Search

An alternative method to conduct a Barton-McCombie deoxygenation in nucleosides is described. The utility of the procedure is limited to structures with an electronegative substituent, particularly fluorine, in the beta-position relative to the radical center. The process is radical in nature and triggered by peroxides. The abstraction of hydrogen from the solvent is favorably influenced by the presence of a beta-fluorine.

Measurement of intracellular drug levels in cell culture systems can be of predictive value in establishing rational clinical dosage schedules. Such in vitro measurements carried out with anti-HIV agents of the 2',3'-dideoxynucleoside (ddN) class have shown that many of the pharmacologically active ddNTP metabolites of these agents have relatively long intracellular half-lives and little or no host-cell cytotoxicity. As a consequence, replication of drug-exposed cells continues at an unperturbed rate so that a systematic dilution error occurs in the measurement of ddNTP decay half-times. The aim of this study is to present a simple general formulation for the correction of measured t1/2-values for ddNTPs and for other agents with similar intracellular pharmacokinetic properties. Two factors of practical interest emerge: first, the error is greater for agents with slow intracellular clearance rates than for agents with rapid rates; and second, for cell lines with long doubling times, the measured t1/2-values approach more closely to the true t1/2-values, until with the extreme case (quiescent or "G(o)" cells), the observed and true decay times are identical. The greatest dilution errors are seen with adenodine-based agents such as ddATP and 2'-F-ddATP, while the smallest errors are seen with rapidly cleared agents of the dideoxythymidine class.

2'-beta-Fluoro-2',3'-dideoxyadenosine (F-ddA) is an acid-stable purine dideoxynucleoside analog active against a wide spectrum of human immunodeficiency virus type 1 (HIV-1) and HIV-2 strains in vitro. F-ddA is presently undergoing a phase I clinical trial at the National Cancer Institute. We induced HIV-1 variants resistant to F-ddA by exposing wild-type HIV-1 (HIV-1LAI) to increasing concentrations of F-ddA in vitro. After 18 passages, the virus was fourfold less sensitive to F-ddA than HIV-1LAI. Sequence analyses of the passage 18 virus revealed changes in three amino acids in the reverse transcriptase (RT)-encoding region of the pol gene: P to S at codon 119 (P119S; present in 3 of 13 and 28 of 28 molecular clones before and after F-ddA exposure, respectively), V179D (0 of 13 and 9 of 28, respectively), and L214F (9 of 13 and 28 of 28, respectively). Drug sensitivity assays using recombinant infectious clones confirmed that P119S was directly responsible for the reduced sensitivity of HIV-1 to F-ddA. Various infectious clones with single or multiple amino acid substitutions conferring viral resistance against nucleoside RT inhibitors, including HIV-1 variants with multi-dideoxynucleoside resistance, were generally sensitive to F-ddA. The moderate level of resistance of HIV-1 to F-ddA, together with the lack of conferment of significant cross-resistance by the F-ddA-associated amino acid substitutions, warrants further investigation of F-ddA as a potential antiviral agent for use in treatment of HIV-1 infection.

A new antiretroviral agent, 2'-beta-fluoro-2',3'-dideoxyadenosine (FddA), is an acid-stable compound whose triphosphate form is a potent reverse transcriptase inhibitor with in vitro anti-human immunodeficiency virus (HIV) activity and a favorable pharmacokinetic profile. Severe combined immunodeficiency (SCID) mice reconstituted with human peripheral blood leukocytes (hu-PBL-SCID mice) provide a useful small-animal model for HIV research. In the present study we utilized this experimental system for the in vivo evaluation of the anti-HIV activity of this new compound when administered prior to infection. Initial studies revealed that, following a challenge with 50 100% tissue culture infective doses of HIV type 1 lymphadenopathy-associated virus, 39 of 42 (93%) control mice developed HIV infection, as evidenced by positive coculture or positive PCR. Administration of zidovudine decreased the infection rate to 5 of 16 (31%), while administration of FddA decreased the infection rate to 0 of 44 (0%). In follow-up controlled studies, the anti-HIV activity of FddA was confirmed, with 18 of 20 control mice showing evidence of HIV infection, compared with 4 of 20 FddA-treated mice. In addition to having direct anti-HIV effects, FddA was found to have a protective effect on human CD4+ T cells in the face of HIV infection. Mice treated with FddA were found to have a significantly higher percentage of CD4+ T cells than controls (10.3% +/- 3.4% versus 0.27% +/- 0.21%; P = 0.01). Thus, FddA, with its potent anti-HIV activity in vivo, high oral bioavailability, long intracellular half-life, and ability to preserve CD4+ cells in the presence of HIV, appears to be a promising agent for clinical investigation.

A series of 6-substituted amino analogs of 9-(2,3-dideoxy-2-fluoro-beta-D-threo-pentofuranosyl) purines (F-ddN) has been synthesized and characterized with the objective of finding compounds which might be superior to existing drugs for the treatment of HIV in the central nervous system. These compounds are intended to be more lipophilic than the currently approved anti-HIV drugs for better blood-brain barrier penetration. Subsequent adenosine deaminase (ADA)-catalyzed hydrolysis of these prodrugs in the brain is expected to produce the anti-HIV agent, 9-(2,3-dideoxy-2-fluoro-beta-D-threo-pentofuranosyl)hypoxanthine (F-ddI). The new compounds, synthesized from the corresponding 6-chloro analog, include F-ddN which contain methylamino, ethylamino, dimethylamino, hydroxylamino, methoxyamino, benzyloxyamino, hydrazino, and nitro substituents in the 6-position. The 6-nitro analog was isolated as an unexpected product during the preparation of the 6-chloro derivative. Among the analogs with anti-HIV activity, the ethylamino and dimethylamino compounds are ca. 100 times more lipophilic than ddI or F-ddI. As expected, 2'-fluoro substitution protects the compounds from acid-catalyzed glycosylic cleavage. Only the hydroxylamino and nitro analogs underwent any nonenzymatic hydrolysis at pH 1.0 or 7.4. This reaction, however, results in hydrolysis of the group in the 6-position rather than glycosylic bond cleavage. ADA catalyzes the hydrolysis of the 6-substituents at rates which vary from slightly slower (NO2, 1.7x) to much slower (NHEt, 5000x) than F-ddA. The 6-dimethylamino analog is the only compound which possesses anti-HIV activity (ED50 18 microM) without ADA hydrolysis. With the exception of the two inactive alkoxyamino compounds, the other prodrugs exhibited cellular protection in the HIV-1/PHA-PBM system with IC50 potencies of 7-40 microM.

