6-(alkylamino)-9-alkylpurines. A new class of potential antipsychotic agents.

A series of 6-(alkylamino)-9-alkylpurines was synthesized and evaluated for the property of antagonizing the behavioral effects in animals of the dopamine agonist apomorphine. This model for identifying potential antipsychotic agents is based on the hypothesis that agents that antagonize apomorphine-induced aggressive behavior in rats and apomorphine-induced climbing in mice, but that do not block stereotyped behavior, could have an antipsychotic effect in humans without producing extrapyramidal side effects. The antiaggressive-behavior activity of lead compound 1 (6-(dimethylamino)-9-(3-phenylalaninamidobenzyl)-9H-purine) was improved 48-fold with 6-(cyclopropylamino)-9-(cyclopropylmethyl)-2-(trifluoromethyl)-9H- purine (80) (po ED50 of 2 mg/kg), which was obtained through an iterative sequence of structure--activity relationship studies that encompassed evaluation of the effects of structure variations at the purine 9-, 6-, and 2-positions. Potency was enhanced with a 9-cyclopropyl group, the duration of action was improved with the 6-(cyclopropylamino) substituent, potency was further enhanced with an N-formyl prodrug, and an agent with reduced cardiovascular effect emerged with the 2-trifluoromethyl purine 80. This potential antipsychotic agent was not developed further due to undesirable effects on the stomach.

Synthesis and anticonvulsant activity of N-benzylpyrrolo[2,3-d]-, -pyrazolo[3,4-d]-, and -triazolo[4,5-d]pyrimidines: imidazole ring-modified analogues of 9-(2-fluorobenzyl)-6-(methylamino)-9H-purine.

Analogues of 9-(2-fluorobenzyl)-6-(methylamino)-9H-purine (1) containing isosteric replacements of the imidazole ring atoms were synthesized and tested for anticonvulsant activity. The pyrrolo[2,3-d]-, pyrazolo[3,4-d]-, and triazolo[4,5-d]pyrimidines were less active than 1 against maximal electroshock-induced seizures (MES) in rats when given po. The differences in anti-MES activity for these analogues was not explained by differences in pKa or lipophilicity. However, the four classes of heterocycles have distinctly different calculated electrostatic isopotential maps, which may be related to optimum anticonvulsant activity.

1-(Fluorobenzyl)-4-amino-1H-1,2,3-triazolo[4,5-c]pyridines: synthesis and anticonvulsant activity.

A series of (fluorobenzyl)triazolo[4,5-c]pyridines was synthesized and tested for activity against maximal electroshock-induced seizures in rodents. The most promising compound, 14 (BW 534U87), which is a carbon-nitrogen isoster of a purine anticonvulsant, has a profile in rodents that suggests 14 will be free of emesis and useful in the treatment of seizure disorders for which phenytoin is presently indicated.

[[(Guaninylalkyl)phosphinico]methyl]phosphonic acids. Multisubstrate analogue inhibitors of human erythrocyte purine nucleoside phosphorylase.

A series of [[(guaninylalkyl)phosphinico]methyl]phosphonic acids, 2, was synthesized and tested as inhibitors of human erythrocyte purine nucleoside phosphorylase (PNPase). The target (phosphinicomethyl)phosphonic acids 2 were synthesized in six or seven steps from alkenylphosphonates 4. The latter were converted to the intermediate alkylmesylates 9 in a series of steps that included (1) conversion of the diethyl phosphonates 4 to the (phosphinoylmethyl)-phosphonates 7 and (2) conversion of the terminal double bond of [(alkenylphosphinoyl)methyl]-phosphonates 7 to the alkylmesylates 9. The pure 9-isomers 2 were obtained by alkylation of 2-amino-6-(2-methoxyethoxy)-9H-purine with alkylmesylates 9 followed by hydrolysis of the protecting groups with concentrated hydrochloric acid and ion exchange chromatography to give 2 as hydrated ammonium salts. The most potent inhibitor of human erythrocyte PNPase, [[[5-(2-amino-1,6-dihydro-6-oxo-9H-purin-9- yl)pentyl]phosphinico]methyl]phosphonic acid (2b), was a multisubstrate analogue inhibitor with a Ki' of 3.1 nM. Optimum PNPase inhibitory activity required the presence of zinc ions in the assay medium. These potent inhibitors of PNPase exhibited only weak activity against human leukemic T-cells in vitro.

