo-5'-amino-2',5'-dideoxyuridine (AldUrd), given as five single i.p. injections on Days 0 to 4 after antigenic challenge with sheep erythrocytes, had no demonstrable effect on serum hemagglutinin titers in doses as high as 2000 mg/kg/day. This was the maximum feasible single dose, but no 10% lethal dose was determinable. Similarly, 5-iodo-2'-deoxyuridine (IdUrd), 50 mg/kg/day (10% lethal dose), on Days 0 to 4 did not significantly affect this humoral response. However, with a more sensitive assay, immunocytoadherence, reductions in the number of hemagglutinin-forming cells in the spleen were found at several levels of AldUrd and IdUrd, but the same level of inhibition was attained by a course of AldUrd, 2000 mg/kg, or IdUrd, 50 mg/kg. Spleen cell-mediated immunity against lethally irradiated L1210 was measured by 4-hr 51Cr release and 48-hr growth inhibition assays. Both drugs showed dose-related immunosuppression. With AldUrd, 2000 mg/kg/day, and IdUrd, 100 mg/kg/day, on Days 0 to 4, cytotoxicity was inhibited by 35 to 68% and 73 to 90%, respectively. In comparison, a similar course of 1-beta-D arabinofuranosylcytosine, 40 mg/kg/day, completely abrogated both humoral and cell-mediated immunity. When AldUrd and IdUrd were administered on Days -5 to -1, little effect on either type of immunity was found, while pretreatment with the alkylating agent, cyclophosphamide, abolished all T-cell-mediated killing as measured on Day 7. Thus, AldUrd appears to be a very mild and IdUrd a moderate to strong cell cycle-dependent immunosuppressive.

As part of our studies on the molecular basis for the antiherpes activity of 5'-AdThd (5'-amino-5'-deoxythymidine), a study of the HSV-1 DNA synthesized in infected Vero cells exposed to 5'-AdThd was undertaken. Unlike many other antiviral nucleoside analogs, 5'-AdThd did not inhibit HSV-1 DNA synthesis. Analysis of the DNA synthesized in the presence of [14C]5'-AdThd revealed that the analog was incorporated into the viral DNA in a dose-dependent manner and that the degree of incorporation correlated with the antiviral activity as measured by yield reduction assays. Analysis of the 5'-AdThd substituted DNA by centrifugation in neutral and alkaline sucrose gradients revealed no double-stranded breaks but an increase in single-stranded breaks, at very high concentrations of the analog. Analysis of HSV-1-specific RNAs revealed a shift from poly(A+) to poly(A-) RNA. The degree of this shift paralleled the substitution of 5'-AdThd for thymidine in the HSV-1 DNA.

The 5'-hydroxymethylene function of several uracil and cytosine nucleosides has been modified to produce a variety of uronic acids, uronates, uronamides, and urononitriles of 2'-deoxy-beta-D-erythro-pentofuranosyl- and beta-D-arabino-pentofuranosylpyrimidines. In addition, the 5 position in many of these nucleosides has been substituted by a halogen atom. Twenty-one of the 35 compounds synthesized and examined for biological activity have not been previously reported. The purity of the products was measured by a high-pressure liquid chromatographic method. They were then evaluated as potential growth inhibitors of murine Sarcoma 180 cells in culture, of herpes simplex virus type 1 in vitro, and of Streptococcus faecium, a folic acid or deoxythymidine dependent bacterial strain. The ability of these nucleoside analogues to inhibit the phosphorylation of deoxythymidine by herpes simplex virus type 1 encoded pyrimidine deoxyribonucleoside kinase was also investigated and a structure-activity relationship examined.

The structure and conformation of 2',3'-dideoxycytidine, a potent inhibitor of the human immunodeficiency virus, was determined by X-ray crystallography. The nucleoside crystallizes in the tetragonal space group P4(1)2(1)2 with cell dimensions a = b = 8.698(4) and c = 26.155(9) A. Atomic parameters were refined by full-matrix least squares to a final value of R = 0.037 for 1926 observed reflections. The conformation of the furanose ring corresponds to the unusual C3'exo/C4'endo (3T4) pucker, similar to that found in one of the molecules of 3'-azidothymidine (AZT). The glycosidic torsion angle is also smaller than expected. The relevance of these unusual structural features to anti-AIDS activity is assessed.

Immunological function was analyzed in mice that received daily inoculations of 5'-amino-5'-deoxythymidine (1000 mg/kg/day, i.p.) or 5-iodo-2'-deoxyuridine (100 mg/kg/day, i.p.) following antigenic stimulation. 5' Amino-5'-deoxythymidine did not suppress the development of: (1) delayed type hypersensitivity response to sheep red blood cells, (2) cell-mediated cytotoxicity response to allogeneic tumor cells, and (3) IgM and IgG antibody response to sheep red blood cells. In contrast 5-iodo-2'-deoxyuridine suppressed the development of all of these responses.