Incorporation of 2'-beta-fluoro-2',3'-dideoxyadenosine (F-ddA), a recently developed anti-HIV agent, into the cellular DNA of human MOLT-4 cells has been compared with the DNA incorporation seen with fialuridine (FIAU; 1-[2'-deoxy-2'-fluoro-beta-D-arabinofuranosyl]-5-iodouracil), a potent anti-hepatitis B (anti-HBV) nucleoside analogue recently found to cause severe hepatic toxicity in human subjects. At equimolar concentrations (10 microM), incorporation of F-ddA was less than 1% of that for FIAU, a difference attributable to the lack of a 3'-hydroxyl group in the former compound and a consequent inability of F-ddA, unlike FIAU, to form DNA internucleotide linkages.

2'-beta-Fluoro-2',3'-dideoxyadenosine (F-ddA), an acid-stable, purine dideoxynucleoside with in vitro anti-HIV activity, has been selected by the NCI as a clinical trial candidate. A recent report that high, single doses of F-ddA produce cardiotoxicity in rats prompted the present investigation whose objective was to quantitate this effect and establish a relationship between this toxicity and F-ddA plasma concentrations. Microscopic examination of cardiac tissues for degenerative lesions established the effects of F-ddA and ddA on three iv schedules [daily x 1(2.5-250 mg/kg); daily x 5(125, 250 mg/kg), and BID x 1 (250 mg/kg)] as well as one oral schedule [BID x 1 (500 mg/kg) using 8- to 12-week old female Sprague-Dawley rats. For both F-ddA and ddA, the group mean severity of the cardiac lesions was dose-dependent and proportional to the measured plasma concentrations of the undeaminated parent drugs. F-ddI and ddI, were essentially nontoxic in this study (iv, 250 mg/kg, daily x 1 and daily x 5), since plasma concentrations exceeding 2 mM produced only minimal cardiac lesions. The cardiomyopathy of F-ddA was minimal to mild for all iv doses except 250 mg/kg (daily x 1) and usually was greater than that of ddA at any given dose. This is a consequence of the fact that F-ddA is deaminated 20 times more slowly than ddA, resulting in higher plasma concentrations of F-ddA relative to ddA at any given time for any given dose. Neither F-ddA nor ddA was more cardiotoxic on a repeated iv schedule (daily x 5) than when administered only once, suggesting that rat cardiotoxicity is related Cmax rather than total exposure. In this most sensitive species, the formation of cardiac lesions above the background level is associated with i.v. F-ddA administration when the F-ddA plasma concentration approaches 300 microM, 30-50 times the anticipated therapeutic level in humans.

Ribonucleotide reductase inhibitors such as hydroxyurea (HU) and related compounds, at low, non-toxic doses, enhance the anti-human immunodeficiency virus type 1 (HIV-1) potency of both purine and pyrimidine 2',3'-dideoxynucleosides (ddNs) in human lymphocytes and macrophages. The most marked enhancement of inhibition of HIV-1 replication reported to date has been seen with the purine ddN 2',3'-dideoxyinosine (ddIno): a low level of HU (0.1 mM) permitted a 4.5-fold reduction in optimal ddIno dosage with no decrease in therapeutic effect or increase in toxicity. We report here even more marked enhancement by HU of the potency of the purine ddN 2'-beta-fluoro-2',3'-dideoxyadenosine (2'-beta-F-ddAdo), where the addition of 0.1 mM HU permitted a 7.1-fold reduction in the optimal dose of 2'-beta-F-ddAdo in the phytohemagglutinin-activated peripheral blood mononuclear cell HIV-1 test system.

Flexible molecules change their conformation upon binding to a protein. This was shown by the analysis of small molecules whose structures have been determined by X-ray crystallography of both the pure compound and the compound bound to a protein. Thirty-three compounds present both in the Cambridge Structural Database and the Brookhaven Protein Data Bank were analyzed, and both were compared with the global energy minimum conformation calculated by the molecular mechanics program CHARMm. It was found that the conformation bound to the protein differs from that in the crystal structure and also from that of the global energy minimum, and the degree of deformation depends upon the number of freely rotatable bonds in the molecule. Analysis of the conformational energies of the flexible molecules showed that, for most of those compounds, both the crystal and the protein-bound conformations are energetically well above the global minimum, and, in many cases, not even in any local energy minimum. Semi-empirical calculations performed for a select number of structures, using both the AM1 and PM3 hamiltonians, confirmed these results. These findings are discussed as to their impact upon contemporary methods of drug design.

A series of 6-halo-(F-, Cl-, Br-, I-) and 6-alkoxy-(OMe-, OEt-) 9-(2,3-dideoxy-2-fluoro-beta-D-threopentofuranosyl) purines (F-ddN) have been synthesized and characterized with the objective of finding compounds which might be superior to existing drugs for the treatment of HIV in the central nervous system. These compounds, which contain lipophilic 6-substituents, were chosen as acid-stable prodrugs for the anti-HIV-active F-ddN, 9-(2,3-dideoxy-2-fluoro-beta-D-threo-pentofuranosyl) hypoxanthine (F-ddI), because of their potential to increase blood-brain-barrier penetration relative to F-ddI. All the new compounds were more lipophilic than the currently approved anti-AIDS drugs. Partition coefficient increases of 30- and 110-fold were achieved, relative to didanosine (ddI), for the 6-chloro- and 6-ethoxy analogues. 2'-Fluoro substitution abolished the pH 1, acid-catalyzed cleavage of the nucleoside glycosylic bond. However, pH 1, acid-catalyzed hydrolysis of the 6-fluoro substituent to produce F-ddI was observed to occur at a rate (t1/2 0.54 h) which was ca. 40-170 times faster than that of the other prodrugs. The utility of the F-ddNs as prodrugs for F-ddI depends upon their ability to act as substrates for adenosine deaminase. The relative rates of adenosine deaminase-catalyzed prodrug hydrolysis to F-ddI varied by a factor of > 25,000 with the 6-fluoro- and 6-ethoxy analogues reacting the fastest and slowest, respectively. All of the prodrugs possessed anti-HIV activity in the phytohemagglutinin-stimulated peripheral blood mononuclear cell test system and a qualitative correlation exists between prodrug anti-HIV activity and adenosine deaminase hydrolysis rates.