9-[(Phosphonoalkyl)benzyl]guanines. Multisubstrate analogue inhibitors of human erythrocyte purine nucleoside phosphorylase.

A series of 9-[(phosphonoalkyl)benzyl]guanines was synthesized and tested for inhibition of human erythrocyte purine nucleoside phosphorylase (PNPase). Inhibitors of PNPase should be T-cell selective, immunosuppressive agents with potential clinical utility in the treatment of a wide variety of disorders in which T-lymphocytes are pathogenic. An initial set of six analogues of the weak PNPase inhibitor 9-benzylguanine (2) contained a phosphonic acid group linked to the ortho, meta, or para position of the aryl moiety via two- or three-atom spacers. These compounds allowed us to probe for a favorable interaction with the phosphate-binding domain. Several additional meta phosphonoalkyl substituents were examined in an effort to optimize the spacer. The two most potent compounds, [[3-[(2-amino-1,6-dihydro-6-oxo-9H-purin-9-yl)methyl] benzyl]oxy]-methylphosphonic acid (3f) and [[3-[(2-amino-1,6-dihydro-6-oxo-9H-purin-9-methyl)methyl] benzyl]-thio]methylphosphonic acid (3j), were inhibitors of PNPase with Ki's of 5.8 and 1.1 nM, respectively. These inhibitors displayed competitive kinetics with respect to inosine and inorganic phosphate, which showed that these compounds possess binding determinants for both the purine- and phosphate-binding domains of the enzyme, characteristics that are consistent with 3f and 3j being multisubstrate analogue inhibitors of PNPase. The potency of 9-benzylguanine (2) was enhanced more than 6000-fold by linking a phosphonic acid residue with a (methylthio)methyl spacer to the meta position of 2 to give 3j, which illustrates the potent enzyme inhibitory properties available to multisubstrate analogue inhibitors.

Benzodiazepine receptor binding activity of 9-(1-phenylethyl)purines.

Several alpha-methyl analogues of the 9-benzylpurines that bind to the benzodiazepine receptor (BZR) were synthesized and tested for BZR-binding activity. Although introduction of a m-amino group and an 8-bromo substituent gave an additive increase in BZR affinity with 9-(3-aminobenzyl)-8-bromo-6-(dimethylamino)-9H-purine (4), addition of an alpha-methyl group to 4 resulted in a loss in BZR affinity. This loss in affinity is apparently due to repulsive, steric interactions between the 8-bromo and 9-(1-phenylethyl) substituents, which results in a conformation that is not optimal for interaction with the BZR. Several compounds were tested on a modified Geller-Seifter conflict schedule, but none exhibited significant anxiolytic activity.

Synthesis and biological activity of an acyclic analogue of 5,6,7,8-tetrahydrofolic acid, N-[4-[[3-(2,4-diamino-1,6-dihydro-6-oxo-5- pyrimidinyl)propyl]amino]-benzoyl]-L-glutamic acid.

The synthesis and biological evaluation of N-[4-[[3-(2,4-diamino-1,6-dihydro-6-oxo-5-pyrimidinyl)propyl]amino]- benzoyl]-L-glutamic acid (1) (5-DACTHF, 543U76), an acyclic analogue of 5,6,7,8-tetrahydrofolic acid (THFA), are described. The key intermediate, hemiaminal 8, was prepared in four stages from 3-chloropropionaldehyde diethyl acetal. Reaction of 8 with dimethyl N-(4-aminobenzoyl)-L-glutamate gave the 2,4-bis(acetylamino) derivative 11, which was hydrolyzed with 1 N sodium hydroxide to give 1; the glycine analogue 16 was prepared in a similar manner. The N-methyl analogue 2 and N-formyl analogue 3 were prepared from 11 and 1, respectively. Compounds 1-3 inhibited growth of Detroit 98 and L cells in cell culture, with IC50s ranging from 2 to 0.018 microM. Cell culture toxicity reversal studies and enzyme inhibition tests showed that 1 was cytotoxic but not by the mechanism of the dihydrofolate reductase inhibitor aminopterin. Compound 1 and its polyglutamylated homologues inhibited glycinamide ribonucleotide transformylase (GAR-TFase) and aminoimidazole ribonucleotide transformylase (AICAR-TFase), the folate-dependent enzymes in de novo purine biosynthesis; and 1 was an effective substrate for mammalian folyl-polyglutamate synthetase. The compound inhibited (IC50 = 20 nM) the conversion of [14C]formate to [14C]-formylglycinamide ribonucleotide by MOLT-4 cells in culture. These data suggest that the site of action of 1 is inhibition of purine de novo biosynthesis. Moderate activity was observed against P388 leukemia in vivo.