The antiherpesvirus agent (E)-5-(2-bromovinyl)-2'-deoxyuridine caused marked alterations in the synthesis and processing of several herpes simplex virus type 1 (HSV-1)-infected-cell polypeptides. Analogous to other thymidine analogs, there was a dose-dependent decrease in several beta and gamma polypeptides and an accumulation of HSV-1 thymidine kinase. In contrast to the action of other thymidine analogs, there were alterations in alpha polypeptides, including an increase in the synthesis and phosphorylation of infected-cell polypeptide 4b and a decrease in the synthesis of infected-cell polypeptide 27. The phosphorylation of several other HSV-1 phosphoproteins was mildly inhibited. (E)-5-(2-Bromovinyl)-2'-deoxyuridine inhibited the glycosylation of the major HSV-1 glycoproteins, and this activity appeared to be independent of the incorporation of the drug into the viral DNA. Thus, the alterations in HSV-1 polypeptide expression appear to be due to the presence of the drug in a low-molecular-weight form as well as its presence in the viral DNA. This suggests that this analog or a phosphorylated derivative might act as an inhibitor of an enzyme(s) responsible for posttranslational modification of polypeptides.

At present, only a few drugs have been approved by the FDA for therapy of viral infections in humans. There is a great need for antiviral drugs with increased potency and decreased toxicity, as well as drugs to treat viral diseases for which no drug or vaccine is currently available. Two approaches for development of antiviral drugs are described--an empirical strategy and a rational strategy--with several examples of each. Although many compounds have potent antiviral activity in cell culture, only a small fraction of these will go on to become antiviral drugs for use in humans. At this time, only seven synthetic compounds and alpha interferon have been approved by the FDA for therapy of viral infections in humans. None of these approved drugs are without toxicities, however, and hence there is a great need for antiviral drugs with increased potency and decreased toxicity, as well as for drugs to treat viral diseases for which no drug or vaccine is currently available. Two approaches for the development of antiviral drugs--the empirical and the rational strategies--and their applications and future directions are discussed.

Evidence was obtained which indicates that the lethal effect of 5-iodouracil (IUra) on bacteriophage T4 is not due to a mutagenic process. T4td8rII (thymine requiring, rapid lysis) double mutants were constructed. Reversion of T4td8rII to r(+) was measured. First, reversion by growth in the presence of the structural analogues chlorouracil (ClUra) and bromouracil (BrUra) did not correlate with their relative lethal effects (for mutagenesis: IUra

E-5-(2-Bromovinyl)-2'-deoxyuridine (BrvdUrd) produced a dose-dependent shift in the density of herpes simplex virus type 1 (HSV-1) DNA at concentrations which yielded potent inhibition of virus replication in cultured Vero cells. Although the density of cellular DNA was not altered by these concentrations of BrvdUrd, incorporation of this analogue into cellular DNA of HSV-1-infected cells has been previously observed in this laboratory. The degree of inhibition correlated with the amount of BrvdUrd substituted for thymidine in HSV-1 DNA. BrvdUrd-substituted DNA was more labile as determined by a dose-dependent increase in single strand breaks when examined by centrifugation in alkaline sucrose gradients. Thus, the potent antiviral action of BrvdUrd observed in cell culture correlates not only with its incorporation into HSV-1 DNA but also with an altered stability of this DNA.

In order to define the mechanism for the enhancement by thymidine (dThd) of the antitumor activity of 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU) and 3'-[3-(2-chloroethyl)-3-nitrosoureido]-3'-deoxythymidine (3'-CTNU) in mice, we have investigated the effect(s) of dThd on the uptake of nitrosourea by L1210 cells in culture, DNA alkylation, and repair of the alkyl lesion. Using a rapid centrifugation technique through silicone:paraffin oil, we observe a 1.3- and 1.5-fold increase in the uptake of radioactivity from 0.1 mM [chloroethyl-14C]BCNU in the presence of a 5- and 25-fold excess of dThd, respectively. Similarly, an enhancement of DNA alkylation was observed upon treatment of L1210 cells for up to 3 h with 0.1 mM [chloroethyl-14C]BCNU from 70 pmol 14C/mg DNA in control to 85, 95, and 120 pmol 14C/mg DNA with equimolar 5- and 25-fold excess dThd, respectively. No effect of dThd on the uptake of 0.1 mM [chloroethyl-14C]-3'-CTNU was observed, although a small increase in DNA alkylation at 3 h was evident. DNA repair, as measured by the amount of radioactivity remaining associated with the DNA after an initial 2-h treatment with labeled BCNU was largely unaffected by dThd. Although dThd appears to enhance the cellular uptake of BCNU and the alkylation of DNA by both BCNU and 3'-CTNU, dealkylative repair proceeds unhindered in the presence of dThd.