Among nine compounds surveyed, cytidine was found to be the most effective in reversing the antiproliferative effects of cyclopentenyl cytosine (CPEC) on human T-lymphoblasts (MOLT-4) in culture. Cytidine, at concentrations of 1-25 microM, enabled cells to maintain normal logarithmic growth when added up to 12 hr after exposure to a 200 nM concentration of the oncolytic nucleoside, CPEC. The most abundant CPEC metabolite, CPEC-5'-triphosphate, is a potent [K1 approximately 6 microM] inhibitor of CTP synthetase (EC 6.3.4.2). Accumulation of this inhibitor resulted in a depletion of CTP levels to 17% of their original cellular concentration. Exogenous cytidine reversed CPEC-induced cellular cytotoxicity by suppressing the formation of CPEC-5'-triphosphate by 70%, and by partially replenishing intracellular CTP to at least 60-70% of its original concentration. In vivo, cytidine (500 mg/kg) administered intraperitoneally 4 hr after each daily dose of CPEC (LD10-LD100) for 9 days reduced the toxicity and abolished the lethality of CPEC to non-tumored mice. Of greater practical importance is the finding that, under these experimental conditions, cytidine did not curtail the antineoplastic properties of CPEC in L1210 tumor-bearing mice. Moreover, the concentration range over which CPEC exhibited antineoplastic activity was extended with cytidine administration.

The anti-human immunodeficiency virus agents 2',3'-dideoxyadenosine (ddAdo) and 2'-beta-fluoro-2',3'-dideoxyadenosine (2'-beta-F-ddAdo) are rapidly converted, both in vitro and in vivo, to the corresponding inosine analogs by the widely distributed enzyme adenosine deaminase (EC 3.5.4.4). We have determined the effects of the potent adenosine deaminase inhibitor 2'-deoxycoformycin (2'-dCF) on ddAdo and 2'-beta-F-ddAdo metabolism in MOLT-4 cells and on ddAdo antiviral activity in the ATH8 test system. At levels as low as 5 nM in the incubation medium, 2'-dCF effectively blocks the extracellular deamination of both agents, thus permitting their rapid cellular uptake as the unchanged parent compounds, rather than as the less lipid-soluble 2',3'-dideoxyinosine or 2'-beta-fluoro-2',3'-dideoxyinosine. The result is a significant increase in intracellular levels of the pharmacologically active forms 2',3'-dideoxyadenosine-5'-triphosphate and 2'-beta-fluoro-2',3'-dideoxyadenosine-5'-triphosphate. The effect becomes maximal over the range of 50-250 nM 2'-dCF and declines to control levels when extracellular 2'-dCF levels exceed 1 microM. This decrease in ddAdo and 2'-beta-F-ddAdo phosphorylation with higher levels of the inhibitor appears to result from intracellular penetration of 2'-dCF and consequent inhibition of intracellular deamination, a critical step in the activation of both agents through the 5'-nucleotidase pathway. In anti-human immunodeficiency virus assays, a 2.2-fold increase in ddAdo antiviral potency was seen at 2'-dCF levels of 20 and 50 nM.

Many dideoxynucleosides that are effective against human immunodeficiency virus (HIV) also are potent inhibitors of mitochondrial DNA (mtDNA) synthesis, and the resulting mtDNA decrease could be responsible for the delayed clinical toxicity sometimes observed with these drugs. The following compounds have been examined for their toxicity to human lymphoid CEM cells, and their ability to suppress mtDNA content: 2',3'-dideoxycytidine (ddC), 2',3'-dideoxyadenosine (ddA), 2',3'-dideoxyinosine (ddI) and 2',3'-dideoxyguanosine (ddG); and their 2'-beta-fluoro analogs; beta-F-ddC, beta-F-ddA, beta-F-ddI and beta-F-ddG. Two other fluoro analogs, 5-F-ddC and 2'-beta,5-di-F-ddC were also examined. The ratio of C-IC50 (concentration that inhibited cell growth by 50%) to mt-IC50 (concentration that inhibited mtDNA synthesis by 50%) was determined for each compound. The rank-order of this ratio was ddC > 5-F-ddC >> ddA > ddI > ddG > beta-F-ddC > beta-F-ddA > beta-F-ddG with the highest ratios indicating the greatest potential for delayed toxicity. In comparison with ddC, beta-F-ddC and beta-F-ddA were 5,000 and 22,000 times less potent, respectively, in suppressing cellular mtDNA content, while their anti-HIV potencies were decreased only modestly relative to their unfluorinated parent compounds. beta-F-ddI and 2'-beta,5-di-F-ddC produced neither cellular toxicity nor mtDNA suppression at concentrations of 500 and 1000 microM, respectively. Lactic acid, the product of compensatory glycolysis that results from the inhibition of mitochondrial oxidative phosphorylation, was measured after cells were treated with these compounds. There appears to be a concentration-related correlation between the increase of lactic acid and the extent of mtDNA inhibition for the compounds examined.

A searcheable database of three-dimensional structures has been developed from the chemistry database of the NCI Drug Information System (DIS), a file of about 450,000 primarily organic compounds which have been tested by NCI for anticancer activity. The DIS database is very similar in size and content to the proprietary databases used in the pharmaceutical industry; its development began in the 1950s; and this history led to a number of problems in the generation of 3D structures.

The mechanism of divergent anti-human immunodeficiency virus type 1 (HIV-1) activity of various 2',3'-dideoxynucleoside analogs (ddNs) in peripheral blood mononuclear cells (PBM) was studied. We demonstrate that the in vitro anti-HIV-1 activity of various ddNs varies profoundly and that the divergent antiviral activity is related to the extent of anabolic phosphorylation of each ddN and its counterpart 2'-deoxynucleoside (dN). We also show that certain ddNs cause a reduction of their counterpart dNTP formation in PBM in the following order: 2',3'-dideoxycytidine (ddC) >> 2',3'-didehydro-2',3'-dideoxythymidine (d4T), 3'-thia-2',3'-dideoxycytidine (3TC), 2',3'-dideoxyinosine (ddI), 2',3'-dideoxyguanosine (ddG) > 3'-azido-2',3'-dideoxythymidine (AZT) > 2'-beta-fluoro-2',3'-dideoxyadenosine (F-ara-ddA). Based on the phosphorylation profiles, anti-HIV-1 ddNs can be classified into two groups: (i) cell-activation-dependent ddNs such as AZT and d4T that are preferentially phosphorylated, yield higher ratios of ddNTP/dNTP, and exert more potent anti-HIV-1 activity in activated cells than in resting cells; and (ii) cell-activation-independent ddNs including ddI (and 2',3'-dideoxyadenosine), F-ara-ddA, ddG, ddC, and 3TC that produce higher ratios of ddNTP/dNTP and exert more potent anti-HIV-1 activity in resting cells. These data should provide a basis for the elucidation of the mechanism of the divergent antiretroviral activity of ddNs.

