Synthesis of chiral carbocyclic ribonucleotides.

The carbocyclic analogs of CMP, UMP, GMP, IMP, and ribo-TMP, of the same absolute configuration as the naturally occurring beta-D-ribofuranose-based ribonucleoside monophosphates, have been synthesized. The synthetic route employed Mitsunobu coupling of the heterocycles, appropriately protected where necessary, with a differentially protected, chiral carbocyclic core.

The human trifunctional enzyme of de novo purine biosynthesis: heterologous expression, purification, and preliminary characterization.

The cDNA for the human trifunctional enzyme of de novo purine biosynthesis, which encodes glycinamide ribonucleotide synthetase, aminoimidazole ribonucleotide synthetase, and glycinamide ribonucleotide trans-formylase, has been overexpressed in Escherichia coli and its protein product has been purified to homogeneity. The glycinamide ribonucleotide transformylase activity, which constitutes the C-terminal domain of the trifunctional enzyme, has been characterized with respect to its kinetic constants, Vmax = 3.03 +/- 0.15 micromol/min-mg and Km values for beta-glycinamide ribonucleotide and 10-formyl-5,8-dideazafolate of 0.94 +/- 0.21 and 1.58 +/- 0.25 microM, respectively, and its kinetic mechanism, which is ordered-sequential with the folate substrate binding first. The correspondence of these data to those obtained for the glycinamide ribonucleotide transformylase activity of the mammalian trifunctional enzyme indicates that the recombinant enzyme is fully functional.

The human glycinamide ribonucleotide transformylase domain: purification, characterization, and kinetic mechanism.

Glycinamide ribonucleotide transformylase catalyzes the third reaction of de novo purine biosynthesis, namely, the conversion of glycinamide ribonucleotide to N-formylglycinamide ribonucleotide, with concomitant conversion of 10-formyltetrahydrofolate to tetrahydrofolate. This activity has been shown to be a target for cancer chemotherapy, which has generated renewed interest in both the enzyme and the pathway. Moreover, in higher eukaryotes this activity constitutes the C-terminal domain of a monomeric protein which also catalyzes two additional reactions of de novo purine biosynthesis. In this study, the human glycinamide ribonucleotide transformylase domain has been expressed to high levels in Escherichia coli and purified to homogeneity. Our improved expression-purification system produces the desired activity exclusively in a soluble form and in higher abundance than previously achieved. The kinetic constants have been determined and the kinetic mechanism has been established as ordered-sequential, with the folate substrate binding first. The correspondence of these data to those obtained for the glycinamide ribonucleotide transformylase activity of the mammalian trifunctional enzyme indicates that the recombinant enzyme is fully functional.

Substrate specificity of glycinamide ribonucleotide synthetase from chicken liver.

Several analogs of glycinamide ribonucleotide and phosphoribosylamine have been prepared and evaluated as substrates for glycinamide ribonucleotide synthetase purified from chicken liver. Glycinamide ribonucleotide analogs include side chain modifications wherein the glycine side chain (R = CH2NH2) has been replaced by R = CH2NHCH3 and R = CH2CH2NH2, ribose ring replacement by chiral cyclopentane and cyclopentene derivatives, and phosphate replacement by phosphonates. All of these, with the exception of the O-phosphonate, served as substrates for the reverse enzymatic reaction, with Vmax values comparable to that obtained with glycinamide ribonucleotide, although the Km values ranged from 21 to 118 times the Km for glycinamide ribonucleotide. Analogs of phosphoribosylamine examined as substrates for the forward reaction consist of chiral derivatives of cyclopentane and cyclopentene and a chiral carbocyclic phosphonate. These also served as substrates, with Km values ranging from 5 to 23 times the Km for phosphoribosylamine and with diminished Vmax values. These studies have begun to define the structural features of the nucleotide substrate necessary to support enzymatic activity. Sarcosine (N-methylglycine) and beta-alanine were also accepted as substrates, albeit with reduced affinity compared with glycine.