Herpes simplex virus type 1 (HSV-1) encoded thymidine kinase converts 5-iodo-5'-amino-2',5'-dideoxyuridine (AIdUrd), a highly specific anti-herpes agent, into the 5'-diphosphate (AIdUDP) derivative in vitro. AIdUDP was identified by its acid lability, sensitivity to alkaline phosphatase hydrolysis, chromatographic behavior, and ratio of double isotope (125I, 32P) labeling. ATP, but not AMP, is a phosphate donor, and the direct transfer of the beta and gamma phosphate of ATP as pyrophosphate to AIdUrd was ruled out. The presence of a phosphoramidate bond was supported by the acid lability of AIdUDP which has a half life (t1/2) of 320 min at pH 3.0. At neutral pH, the hydrolysis products are AIdUrd and orthophosphate, with AIdUrd monophosphate being the probable hydrolytic intermediate at these pH values. However, at acidic pH, some pyrophosphate was detected in addition to AIdUrd and orthophosphate. AIdUrd competitively inhibited the phosphorylation of thymidine and deoxycytidine. Escherichia coli thymidine kinase, even though 100-fold higher in activity, was unable to phosphorylate AId-Urd under similar incubation conditions.

The nucleoside analogue 1-(2,3-dideoxy-beta-D-glycero-pent-2-enofuranosyl)thymine (d4T, 1) was prepared by ring opening of the 3',5'-anhydro compound 5. This method has been refined such that it can be used to prepare d4T on a large scale. The triphosphate of d4T was also synthesized from 1 in order to examine the mode of action. The in vitro inhibitory activity of d4T was found to be comparable to that of AZT in HIV-infected CEM cells. The triphosphate of d4T (8) and that of AZT inhibited the HIV reverse transcriptase with poly(rA):oligo(dT) as the template:primer with Ki values of 0.032 and 0.007 microM, respectively. The in vitro toxicity of d4T against normal human hematopoietic progenitor cells (CFU-GM) was measured in comparison to AZT. While d4T reduces colony-forming units by 50% at a concentration of 100 microM, it takes only 1 microM AZT to have a similar toxic effect. With erythrocyte burst forming units (BFU-E) the in vitro toxicities for d4T and AZT have comparable ID50 values of 10 and 6.7 microM, respectively.

Treatment of herpes simplex virus type 1 (HSV-1) infected Vero, BHK, BHKtk- and LMtk- cells with 5-iodo-5'-amino-2',5'-dideoxyuridine (AIdUrd) caused increased synthesis of ICP36 and an increase in HSV-1 thymidine kinase (tk) activity at late times of infection. The overproduced ICP36 was identified as the HSV-1 encoded tk protein by immunoprecipitation. Whereas the thymidine analogue 5'-amino-5'-deoxythymidine (AdThd) caused an increase in HSV-1 tk synthesis and activity in wild type Vero and BHK cells, 5-iodo-2'-deoxyuridine (IdUrd) caused a similar increase only in tk- cells (LMtk-, BHKtk-). In vivo and in vitro stabilization studies using a [35S]methionine pulse-chase experiment or heat inactivation studies with purified HSV-1 tk revealed that stabilization of tk by the analogues could not account for the extent of the observed increase. Since overproduction of tk is observed only at late times of infection, it is suggested that the presence of these thymidine analogues in either the viral DNA or the cellular nucleotide pools is responsible for the observed differential effects.

The (chloroethyl)nitrosourea analogues of 2'-deoxyuridine and 2'-deoxy-5-fluorouridine, 3'-[3-(2-chloroethyl)-3-nitrosoureido]-2',3'-dideoxyuridine (3'-CdUNU, 7) and 3'-[3-(2-chloroethyl)-3-nitrosoureido]-2,3'-dideoxy-5-fluorouridine (3'-CFdUNU, 8), have been synthesized by treatment of the corresponding 3'-amino nucleosides with chloroethyl isocyanate, followed by nitrosation of the resulting ureas. Nucleoside nitrosoureas 7 and 8 exhibited marked anticancer activity against L1210 leukemia in tumor-bearing mice. At an optimum dosage level of 40 mg/kg, 7 and 8 produced 90% and 60% "cures" (greater than 60-day survivors), respectively. The structure-activity relationships are discussed.