In 1979, an unusual, carbocyclic nucleoside was discovered in a Japanese fermentation broth and designated neplanocin A. This compound is an analog of adenosine possessing a cyclopentene-containing "sugar" glycon. Although neplanocin A was biologically active, it was quite toxic. It therefore became a lead compound for analog synthesis in an attempt to maximize antitumor and antiviral activity while minimizing toxicity. First, a total synthesis of naturally occurring (-)-neplanocin A was accomplished using a new, versatile cyclopentenone carbocyclic "sugar" intermediate. This intermediate was then used to synthesize some 20 purine and pyrimidine analogs of neplanocin A which were evaluated for their antitumor and antiviral properties. Among the purine analogs, 3-deazaneplanocin A, a powerful inhibitor of S-adenosylhomocysteine hydrolase, was found to have excellent antiviral activity both in vitro and in vivo. Cyclopentenyl cytosine (CPE-C) was found to be the most biologically active compound among the carbocyclic pyrimidine nucleosides. In addition to activity against over 20 viruses, this compound had excellent preclinical antitumor activity against both murine leukemias and human tumor xenografts. CPE-C is currently under clinical evaluation as an anticancer drug.

The purine dideoxynucleosides 2'-beta-fluoro-2',3'-dideoxyadenosine (2'-beta-F-ddAdo), 2'-beta-fluoro-2',3'-dideoxyinosine, and 2'-beta-fluoro-2',3'-dideoxyguanosine (2'-beta-F-ddGuo) are active inhibitors of the replication of the human immunodeficiency virus (HIV) in the ATH8 assay system, with 2'-beta-F-ddAdo and 2'-beta-fluoro-2',3'- dideoxyinosine showing activity and potency equivalent to those of their respective parent compounds, 2',3'-dideoxyadenosine (ddAdo) and 2',3'-dideoxyinosine. Because inhibitors of IMP dehydrogenase such as ribavirin and tiazofurin stimulate the 5'-phosphorylation and consequently the anti-HIV activity of the three nonfluorinated parent compounds (ddAdo, 2',3'-dideoxyinosine, and 2',3'-dideoxyguanosine), we have undertaken a study in MOLT-4 cells to determine whether a similar stimulatory effect is observed with their 2'-beta-fluorinated analogs. The 5'-phosphorylation of all the fluoro compounds was found to be greatly enhanced by low levels (10 microM) of either ribavirin or tiazofurin, with the greatest increase being seen with 2'-beta-F-ddAdo, where stimulation of the formation of the 5'-mono-, di-, and triphosphorylated nucleotides was approximately 20-fold, 6-fold, and 5-fold, respectively. These increases were approximately 3-fold greater than the increases seen with the nonfluorinated parent compound ddAdo. In the case of 2'-beta-F-ddGuo, the greatest stimulation (8-fold) was seen in the formation of the 5'-diphosphate. In parallel with the increased phosphorylation of 2'-beta-F-ddAdo and 2'-beta-F-ddGuo, the anti-HIV potency of these two compounds at the 5 microM level was approximately doubled in the presence of ribavirin (5 microM).

The synthesis, chemistry, biochemistry, and anti-HIV activity of a series of 1-(2,3-dideoxy-2-fluoro-beta-D-threopentofuranosyl)pyrimidines have been studied in an attempt to find useful anti-AIDS drugs. Synthesis is carried out via a 2,3-dideoxyribose intermediate which facilitates the preparation of analogues by removing the sugar 3'-hydroxyl group prior to, rather than after, condensation with a uracil or cytosine aglycon. The 2'-F-dd-uridine analogues 7a-d (with H, F, Cl, and CH3 substitution in the 5-position) as well as the 4-deoxy compound (12b) are nonprotective to ATH8 or CEM cells infected with HIV-1. In the corresponding cytidine series, the 5-chloro analogue (11) is inactive. However, 2'-fluoro-2',3'-dideoxyarabinosylcytosine, 10a, and its 5-fluoro analogue, 10b, are both active. While neither compounds is a potent as ddC or 5-F-ddC (2b), 10b gives complete protection against the cytopathic effects of HIV in both host cell lines. 2'-Fluoro substitution confers increased chemical and enzymatic stability on dideoxynucleosides. Even though dideoxy pyrimidine nucleosides are inherently more stable than the corresponding purine analogues toward acid-catalyzed cleavage of the glycosidic bond, 2'-fluoro substitution (10a) still increases stabilization relative to ddC (2b). No detectable deamination by partially purified cytidine deaminase is observed with the 2'-fluoro compounds 10a, 10b, or 11 under conditions which rapidly deaminate cytidine. A small amount of 2'-F-dd-ara-U (7a) is formed from 10a in monkey plasma after greater than 24 h of exposure. The octanol-water partition coefficients for the dideoxynucleosides in this study indicate their hydrophilic character, with log P values varying from -0.28 to -1.18.

2(1H)-Pyrimidinone riboside (zebularine, 1b) and its 5-fluoro (6b) and 2'-ara-fluoro (7b) analogues have been synthesized and evaluated in vivo as antitumor agents. Zebularine provides increase in life span (ILS) values of ca. 70% against intraperitoneal (ip) murine B16 melanoma and 50% against P388 leukemia. This compound is active when administered either ip or orally against ip or subcutaneously implanted L1210 leukemia, producing ILS values of about 100% at an optimum dose of 400 mg/kg. 1b is also active (60% ILS) against ara-C-resistant L1210. The analogous unsubstituted purine riboside nebularine (2) has modest activity against P388 leukemia (60% ILS). While 2'-ara-fluorozebularine (7b) is only marginally active (40% ILS) at high doses against L1210 leukemia, 5-fluoro analogue 6b is more active than zebularine and is ca. 100 times more potent. Although the activity of 6b is about the same as that of 1b against P388 leukemia, greater potency also is realized in this model. Zebularine is a strong inhibitor of cytidine deaminase, but in contrast to tetrahydrouridine, 1b is acid-stable. In an attempt to use this property to advantage in oral administration, 1b and ara-C have been orally coadministered to mice with ip L1210 leukemia. When zebularine is given in divided doses, up to a 2-fold increase in activity is realized, relative to treatment with the same dose of ara-C alone.