Substrate specificity of glycinamide ribonucleotide transformylase from chicken liver.

Several glycinamide ribonucleotide analogs have been prepared and evaluated as substrates and/or inhibitors of glycinamide ribonucleotide transformylase from chicken liver. The side chain modified analogs, in which the glycine side chain, R = CH2NH2, has been replaced by R = CH2NHCH3 and R = CH2CH2NH2, are substrates, with V/K (relative intensity) of 2.4% and 16.3%, respectively. Several carbocyclic analogs of glycinamide ribonucleotide, including the phosphonate derivative of carbocyclic glycinamide ribonucleotide, did not serve as substrates, but were inhibitors of the enzyme, competitive against glycinamide ribonucleotide, with Ki values ranging from 7.4 to 23.6 times the Km for glycinamide ribonucleotide. However, the O-phosphonate analog of carbocyclic glycinamide ribonucleotide did support enzymatic activity, with V/K (relative intensity) of 0.8%. In addition, glycinamide ribonucleoside was neither a substrate for, nor an inhibitor of, glycinamide ribonucleotide transformylase. Furthermore, alpha-glycinamide ribonucleotide had no effect on enzyme activity. These studies have begun to define the structural features of the nucleotide substrate required to support enzymatic activity.

Carbocyclic substrates for de novo purine biosynthesis. Enantiospecific synthesis and enantiospecificity of enzymatic utilization.

The carbocyclic analogues of phosphoribosylamine, glycinamide ribonucleotide, and formylglycinamide ribonucleotide have been prepared enantiospecifically from D-ribonic acid gamma-lactone. These carbocycles, which have the same absolute configuration as the natural D-ribose-derived intermediates of de novo purine biosynthesis, are utilized stoichiometrically by the initial enzymes of the pathway. A comparison of the enzymatic processing of the (-)-enantiomers with those of the racemates indicates that in some cases, the (+)-enantiomer acts to inhibit the enzymatic activity.

Exploring the hexokinase glucose binding site through correlation analysis and molecular modeling of glucosamine inhibitors.

A series of N-substituted glucosamines has been designed, synthesized, and tested as inhibitors of yeast hexokinase. All derivatives exhibited competitive inhibition kinetics with respect to glucose. Quantitative structure-activity relationships were derived from the resulting inhibition data. The most significant equation demonstrated the existence of highly specific steric effects for the seven meta-substituted benzoylglucosamines included in the relationship. Molecular modeling of potential complexes between the inhibitors and the hexokinase substrate binding site strongly suggests that the steric effects arise from potential contacts with two amino acid residues lying in the region occupied by the amide substituents.

Carbocyclic substrates for de novo purine biosynthesis.

The carbocyclic analogues of phosphoribosylamine, glycinamide ribonucleotide, and formylglycinamide ribonucleotide have been prepared as the racemates. Carbocyclic phosphoribosylamine was utilized as a substrate by the monofunctional glycinamide ribonucleotide synthetase from Escherichia coli as well as the glycinamide ribonucleotide synthetase activity of the eucaryotic trifunctional enzyme of de novo purine biosynthesis. Furthermore, carbocyclic glycinamide ribonucleotide was processed in the reverse reaction catalyzed by these enzymes. In addition, carbocyclic formylglycinamide ribonucleotide was converted, by E. coli formylglycinamide ribonucleotide synthetase, to carbocyclic formylglycinamidine ribonucleotide, which was accepted as a substrate by the aminoimidazole ribonucleotide synthetase activity of the trifunctional enzyme. This study has afforded carbocyclic substrate analogues, in particular for the chemically labile phosphoribosyl amine, for the initial steps of de novo purine biosynthesis.

Mammalian glycinamide ribonucleotide transformylase. Kinetic mechanism and associated de novo purine biosynthetic activities.

