Inhibition of human immunodeficiency virus (HIV-1/HTLV-IIIBa-L) replication in fresh and cultured human peripheral blood monocytes/macrophages by azidothymidine and related 2',3'-dideoxynucleosides.

Because of the probable role of HIV-infected monocyte/macrophages in the pathogenesis and progression of AIDS, it is essential that antiretroviral therapy address viral replication in cells of this lineage. Several dideoxynucleosides have been shown to have potent in vitro and, in the case of 3'-azido-2',3'-dideoxythymidine (AZT) and 2',3'-dideoxycytidine (ddC), in vivo activity against HIV. However, because these compounds must be phosphorylated (activated) in target cells, and because monocyte/macrophages may have levels of kinases that differ from those in lymphocytes, we investigated the capacity of these drugs to suppress HIV replication in monocyte/macrophages using HIV-1/HTLV-IIIBa-L (a monocytotropic isolate). In the present study, we observed that HTLV-IIIBa-L replication in fresh human peripheral blood monocyte/macrophages was suppressed by each of three dideoxynucleosides: 3'-azido-2',3'-dideoxythymidine (AZT), 2',3'-dideoxycytidine (ddC), and 2',3'-dideoxyadenosine (ddA). Similar results were observed in 5-d-cultured monocyte/macrophages, although higher concentrations of the drugs were required. We then studied the metabolism of AZT and ddC in such cells. The phosphorylation of ddC to a triphosphate moiety was somewhat decreased in monocyte/macrophages as compared with H9 T cells. On the other hand, the phosphorylation of AZT in monocyte/macrophages was markedly decreased to 25% or less of the level in T cells. However, when we examined the level of the normal endogenous 2'-deoxynucleoside triphosphate pools, which compete with 2',3'-dideoxynucleoside triphosphate for viral reverse transcriptase, we found that the level of 2'-deoxycytidine-triphosphate (dCTP) was six- to eightfold reduced, and that of 2'-deoxythymidine-triphosphate (dTTP) was only a small fraction of that found in T cell lines. These results suggest that the ratio of dideoxynucleoside triphosphate to normal deoxynucleoside triphosphate is a crucial factor in determining the antiviral activity of dideoxynucleosides in HIV target cells, and that the lower levels of dTTP may account for the antiretroviral activity of AZT in the face of inefficient phosphorylation of this compound.

Arabinosyl-5-azacytosine: mechanisms of native and acquired resistance.

Factors influencing the activity of the nucleoside analogue arabinosyl-5-azacytosine (ara-AC) were studied in P388 murine lymphoblasts in vitro and in vivo, in variants of these cells with artificially acquired resistance, in the naturally resistant colon 38 carcinoma in vivo, and in a panel of six human tumors maintained in continuous culture. Differences were noted not only between the sensitive and artificially developed resistant variants of P388, but also between the naturally sensitive (P388) and naturally resistant (colon 38) tumors. The artificially developed resistant P388 cell lines showed an inhibited capacity to accumulate nucleotides derived from ara-AC and deoxycytidine, whereas the accumulation of cytidine nucleotides remained unchanged. Studies of the initial velocity of facilitated diffusion of ara-AC showed only minor differences between parental and resistant lines, while the nucleotide formation rates from both ara-AC and deoxycytidine were markedly depressed in the latter cells. It is concluded, therefore, that the failure of resistant P388 cells to accumulate these compounds results not from a transport deficit per se but rather from a failure to convert the nucleosides to nondiffusible (i.e., phosphorylated) species inside the cell. This failure was accompanied by a substantial reduction in the incorporation of a radiolabeled product derived from deoxycytidine into the nucleic acids of the resistant clones. The common factor responsible for the resistance of P388 variants toward ara-AC appears to be a markedly decreased level of deoxycytidine kinase activity. The naturally resistant colon 38 carcinoma, on the other hand, in addition to a decrease in the activity of its deoxycytidine kinase, showed a lower level of activity of all its purine and pyrimidine kinases, along with a notably elevated nucleoside triphosphatase activity (with ATP as substrate) when compared to P388. These differences were reflected in lower endogenous nucleoside triphosphate pool sizes in colon 38, and in a lower level of ara-AC-5'-triphosphate accumulation in colon 38 than in P388 after comparable drug exposure. In the six human tumor lines, a positive correlation was established between sensitivity to ara-AC (as determined by its median inhibitory concentration) and cellular content of deoxycytidine kinase. It is concluded that this latter enzyme is a generally important determinant of sensitivity to arabinosyl-5-azacytosine.

Thiazole-4-carboxamide adenine dinucleotide (TAD). Analogues stable to phosphodiesterase hydrolysis.