A series of 6-substituted 2',3'-dideoxypurine ribofuranosides (ddP) was enzymatically synthesized with live E. coli in an effort to enhance the lipophilicity of this class of anti-human immunodeficiency virus (HIV) compounds and thereby facilitate drug delivery into the central nervous system. All 6-halo-substituted ddPs were substantially more lipophilic, as defined by their octanol-water partition coefficient (P), than their nonhalogenated congeners 2',3'-dideoxyinosine (ddI) or 2',3'-dideoxyguanosine (ddG). For this class of compounds, log P's ranged from +0.5 to -1.2 in the following order: 6-iodo, 2-amino-6-iodo greater than 6-bromo, 2-amino-6-bromo greater than 6-chloro, 2-amino-6-chloro greater than 6-fluoro, 2-amino-6-fluoro much greater than ddG greater than ddI. These compounds were evaluated in vitro for ability to suppress the infectivity, replication, and cytopathic effect of HIV. 2-Amino-6-fluoro-, 2-amino-6-chloro-, and 6-fluoro-ddP exhibited a potent activity against HIV comparable to that of ddI or ddG and completely blocked the infectivity of HIV without affecting the growth of target cells. The comparative order of in vitro anti-HIV activity was 2-amino-6-fluoro, 2-amino-6-chloro, 6-fluoro greater than 2-amino-6-bromo greater than 2-amino-6-iodo, 6-chloro greater than 6-bromo greater than 6-iodo. These compounds also exhibited potent in vitro activity against HIV-2 and 3'-azido-3'-deoxythymidine-resistant HIV-1 variants. All 2-amino-6-halo-ddPs and 6-halo-ddPs were substrates for adenosine deaminase (ADA) and were converted to ddG or ddI, respectively. In the presence of the potent ADA inhibitor 2'-deoxycoformycin, 6-halo-substituted ddPs failed to exert an in vitro antiretroviral effect. These dideoxypurine nucleoside analogues represent a new class of lipophilic prodrugs of ddG and ddI that possess the potential for more effective therapy of HIV-induced neurologic disorders.

Selected acid-stable (2'-fluoro-2',3'-dideoxyarabinofuranosyl)adenine nucleosides containing methyl groups and other lipophilic functions at various positions in the adenine ring were prepared and evaluated as anti-HIV agents. The N6-methyl (1f), N6-benzoyl (1g), and 6-chloro (1i) analogues had modest activity, giving 30-50% protection to ATH8 cells infected with HIV. 2-Methyl (1d), 8-methyl (1h), and 2,N6-dimethyl (1e) substitution, as well as N1-oxide (21) formation, abolished the activity of the parent compound (1a). Several of these compounds, originally designed as agents for treating HIV in the central nervous system, were further investigated as substrates for adenosine deaminase (ADA). Kinetic experiments showed that ADA catalyzed the formation of the anti-HIV active inosine compound 1b from the N6-methyl analogue 1f in a quantitative manner. The anti-HIV activity of 1f and 1i was abolished when the ADA inhibitor, 2'-deoxycoformycin, was added to the test mixture. In contrast, the activity of 1f was significantly enhanced when ADA was added to the test system. These data indicate that 1f and 1i are prodrug forms of 1b in systems containing ADA.

Two ring-expanded analogues (compounds 2 and 3) of the anti-HIV fermentation product oxetanocin A (1) were synthesized from commercially available diacetone D-glucose. Antiviral testing against HIV in ATH8 cells revealed that the ring-expanded analogue 2 possessed a similar activity profile as oxetanocin A. Neither compound, however, was capable of providing full protection to the cells against HIV infection. The isomeric ring-expanded analogue 3 was totally devoid of anti-HIV activity. Molecular modeling suggested that while oxetanocin A and compounds 2 and 3 share a large common substructure with the potent anti-HIV drug, dideoxyadenosine (ddA), the extra hydroxymethyl substituent may contribute negatively to the binding of these molecules to a critical enzyme. The negative contribution may be less important in oxetanocin and isomer 2 than in isomer 3. From these studies it would appear that both oxetane and tetrahydrofuran rings are equivalent templates to support the adenine base in terms of anti-HIV activity.

In this report, we have compared the uptake, metabolism, and relevant enzymology of a novel anti-acquired immunodeficiency syndrome drug, 2'-fluoro-2',3'-dideoxyarabinosyladenine (2'-F-dd-ara-A) with the corresponding properties of its parent compound 2',3'-dideoxyadenosine (2',3'-ddAdo) in three human T cell lines, MOLT-4, ATH8, and CEM. In previous communications, we have reported that the primary route of metabolism of 2',3'-ddAdo in human T lymphoblasts is catabolic, i.e., deamination to 2',3'-dideoxyinosine (2',3'-ddlno). At this point, the metabolic pathway diverges, to result in either cleavage and inactivation of 2',3'-ddlno by purine nucleoside phosphorylase or in 5'-phosphorylation by a phosphotransferase, a reaction that generates 2',3'-inosine monophosphate and ultimately the putative active metabolite 2',3'-dideoxy-ATP. Studies with kinase-deficient mutant CEM lines indicate, however, that 2'-F-dd-ara-A favors a more direct anabolic route toward formation of 2'-fluoro-dideoxynucleotides, catalyzed initially by 2'-deoxycytidine kinase. In MOLT-4 cells, amounts of 2'-fluoro-dideoxyarabinosyladenine di- and triphosphate formed were approximately 20-fold and 5-fold greater than the respective accumulation of 2',3'-dideoxy-ADP and 2',3'-dideoxy-ATP over the same time of exposure. This metabolic profile was supported by enzymological studies, which revealed that 2'-F-dd-ara-A is deaminated 10 times less rapidly than ddAdo and that the resulting deaminated product is resistant to hydrolysis by purine nucleoside phosphorylase. Under similar conditions, ddAdo was rapidly degraded through cleavage of its deamination product ddlno. Like ddAdo, 2'-F-dd-ara-A was found to be transported by passive diffusion and does not enter cells via the purine nucleoside transport carrier system. However, the rate of entry of 2'-F-dd-ara-A was about half that of ddAdo (9.7 pmol/10(6) cells/min for 2'-F-dd-ara-A versus 18.4 pmol/10(6) cells/min for ddAdo). This investigation, therefore, demonstrates that, under the conditions studied, 2'-F-dd-ara-A and its deamination product 2'-fluoro-2',3'-dideoxyarabinosylhypoxanthine have metabolic properties that differ significantly from those of their parent compounds ddAdo and ddlno. These properties, combined with the previously reported resistance of the fluorinated nucleosides to acid degradation, make these compounds interesting candidates for further study as orally administered agents for the inhibition of human immunodeficiency virus replication in patients with acquired immunodeficiency syndrome.

