NDP kinase reactivity towards 3TC nucleotides.

Nucleoside diphosphate (NDP) kinase is usually considered as the enzyme responsible for the last step of the cellular phosphorylation pathway leading to the synthesis of biologically active triphospho-derivatives of nucleoside analogs used in antiviral therapies and in particular in the treatment of AIDS. NDP kinase lacks specificity for the nucleobase and can use as substrate both ribo- or 2'-deoxyribonucleotides. However, only nucleoside analogs with a sugar moiety in the D-configuration (e.g. 3'-deoxy-3'-azidothymidine (AZT), 2',3'-didehydro-2',3'-dideoxythymidine (d4T)) have so far been analyzed as substrates of NDP kinase. In contrast, beta-L-2',3'-dideoxy-3'-thiacytidine (3TC), also called lamivudine, is a nucleoside analog that is now widely used in AIDS therapy and has a sugar moiety in the L-configuration. Using protein fluorescence to monitor the phosphotransfer between the enzyme and the nucleotide derivative at the presteady state, we have studied the reactivity of 3TC triphosphate and of other L-dideoxynucleotides with NDP kinase. We found that L-dideoxynucleoside triphosphates have a poor affinity for NDP kinase and that the catalytic efficiency of the phosphorylation of L-dideoxyderivatives is very low as compared with their D-enantiomers. We discuss these results using a computer model of 3TC diphosphate bound to the NDP kinase active site. NDP kinase may not seem to be the major enzyme phosphorylating 3TC-DP, in contrast to current opinion.

Evaluation and biological properties of reactive ligands for the mapping of the glycine site on the N-methyl-D-aspartate (NMDA) receptor.

The glycine-binding site of the N-methyl-D-aspartate (NMDA) receptor, given its potential as pharmacological target, has been thoroughly studied by structure-activity relationships, which has made possible its description in terms of spatial limits and interactions of various types. A structural model, based on mutational analysis and sequence alignements, has been proposed. Yet, the amino acid residues responsible for the interactions with the ligand have not been unambiguously characterized. To evidence nucleophilic pocket-lining residues, we have designed and synthesized reactive glycine-site ligands derived from 3-substituted 4-hydroxy-quinolin-2(1H)-ones by introducing various electrophilic groups at different positions of the molecule. These ligands were found to have high affinity at the glycine site and to be functional antagonists by inhibiting glycine/glutamate-induced currents in transfected oocytes. The correlation between their potency and their substitution pattern was strictly consistent with previously established structure-activity relationships. Most ligands displayed intrinsic reactivity toward cysteine, but none inactivated wild-type receptors. This is consistent with the model since it indicates the absence of exposed cysteine in the glycine-binding site. A strategy of cysteine incorporation by point mutations at selected polypeptide positions will create unambiguously localized targets for our reactive probes.

Reactive affinity probes for the mapping of the glycine-binding site of the NMDA receptor NR1 subunit.

The glycine co-agonist binding site of the NMDA receptor is a target for the prevention and treatment of neurotoxic and neurodegenerative conditions. Until now, the interactions taking place at this site, and its structure, have been investigated by ligand structure-activity relationships and by site-directed mutagenesis. On the basis of a structural model which is currently proposed for this site, we have designed and synthesized six affinity markers by substituting electrophilic reactive groups in the 4, the 7 and the 3' positions of L 701,324, a high-affinity glycine site antagonist. These compounds compete with 3H-DCKA binding to rat brain membranes at equilibrium with nanomolar to low-micromolar affinities, and antagonize glycine-evoked currents in oocytes transfected with wild-type NR1-NR2B. However, they do not induce a time-shift in binding equilibria, and do not inactivate irreversibly the glycine evoked currents. Since they react only with cysteine at physiological pH, we conclude that there is no such residue in the site, in agreement with the model. Our affinity markers therefore represent potential topological probes for NMDA receptors with sequence positions related to the glycine-binding site mutated into cysteine.

Suramin analogues with a 2-phenylbenzimidazole moiety as partial structure; potential anti HIV- and angiostatic drugs, 2: Sulfanilic acid-, benzenedisulfonic acid-, and naphthalenetrisulfonic acid analogues.

The synthesis of suramin analogues bearing a 2-phenyl-benzimidazole moiety is described. Aminoarene sulfonic acids 2a-e are acylated with 3,4-dinitrobenzoyl chloride 3 yielding the amides 4a-e which are hydrogenated to the corresponding diamines 5a-e. These are treated with 3-nitrobenzaldehyde, yielding the azomethines 7a-e and their isomers 8a-e and 9a-e. Key step in the synthesis of the target compounds 12a-e is the oxidation of the azomethines with oxygen to the benzimidazoles 10a-e. These are hydrogenated to the amines 11a-e reacting with phosgene to yield the symmetric ureas 12a-e. Results of the anti-HIV, cytostatic, and antiangiogenic screening are presented.

Gsalpha-selective G protein antagonists.