Benzodiazepine receptor binding activity of 8-substituted-9-(3-substituted-benzyl)-6-(dimethylamino)-9H-purines.

A series of 8-substituted analogues of 9-(3-aminobenzyl)-6-(dimethylamino)-9H-purine (8) were synthesized and tested for their ability to bind to the benzodiazepine receptor (BZR) in rat brain tissue. The most active compound was the 8-bromo-9-(3-formamidobenzyl) analogue 16 (IC50 = 0.011 microM), which was 1000-fold more active than the parent 9-benzyl-6-(dimethylamino)-9H-purine (1) and nearly as active as diazepam. Although substitution of a m-formamido group and an 8-bromo substituent on 1 imparted potent BZR binding activity, neither 16 nor 11 analogues exhibited significant anxiolytic activity on a modified Geller-Seifter conflict schedule.

Benzodiazepine receptor binding activity of 6,9-disubstituted purines.

A series of 6,9-disubstituted purines were tested for their ability to bind to the benzodiazepine receptor in rat brain tissue. One of the most active compounds was 9-(3-aminobenzyl)-6-(dimethylamino)-9H-purine (44) with an IC50 = 0.9 microM, which was only 4.5-fold higher than the IC50 for chlordiazepoxide. Substitution of a 3-aminobenzyl or 3-hydroxybenzyl group at the 9-position of 6-(dimethylamino)purine led to over a 50-fold increase in receptor affinity. Compound 44 did not exhibit significant anxiolytic activity, nor did anticonvulsant activity correlate with relative receptor binding affinity.

9-(2-Fluorobenzyl)-6-(alkylamino)-9H-purines. A new class of anticonvulsant agents.

Several substituted aryl and 6-alkylamino analogues of the anticonvulsant purine 9-(2-fluorobenzyl)-6-(methyl-amino)-9H-purine (1) were synthesized and tested for anticonvulsant activity against maximal electroshock-induced seizures (MES) in rats. Derivatives with a second fluoro substituent in the 5- or 6-position of the aryl moiety were very active with ip ED50's that ranged from 2 to 4 mg/kg. Congeners in which the purine 6-substituent was varied among a number of alkylamino groups possessed potent activity against MES that was comparable to or several times better than phenytoin.

6-(Alkylamino)-9-benzyl-9H-purines. A new class of anticonvulsant agents.

Several 9-alkyl-6-substituted-purines were synthesized and tested for anticonvulsant activity against maximal electroshock-induced seizures (MES) in rats. Most compounds were prepared in three steps from 5-amino-4,6-dichloropyrimidine or in two steps via alkylation of 6-chloropurine. Potent anticonvulsant activity against MES resided in compounds that contain a benzyl substituent at the 9-position of 6-(methylamino)- or 6-(dimethyl-amino)purine. Among commonly used agents for control of seizures, this type of structure represents a new class of potent anticonvulsant agents.

Inhibition of histidine decarboxylase by imidazole derivatives.

A series of 4-imidazolylmethyl arylsulfides, sulfoxides, and sulfones and two carbon isosteres of 4-imidazolylmethylphenyl sulfide (I) were synthesized and tested for inhibition of histidine decarboxylase from rat stomach. None of these analogs of I met the criterion of a potent and specific inhibitor of histidine decarboxylase.