2',3'-Dideoxy purine nucleosides have anti-HIV activity in vitro and the inosine analogue is being clinically evaluated. The instability of these compounds toward acidic conditions complicates oral administration. The effect of the addition of a fluorine atom to the 2'-position was investigated by preparing the fluorine-containing 2'-erythro and 2'-threo isomers of ddA and the threo isomer of ddI. All fluorine-containing compounds were indefinitely stable to acidic conditions which completely decomposed ddI (1) and ddA (2) in minutes. While the fluorine-containing erythro isomer, 5, was inactive, the threo isomers, 2'-F-dd-ara-A (3) and 2'-F-dd-ara-I (4), were just as potent and active in protecting CD4+ ATH8 cells from the cytopathogenic effects of HIV-1 as the parent drugs. Exposure to pH 1 at 37 degrees C prior to testing destroyed the activity of ddA and ddI but left the anti-HIV properties of 3 and 4 unchanged. The fluorinated analogues also protected cells exposed to HIV-2 and inhibited gag gene product expression but not as effectively as the parent compounds. The fluorine-containing analogues appear to be somewhat more toxic in vitro to the antigen- and mitogen-driven proliferation of immunocompetent cells than their corresponding parent compounds.

Of the dideoxynucleosides described to date, the purine analogues ddA and ddI have exhibited very favorable therapeutic ratios in vitro. ddI is presently undergoing extensive phase I-II clinical trials. Whereas the action of adenosine deaminase (ADA) and purine nucleoside phosphorylase (PNP) is usually to convert a given analogue of Ado to an inactive or less active form, ddI appears to retain the same biological activity as that of the parent ddA. An explanation for these observations was possible when we found that ddI (1) underwent only a slow cleavage to hypoxanthine through the action of PNP and (2) accumulated the same active antiviral metabolite (i.e., ddATP) as ddA in human lymphoid cells. The use of human lymphoid cells with deficiencies in cellular nucleoside kinases and of inhibitors of pathways of nucleotide metabolism have also revealed new aspects of dideoxypurine metabolism in human lymphoid cells, including the identification of a salvage pathway (phosphotransferase/5'-nucleotide pathway) by which ddA/ddI may be metabolized preferentially to the active nucleotide. The effectiveness of ddA and ddI as orally administered antiviral agents may be limited by their susceptibility to acid hydrolysis and the low efficiency for nucleotide conversion in human lymphoid cells. The presence of a fluorine atom in the arabinose configuration on C-2 confers resistance to solvolysis and renders the analogue less susceptible to enzymatic deamination and resistant to phosphorylytic cleavage by PNP. In addition, human lymphoid cells accumulated several fold higher levels of the putative active triphosphate, 2'-F-dd-ara-ATP, than those of ddA or ddI. This increased accumulation of the analogue triphosphate could be accounted for by a more direct conversion of 2'-F-dd-ara-A by a direct phosphorylation through dCyd kinase than ddA. Thus, a single substitution with fluorine at the 2' "up" position of the sugar moiety of ddA markedly improves several biochemical properties relating to dideoxynucleotide accumulation in human lymphoid cells. Whether there are significant alterations of other biochemical properties, such as the ability of the analogue triphosphate to interact with the target enzyme reverse transcriptase, has not yet been determined. Thus, a definitive resolution of the relative merit of ddA/ddI and its 2'-fluoro-arabinosyl analogue is not yet possible on the basis of the studies described here.

The structural requirements for murine P-388 leukemia activity in the dopamine series were investigated. Eight of the 31 analogs evaluated possessed reproducible antitumor activity. The ortho-aromatic hydroxyl groups were required for activity, but the aminoethyl side chain was not crucial since this group could be replaced by methyl or aminomethyl groups with the retention of activity. The lack of activity of O-alkylated and monosubstituted analogs suggests that o-quinone formation may be important for activity, a possibility supported by the observed P-388 activity of 5-hydroxydopamine compared with the inactivity of the 6-hydroxy isomer.

The ribo- and arabinofuranosyl nucleosides of antitumor active 2- and 4-pyridones 1a and 2a were prepared by direct condensation of the silylated bases with either 1-O-acetyl-2,3,5-tri-O-benzoyl-D-ribofuranose (4a) or 2,3,5-tri-O-benzyl-1-p-nitrobenzoyl-D-arabinofuranose (7) in the presence of trimethylsilyl triflate (Me3SiOTf). In the case of the arabinofuranosyl nucleosides, separation of the alpha and beta anomers was accomplished at the stage of O-benzyl-protected compounds (8b + 9b, and 10b + 11b) after chemical functionalization of the 3-hydroxy group of the pyridone aglycons with acetyl and benzyl groups, respectively. Deblocking of the protected ribo- and arabinofuranosyl nucleosides was performed by the standard methods. In vitro activity against P-388 cells in culture indicated that the 4-pyridone riboside 6d was the most active member of the series with a twofold lower ID50 than the parent pyridone 2a. However, this and all the other compounds tested in this series showed no activity against the in vivo model system of murine P-388 leukemia at doses ranging from 25 to 400 mg/kg qd 1-5.

The discovery and development of anticancer drugs with clinical potential are the responsibility of the Developmental Therapeutics Program (DTP), Division of Cancer Treatment, National Cancer Institute (NCI). Approximately 10,000 compounds/year are selectively acquired and screened against murine tumor models in order to discover new, active materials. The program required to accomplish this objective, as well as the subsequent tasks of formulation development and toxicology testing, is described. Since its inception in 1955, the preclinical new drug research program of the NCI has played a major role in the discovery and development of new agents which have been entered into clinical trial. The NCI has been responsible for the discovery of eight of the 16 commercially available drugs discovered since 1955. In addition, the NCI has played an important role in the clinical evaluation of all 16 of these New Drug Application (NDA)-approved drugs. During 1977-1982, the NCI filed Investigational New Drug Applications (INDA) for 33 cytotoxic agents. It was responsible for the discovery of the antitumor activity of 73% of these compounds. Most of the INDA compounds were acquired directly through NCI efforts. The DTP active acquisition program was responsible for obtaining 69% of these materials, with an additional 12% coming from the DTP intramural research program. Only 19% were received as voluntary submissions. The DTP active acquisition and screening effort is shown to have played even a larger role in identifying and obtaining those compounds which are currently in earlier stages of the NCI drug discovery and development process.

With the objective of developing new antitumor agents, two groups of hydrazine compounds, having structural features in common with the antitumor agents procarbazine and 1-acetyl-2-picolinoylhydrazine, were synthesized. The L-1210 leukemia system was used to evaluate compounds of both groups. The aliphatic procarbazines also were screened for antitumor activity as bis(benzyloxycarbonyl) derivatives and as derivatives having a phthalazine nucleus. No L-1210 antitumor activity was exhibited by these compounds.