Thiazole-4-carboxamide adenine dinucleotide (TAD), the active metabolite of the oncolytic C-nucleoside tiazofurin (TR), is susceptible to phosphodiesteratic breakdown by a unique phosphodiesterase present at high levels in TR-resistant tumors. Since accumulation of TAD, as regulated by its synthetic and degradative enzymes, appears to be an important determinant for sensitivity to the drug, a series of hydrolytically resistant phosphonate analogues of TAD were synthesized with the intent of producing more stable compounds with an ability to inhibit IMP dehydrogenase equivalent to TAD itself. Isosteric phosphonic acid analogues of TR and adenosine nucleotides were coupled with activated forms of AMP and TR monophosphate to give dinucleotides 2 and 4. Coupling of protected adenosine 5'-(alpha, beta-methylene)diphosphate with isopropylidene-TR in the presence of DCC afforded compound 3 after deprotection. These compounds are more resistant than TAD toward hydrolysis and still retain potent activity against IMP dehydrogenase in vitro. beta-Methylene-TAD (3), the most stable of the TAD phosphonate analogues, produced a depletion of guanine nucleotide pools in an experimentally induced TR-resistant P388 tumor variant that was superior to that obtained with TR in the corresponding sensitive line.

Ribavirin, tiazofurin, and selenazofurin: mononucleotides and nicotinamide adenine dinucleotide analogues. Synthesis, structure, and interactions with IMP dehydrogenase.

A series of dinucleotides, analogous to nicotinamide adenine dinucleotide but containing the five-membered base nucleosides tiazofurin (1a), selenazofurin (1b), ribavirin (2), and AICAR (3) in place of nicotinamide ribonucleoside, were prepared in greater than 50% yield by reacting the corresponding nucleotide imidazolidates (6a-d) with adenosine 5'-monophosphate (AMP). The symmetric dinucleotides of tiazofurin (TTD, 8e) and selenazofurin (SSD, 8f) were also prepared by a similar methodology. These dinucleotides were characterized by 1H NMR and fast atom bombardment MS and were evaluated for their inhibitory potency against a partially purified preparation of tumoral inosine monophosphate dehydrogenase (IMPD) from P388 cells. The order of potency found was SAD (8b) greater than TAD (8a) much greater than SSD (8f) congruent to TTD (8e) congruent to RAD (8c) much much greater than ZAD (8d). On kinetic analysis none of the dinucleotides produced competitive inhibition of IMPD with NAD as a variable substrate. In addition to their superior IMPD inhibitory activity, SAD and TAD appear to be the only dinucleotides, besides NAD, that are formed naturally by the NAD pyrophosphorylase from P388 lymphoblasts.

The disposition and metabolism of tiazofurin in rodents, rabbits, and dogs.

The pharmacokinetics and metabolism of tiazofurin (2-beta-D-ribofuranosylthiazole-4-carboxamide) have been examined in the mouse, rat, rabbit, and dog using tritiated drug as a marker. In all four species, tiazofurin, given as a single bolus iv injection, is removed from the circulation in a triphasic manner, with a generally prolonged terminal half-life. In all cases, the mean concentration of unchanged drug prevailing during this terminal phase was well within the cytotoxic range (IC50 vs. P388 cells is 2 microM in vitro). Urinary excretion accounted for between approximately 40 and 90% of the administered dose in all four species, with only minor quantities (less than 3%) of drug-derived radioactivity detected in the feces. The metabolism of tiazofurin was examined in mice and rats: although no evidence was uncovered for hydroxylation of tiazofurin at carbon atom 5 of the thiazole ring, phosphorylation of the drug at its 5'-hydroxyl was demonstrable in nearly every organ of both species, but, liver, striated muscles, and kidney were the only tissues catalyzing the synthesis of thiazole-4-carboxamide adenine dinucleotide to any prominent degree. This synthesis did not appear to be a saturated process, even at doses as high as 8000 mg/m2. Since rodent skeletal muscle accumulated high concentrations of tiazofurin phosphates in vivo, it is suggested that musculature may serve as a reservoir for the drug, and contribute to its rather protracted terminal half-life in plasma.

Interaction of digitalis and spironolactone with human sex steroid receptors.

Spironolactone and digitoxin have previously been shown to interact with cytosol androgen and estrogen receptors, respectively, in the rat. The interaction of digitoxin with human uterine cytosol estrogen binding protein and spironolactone with human prostate and newborn prepuce cytosol dihydrotestosterone (DHT) binding protein has been analyzed in this study. Specific estradiol binding was found only in premenopausal uteri. The dissociation constant for estradiol binding was 0.6--2.3 X 10(-9) M (n - 12). Digitoxin in concentrations varying between 0.5--2.0 X 10(-6) M inhibited specific estradiol binding with a Ki of 2.0--7.3 X 10(-7) M (n = 9). The dissociation constants for DHT and the human androgen cytosol binding protein in prostate and newborn prepuce were 0.27--3.0 X 10(-8) M (n = 12) and 0.6--2.0 X 10(-8) M (n = 5), respectively. Spironolactone at concentrations of 0.3--2.0 X 10(-6) M competitively inhibited this binding with an affinity about one order of magnitude less than that of DHT. Digitoxin and spironolactone did not displace estradiol and DHT, respectively, from testosterone-estrogen binding globulin in male or female plasma. The interaction of digitoxin with the human uterus estrogen binding protein and spironolactone with the human prostate and prepuce androgen binding protein is similar to our previous observations in the rat, and may explain the weak estrogenic effects of digitoxin and spironolactone in man.