Glycinamide ribonucleotide transformylase catalyzes the conversion of glycinamide ribonucleotide and 10-formyltetrahydrofolate to formylglycinamide ribonucleotide and tetrahydrofolate. The enzyme purified from the murine lymphoma cell line L5178Y also catalyzes two other de novo purine biosynthetic activities, glycinamide ribonucleotide synthetase and aminoimidazole ribonucleotide synthetase. The transformylase reaction shows a 1:1 stoichiometry for substrate utilization and an optimum rate between pH 7.9 and 8.3. Initial velocity and dead-end inhibition patterns indicate that the kinetic mechanism of the transformylation reaction is ordered-sequential, with 10-formyltetrahydrofolate binding first. alpha, beta-Hydroxyacetamide ribonucleotide (alpha, beta-N-(hydroxyacetyl)-D-ribofuranosylamine) is shown to be an inhibitor of the transformylase, competitive against glycinamide ribonucleotide.

Carbocyclic glycinamide ribonucleotide is a substrate for glycinamide ribonucleotide transformylase.

The carbocyclic analog of glycinamide ribonucleotide has been synthesized from the racemic parent trihydroxy cyclopentyl amine (B.L. Kam and N.J. Oppenheimer (1981) J. Org. Chem. 46, 3268-3272). This analog was accepted as a substrate (Km = 18 microM, Vmax = 0.23 mM/min) by mammalian glycinamide ribonucleotide transformylase (EC 2.1.2.2) with an efficiency comparable to that of the natural substrate glycinamide ribonucleotide (Km = 10 microM, Vmax = 0.27 mM/min). For each molecule of 10-formyl-5,8-dideazafolate cosubstrate consumed, 0.92 molecule of N-formyl carbocyclic glycinamide ribonucleotide was produced in the enzymatic reaction, indicating a 1:1 stoichiometry. These studies afford the first alternate nucleotide substrate for glycinamide ribonucleotide transformylase and suggest that the ribose ring oxygen of glycinamide ribonucleotide is not critical for enzyme recognition and binding.

N10-substituted 5,8-dideazafolate inhibitors of glycinamide ribonucleotide transformylase.

A series of 5,8-dideazafolates bearing ethyl, isopropyl, cyclopropylmethyl, propargyl, 3-cyanopropyl, carboxymethyl, 2-carboxyethyl, phenacyl, 3-fluorobenzyl, and 5-uracilylmethyl substituents at N10 were tested as inhibitors of purified L5178Y glycinamide ribonucleotide transformylase (GAR TFase), which requires 10-formyltetrahydrofolate as cofactor. All of these cofactor analogues exhibited competitive inhibition against N10-formyl-5,8-dideazafolate, with Ki's ranging from 2 to 32 microM.

Synthesis of 10-acetyl-5,8-dideazafolic acid: a potent inhibitor of glycinamide ribonucleotide transformylase.

10-Acetyl-5,8-dideazafolic acid has been synthesized in good yield from the parent compound, 5,8-dideazafolic acid. This quinazoline folate analogue showed no activity as a substrate for the folate-requiring de novo purine biosynthetic enzyme glycinamide ribonucleotide transformylase isolated from the murine lymphoma cell line L5178Y, but proved to be a potent competitive inhibitor, Ki = 1.3 microM, of the purified enzyme.

Mammalian glycinamide ribonucleotide transformylase: purification and some properties.

Glycinamide ribonucleotide transformylase, the first of the two formyl group transferases of de novo purine biosynthesis requiring 10-formyltetrahydrofolate, has been purified 1500-fold, nearly to homogeneity, from the murine lymphoma cell line L5178Y. Purification of the enzyme was facilitated by the use of a gelatin protease "affinity" resin. This mammalian enzyme is a monomer of approximate Mr 110 000. The kinetic studies are consistent with a sequential reaction mechanism and yield Michaelis constants of 0.4 mM for the substrate, glycinamide ribonucleotide, and 0.25 microM for the cofactor analogue 10-formyl-5,8-dideazafolate. A minimum Vmax of 2 mumol/(min . mg) was obtained for the purified enzyme, from which a turnover number of 4 s-1 was calculated.

Synthesis of 5,11-methenyltetrahydrohomofolate and its antifolate activity in vitro.