5-Azacytosine arabinoside (ara-AC) can be considered a combination of structural elements derived from the antitumor nucleosides cytosine arabinoside (ara-C) and 5-azacytidine (5-AC). The synthesis of ara-AC, for which standard methods were inadequate, was accomplished using the stable dihydro derivative as a synthetic intermediate. A novel dehydrogenation of the latter through the application of a trimethylsilylation-oxidation procedure gave ara-AC in good yield. Using murine L1210 leukemia as a test system, ara-AC was evaluated for antitumor properties in parallel determinations with 5-AC and ara-C. Although higher dose levels were necessary, ara-AC demonstrated a reproducibly greater efficacy in the L1210 system (% ILS = 144-148) than that shown by 5-AC (% ILS = 126-124) or ara-C (% ILS=127-121 ). Moreover, initial data suggest that ara-AC exhibits less host toxicity than either 5-AC or ARA-C. Although ara-AC can equally be considered an analogue of either 5-AC or ara-C, preliminary results indicate that ara-AC is chemically similar to 5-AC but biologically more closely related to ara-C.

Arabinofuranosyl-5-azacytosine (ara-AC), a nucleoside combining the structural elements of 5-azacytidine and arabinofuranosylcytosine, exhibited unusually wide therapeutic activity against several murine leukemias and all three human xenografts of the National Cancer Institute tumor panel. Activity was observed following either a daily or an intermittent regimen of treatment in the i.p. L1210 model. However, when multiple doses were administered on each treatment day, a greater therapeutic effect was produced and the total dose was reduced. Extensive necrosis was observed by light and electron microscopy in P388 tumors treated with ara-AC. Following s.c. administration, ara-AC caused regression of the mammary and lung xenografts (MX-1 and LX-1) and a 93% inhibition of the human colon tumor (CX-1); other analogues of this drug failed to demonstrate a comparably broad spectrum of activity. Morphological assessment of treated xenografts revealed a general loss of cell-to-cell contact and abundant necrosis 24 h after the administration of ara-AC. In culture, the 50% inhibitory concentrations of ara-AC for P388 and L1210 cells at 24 h were 1.9 and 4.5 microM, respectively, and the decline in replication rates was dependent on drug concentration. The cytocidal nature of the drug was demonstrated by cloning experiments in which it was observed that ara-AC abolished the clonogenicity of lymphoblasts but was only minimally cytotoxic to normal murine bone marrow progenitor cells. As adjudged by flow cytometry, the drug induced a distinct slowing of cell cycle traverse through S phase. Induction of the differentiation of HL-60 cells in culture was another cytotropic effect of this drug. At 44% differentiation (10 microM ara-AC), 50% of the cultured cells were viable. Its broad spectrum antitumor activity, its selective toxicity to tumor cells, and its ability to produce cytodifferentiation render ara-AC of interest as a potential antineoplastic agent in humans.

Tetrahydrouridine (THU, 2) and other fully reduced cyclic urea ribofuranosyl nucleosides undergo a rapid, acid-catalyzed isomerization to their more stable ribopyranosyl form. This isomerization is characterized by a change in spectral properties and by a greater than 10-fold decrease in potency for those nucleosides that act as potent inhibitors of cytidine deaminase in their ribofuranose form. 1-(beta-D-Ribopyranosyl)hexahydropyrimidin-2-one (7) was synthesized and used in conjunction with its furanose isomer 6 as a model compound for more extensive 1H and 13C NMR, mass spectral, and kinetic studies of this isomerization. The 0.4 delta upfield shift and 4-Hz increase in the J1',2' coupling constant for the pyranose anomeric proton in the 1H NMR spectrum is indicative of a pyranose beta-CI conformation in which the aglycon and C-2' and C-4' hydroxyls are equatorial. The mass spectra of trimethylsilylated pyranose nucleosides also show a characteristic large shift in the m/z 204-217 abundance and the appearance of two new rearrangement ions at M-133 and M-206. For furanose 6 the rate of isomerization is pH and temperature dependent with pyranose 7 predominating by a factor of 6-9 equilibrium. At pH 1 and 37 degrees C, furanose 6 has an initial half-life of less than 12 min. Accordingly, this isomerization may explain the observed lack of enhanced ara-C levels in studies evaluating the oral administration of an ara-C and THU combination to species with an acidic stomach content.

Spiromustine is a hydantoin-containing nitrogen mustard currently in Phase I clinical trial. Since the in vitro plasma half-life of this compound (6.4 min, 37 degrees C, pH 7.4) appeared to be influenced by the hydantoin ring, analogues containing 2-5 methylene spacer groups between this ring and the nitrogen mustard moiety were prepared and evaluated for hydrolytic stability and antitumor activity. Stability correlated with structure and pKa values. The proximity of the hydantoin ring to the mustard function was a stabilizing factor. Activity against murine P-388 leukemia was demonstrated and a gradual decrease in this activity was observed as the hydrolytic instability increased. A relationship between analogue structure and a mass spectral rearrangement ion was identified.

Fluphenazine was found to possess moderate reproducible activity against the intraperitoneal L-1210 and P-388 leukemia murine tumor models. Seven ether derivatives of fluphenazine and eight compounds in which the terminal side-chain hydroxyl group was replaced by an amine function were prepared and evaluated in the intraperitoneal L-1210, P-388, and B16 melanoma systems as well as the intracerebral L-1210 and ependymoblastoma brain tumor models. While no substantial intracerebral activity was observed, seven derivatives possessed reproducible activity in the intraperitoneal L-1210 or P-388 system. Several gave T/C values of 150%. No B16 melanoma activity was observed. These compounds were also tested for their cytotoxic properties in culture against L-1210, P-388, and KB cells. The amine isosteres, while possessing little in vivo activity, were the most cytotoxic of the compounds prepared, with several having ED50 values less than 1 microgram/ml.

Arabinosyl-5-azacytosine (ara-AC) is a new compound which combines the structural characteristics of arabinosyl cytosine (ara-C) and 5-azacytidine (5-AC). These three compounds, injected intraperitoneally, were evaluated in direct comparison against the intracerebral L1210 leukemia model. 5-AC was active in a non-schedule dependent manner producing increase in life span (ILS) values of 70%. The effects of both ara-C and ara-AC were schedule-dependent with the best activity (ILS ca. 600%; multiple long term survivors) observed using an around-the-clock treatment schedule. In all experiments, ara-AC appeared to be more efficacious than ara-C. In some instances, ara-AC was as active as the positive control compound, BCNU. Excellent activity for ara-AC was also observed against the intracerebral P388 leukemia system.