Suramin acts as a G protein inhibitor because it inhibits the rate-limiting step in activation of the Galpha subunit, i.e., the exchange of GDP for GTP. Here, we have searched for analogues that are selective for Gsalpha. Two compounds have been identified: NF449 (4,4',4",4'"-[carbonyl-bis[imino-5,1,3-benzenetriyl bis-(carbonylimino)]]tetrakis-(benzene-1,3-disulfonate) and NF503 (4, 4'-[carbonylbis[imino-3,1-phenylene-(2, 5-benzimidazolylene)carbonylimino]]bis-benzenesulfonate). These compounds (i) suppress the association rate of guanosine 5'-[gamma-thio]triphosphate ([35S]GTP[gammaS]) binding to Gsalpha-s but not to Gialpha-1, (ii) inhibit stimulation of adenylyl cyclase activity in S49 cyc- membranes (deficient in endogenous Gsalpha) by exogenously added Gsalpha-s, and (iii) block the coupling of beta-adrenergic receptors to Gs with half-maximum effects in the low micromolar range. In contrast to suramin, which is not selective, NF503 and NF449 disrupt the interaction of the A1-adenosine receptor with its cognate G proteins (Gi/Go) at concentrations that are >30-fold higher than those required for uncoupling of beta-adrenergic receptor/Gs tandems; similarly, the angiotensin II type-1 receptor (a prototypical Gq-coupled receptor) is barely affected by the compounds. Thus, NF503 and NF449 fulfill essential criteria for Gsalpha-selective antagonists. The observations demonstrate the feasibility of subtype-selective G protein inhibition.

Antiangiogenic and antiproliferative activity of suramin analogues.

The purpose of this study was to test the ability of 70 polyanionic analogues of suramin to inhibit angiogenesis. The ID50, the dose that produced 50% inhibition of angiogenesis, was determined for suramin and each of the analogues by measuring the ability of various amounts to inhibit angiogenesis in vivo in the chick egg chorioallantoic membrane (CAM) assay. Of the 70 analogues, 11 had antiangiogenic activities similar to suramin and an additional 7 were significantly more potent than suramin. All seven of these analogues were from the naphthalenetrisulfonic acid group and contained large urea groups. The benzene sulfonic and disulfonic acid analogues were less active inhibitors of angiogenesis than the naphthalenetrisulfonic acid analogues. Replacement of the naphthalenetrisulfonic acid groups by aliphatic carboxylic acids or benzoic acid gave analogues with very little antiangiogenic activity. In subsequent experiments, the antiproliferative activity of selected analogues on basic FGF (bFGF)-stimulated growth of immortalized human microvascular endothelial cells in vitro was determined. Analogues that inhibited angiogenesis to a greater extent than suramin in the CAM assay generally showed a greater antiproliferative effect on bFGF-induced growth of human microvascular endothelial cells. These results suggest that some of the polyanionic analogues may be potent therapeutic agents for cancers and angiogenesis-dependent diseases.

Alkylation-induced mono(ADP-ribosyl)-histones H1 and H2B. Hydroxylamine-resistant linkage in hepatoma cells.

Treatment of hepatoma AH 7974 cells with dimethyl sulfate led to a marked accumulation in vivo of mono)ADP-ribosyl)-histone H1A, H1B, H1 and H2B, respectively. In these conjugates, most of the modifying groups were linked to the acceptor proteins by an 'unusual' bond not described so far for ADP-ribosyl histone conjugates. It resisted treatment with 3M hydroxylamine, 0.1M picrylsulfonate and mild alkali, which excluded a linkage through carboxyl or guanidino residues. The stability of these conjugates formed endogenously differed also from 'non-enzymic' histone H1 conjugates formed by incubation of free ADP-ribose with the histone. Histone-linked mono(ADP-ribosyl) residues synthesized in hepatoma cells in response to alkylation were located exclusively in the domains that interact with DNA, i.e. in the non-globular C-terminal tail of histone H1 and in the N-terminus of histone H2B. Besides poly(ADP-ribosyl)ation, the modification of histones by single ADP-ribose groups may represent an independent process to modulate DNA/histone interaction.

DNA repair-associated ADP-ribosylation in vivo. Modification of histone H1 differs from that of the principal acceptor proteins.

ADP-ribosylation in vivo of histone H1 was studied in hepatoma cells (Yoshida AH 7974) after treatment with the alkylating agent dimethyl sulfate for 30 min and compared with that of other polypeptides. In unstimulated cells, histone H1 was only a minor acceptor (less than 4%) of total monomeric and polymeric ADP-ribosyl residues. Induction of DNA repair by dimethyl sulfate treatment increased total mono(ADP-ribosyl) protein conjugates 1.6-fold whereas histone H1-linked mono(ADP-ribosyl) groups were elevated greater than 30-fold, thus accounting for nearly one-fourth of the net increase in monomeric ADP-ribosyl residues. In contrast, histone H1-associated poly(ADP-ribosyl) residues comprised only 2% of the total increase in poly(ADP-ribose). The extent to which the histone H1 population became ADP-ribosylated was low even in dimethyl sulfate-treated cells. Less than 2% of the histone H1 molecules were mono(ADP-ribosyl)ated and only 0.003% carried poly(ADP-ribosyl) chains when an average chain length of 10 is assumed. The principal polypeptide acceptors of alkylation-induced ADP-ribosylation were concentrated in two peaks, one migrating close to the position of core histones H3/H2B and accepting most of the induced mono(ADP-ribosyl) and poly(ADP-ribosyl) residues. The other (Mr = 110,000-160,000) resembled auto-modified poly(ADP-ribose) polymerase. Our data demonstrate marked differences of alkylation-induced (ADP-ribosyl)n protein patterns to analyses performed in vitro.