The synthesis of 5,11-methenyltetrahydrohomofolate was accomplished by treatment of tetrahydrohomofolate (H4homofolate) with triethyl orthoformate in glacial acetic acid. This compound is a homofolate analogue of 5,10-methenyltetrahydrofolate which serves as one precursor to the 10-formyl one-carbon donor for the first transformylation in de novo purine biosynthesis, namely, the conversion of glycinamide ribonucleotide (GAR) to N-formylglycinamide ribonucleotide (FGAR), catalyzed by the enzyme glycinamide ribonucleotide transformylase (EC 2.1.2.2). The analogue proved to retard the rate of formation of formylglycinamide ribonucleotide apparently by inhibiting the rate of synthesis of 10-formyltetrahydrofolate, the actual cofactor for the transformylase enzyme, from 5,10-methenyltetrahydrofolate. Its inhibition of the enzyme, 5,10-methenyltetrahydrofolate cyclohydrolase (EC 3.5.4.9), was competitive against (+)-L-5,10-methenyltetrahydrofolate, with a Ki = 41 micro M. This derivative of homofolate may be responsible for its inhibition of purine biosynthesis in Sarcoma 180 cells.

Isotope-tapping experiments with rabbit liver fructose bisphosphatase.

Isotope-trapping experiments with mental-free rabbit liver fructose 1,6-bisphosphatase have shown that enzyme-bound D-fructose 1,6-bisphosphate completely dissociates prior to enzyme turnover initiated by Mn2+ as the catalytic metal. The exchange rate of the binary enzyme-D-fructose 1,6-bisphosphate complex with the substrate pool is, therefore, more rapid than its conversion to products, suggesting that structural Mn2+ is necessary for productive substarate binding. Rapid-quench isotope-trapping experiments confirm the requirement for structural Mn2+ ions for productive binding to occur. These experiments also show that an ordered formation of the enzyme-Mn2+ s-D-fructose 1,6-bisphosphate ternary complex which features metal-ion addition prior to substrate constitutes a catalytically competent pathway in the mechanism of fructose 1,6-bisphosphatase and that all four subunits are active in a single turnover event.

Binding and kinetic data for rabbit liver fructose-1,6-bisphosphatase with Zn2+ as cofactor.

Atomic absorption determinations of zinc content were employed to demonstrate the technique to obtain zinc-free rabbit liver fructose-1,6-bisphosphatase (D-fructose-1,6-bisphosphate 1-phosphohydrolase, EC 3.1.3.11). Reactivation of the apoenzyme by Zn(2+) is rapid (within 1 min) and restores up to 96% of the initial specific activity. Gel filtration measurements showed that the enzyme contains four binding sites for zinc per molecule, one per subunit. The dissociation constants for the initial two binding sites are less than 0.1 muM. In the presence of a substrate analog, (alpha + beta) methyl D-fructofuranoside 1,6-bisphosphate, at a level where two analog molecules are bound per phosphatase molecule, a total of eight Zn(2+) ions bind at 8 muM Zn(2+), revealing the presence of additional binding sites, including the catalytic one. The activity in the presence of Zn(2+) is maximal at ca. 8 muM Zn(2+), which corresponds to saturation of the four subunit sites plus the catalytic sites in the presence of substrate. At metal ion concentrations less than 10 muM, the order of activation is Zn(2+) > Mn(2+) > Mg(2+). In kinetic assays with two metal cofactors the effect of Zn(2+) at concentrations less than 10 muM on either the Mg(2+) or the Mn(2+) assays is inhibitory owing to the apparent formation of mixed (two different elements) metal ion-enzyme complexes possessing a catalytic activity that is measureable but lower than anticipated if the catalysis by the various metal ions is simply additive. Hence the activation by EDTA of the Mg(2+) and Mn(2+) assays is explicable in terms of Zn(2+) removal, thus eliminating mixed metal species. Collectively these observations suggest that fructose-1,6-bisphosphatase may function in vivo as a Zn(2+) metalloprotein.