Compounds with known psychotropic properties were tested for activity in murine ip L1210 leukemia and B 16 melanoma in a protocol designed to obtain leads for new antitumor agents which might also possess central nervous system (CNS) antitumor properties. Barbiturates and hallucinogenic compounds were the only compound types deliberately excluded. Representatives from most of the other known CNS agent classes were included among the 297 psychotropic drugs evaluated. Sixteen of these agents were reproducibly active against the L1210 tumor system with T/C values of 125%. Phenothiazines such as fluphenazine and butyrophenones such as triperidol were prominent among the confirmed active structural types. Dopamine, a beta-phenethylamine neurotrasmitter, was active. While reproducible B16 melanoma activity was not observed among the psychotropic drugs, most of the L1210 confirmed active agents were effective against the ip P388 tumor model and also were active in vitro against KB cells. Ic L1210 activity was not observed among the few compounds chosen for testing in that tumor system. The yield of ip L1210 confirmed actives from this group of psychotropic agents was 18 times that which would have been expected from the random screening of compounds.

A nitrogen mustard analog of propranolol was synthesized as a potential lung-specific antitumor agent. Since dl-propranolol concentrates in lung tissue and beta-blocking activity resides only with the l-enantiomer, the d-modification could serve as a lung-directed carrier for a cytotoxic group. Reaction of 1-(1-naphthyloxy)-3-[bis(2-hydroxyethyl)amino]-2-propanol with thionyl chloride resulted in replacement of all three hydroxyl groups with chlorine. The necessary chlorination selectivity was achieved with p-toluenesulfonyl chloride in dimethylformamide solution to provide propranolol mustard, 1-(1-naphthyloxy)-3-[bis(2-chloroethyl)amino]-2-propanol. Both the trichloro compound and propranolol mustard showed reproducible activity against P-388 leukemia. Neither compound was active against the B16 tumor or Lewis lung carcinoma.

Pyridone structural requirements for activity against murine P-388 leukemia have been extended to isosteric analogs of 3-hydroxy-4-pyridone, a compound previously found to have activity. An amino group can be substituted for the 3-hydroxyl function with retention of activity. A sulfur, but not an amino function, can replace the lactam oxygen in the 2-position. Relocation of the lactam oxygen from the 2- to the 4-position in the pyridine ring also produces active pyridones, including 2-methyl-3-acetoxy-4-pyridone. This compound, which has a T/C value of 179%, is the most active material discovered thus far in the pyridone studies.

Agents were designed to exploit the high tyrosinase activity in melanotic melanoma relative to normal tissues. If specific tyrosinase activation of these agents occurred, the production of toxic metabolites in the melanoma cells would produce selective cell kill. Synthesis and antitumor activities of three new amino acids, 1a [beta-[(p-hydroxyphenyl)amino]alanine hydrochloride], 1b [N delta-(p-hydroxyphenyl)ornithine hydrochloride], and 1c [N delta-(m-hydroxyphenyl)ornithine dihydrochloride], were described. Compounds 1a and 1b were approximately 2-fold more active against the B-16 melanotic melanoma than the amelanotic melanoma cell line in vitro. Compound 1b was approximately 2-fold more potent than compound 1a against either cll line and was 8-fold more potent than L-glutamic acid gamma-(4-hydroxyanilide), a natural product isolated from mushroom. No significant growth inhibitory activity was found for the m-hydroxy analogue 1c at 100 micrometers, the highest concentration tested. Similarly, compound 1b exhibited better activity against P-388 (ED50 = 9.5 x 10(-6) M) than 1a (ED50 = 3.2 x 10(-5) M) and was about 30-fold more potent than 1c. Against human epidermoid carcinoma of the nasopharynx (KB), these agents showed modest inhibitory activity with ED50 values in the range of 1.2 to 3 x 10(-4) M. No in vivo activity against P-388 and B-16 at doses up to 150-200 mg/kg was observed. The biological results suggest that a nonspecific oxidation rather than a specific tyrosinase activation is involved in the biological action of these new compounds.

A series of 15 2,5-diaziridinyl-3,6-bis(alkylamino)-1,4-benzoquinone derivatives was synthesized and evaluated as central nervous system antitumor agents in the murine intracerebral L1210 and ependymoblastoma brain tumor systems. Intraperitoneal activity was evaluted in the leukemia L1210, P388, and B16 melanocarcinoma tumor models. The more hydrophilic hydroxyalkylamino compounds were the most effective in the intraperitoneal ascites systems (L1210, P388) with the dihydroxypropylamino (18) and hydroxyethylamino (17) analogues producing long-term survivors. The simple, more lipophilic mono- and dialkylamino derivatives were most effective in the intracerebral systems. Multiple long-term survivors were obtained with the methyl (13), ethyl (14), and dimethylamino (20) compounds in the ependymoblastoma brain tumor system. The parent amino analogue 12 was very active in several tumor models. The relationship between structure, activity, and water solubility are discussed.

While pure methyl 5-(2-chloroethylamino)-5-deoxy-2,3-O-isopropylidene-beta-D-ribofuranoside hydrochloride has no L-1210 leukemia activity, a decomposed sample was found to be very active. One of several approaches taken to determine the nature of the active component involved a study of how sugar structure affects antitumor activity. A number of aminoribose derivatives were prepared and tested against the murine L-1210 and P-388 leukemia and the B-16 melanoma tumor systems. Compounds were tested as pure materials and as synthetically degraded mixtures. Both the beta-haloethyl group and a secondary amine were required for highest activity.

Thirty-one aziridinylbenzoquinones were compared against five murine tumor models in vivo. Two intracerebral (ependymoblastoma and L-1210 leukemia) and three intraperitoneal (P-388 and L-1210 leukemia and B16 melanoma) systems were utilized. Excellent activity was observed for many compounds. Multiple long-term survivors were produced in the ependymoblastoma, P-388, and intraperitoneal L-1210 systems. Diethyl 2,5-bis(1-aziridinyl)-3,6-dioxo-1,4-cyclohexadiene-1,4-dicarbamate demonstrated superior activity in all five test systems. This compound also was reproducibly active against two colon tumors, a mammary tumor, and the intracerebrally implanted P-388 leukemia model.

Five beta-D-ribofuranosyl cyclic urea nucleosides (14-18), ranging in ring size from five to eight membered, were synthesized and evaluated as cytidine deaminase (CDA) inhibitors. The precursor protected nucleosides (9-13) were prepared by a condensation procedure utilizing persilylated ureas with a halo sugar under the specific catalytic activity of a HgO/HgBr2 mixture which provided exclusively the beta-anomers. Catalytic hydrogenation of known 1-(2,3,5-tri-O-benzoyl-beta-ribofuranosyl)-1,2-dihydropyrimidin-2-one (19) afforded nucleoside 10 identical with that obtained by the mercury-catalyzed condensation procedure. CDA activity varies significantly with the ring size of the urea aglycon the reaches its maximum level for the seven-membered analogues 16 and 17. The unexpected high potency of nucleoside 17 (Ki = 2.5 X 10(-8) M, human liver enzyme) is reported. This compound represents the most potent inhibitor of human liver CDA yet discovered.