The GABA transporter and its inhibitors.

GABA is the major inhibitory neurotransmitter in the brain and GABA re- uptake from the synaptic cleft is one important mechanism in the regulation of GABA activity. Inhibition of the re-uptake of GABA by potent and selective inhibitors of the GABA transporter enhances GABA activity. This property can be used therapeutically in for instance epilepsy or psychiatric disorders. In this paper putative structures of the GABA transporter, its mechanism of action, the progress made in the search for its amino acids involved in the binding of substrates and the SAR of inhibitors acting at the transporter will be discussed. To date only highly selective GAT-1 inhibitors are available. These compounds are lipophilic derivatives of (R)-nipecotic acid and guvacine. The most potent inhibitors of the cloned human GAT-1 are NNC-711 (IC50 = 0.04 mM) and tiagabine (IC50 = 0.07 mM). A diheteroarylvinyloxy analogue of tiagabine, 5 times more potent than tiagabine, has been reported recently. For the GAT-2, GAT-3 and BGT-1 subtypes only compounds with a small preference for one of the subtypes have been published.

Serotonin, dopamine and norepinephrine transporters in the central nervous system and their inhibitors.

An overview is presented on progress made in the research on neuronal transporters of serotonin, dopamine and norepinephrine in the central nervous system. Tools developed by molecular biology, such as expression of cloned transporters, their mutants and chimera in non-neuronal cells offered the opportunity to study the putative domains for binding of substrates and uptake inhibitors and discover factors in the regulation of the transporter function. The study of the distribution of monoamine transporters in human brain became possible by the development of selective radiolabelled transport inhibitors. The relationships between the chemical structure of the uptake inhibitors and the affinity for the monoamine transporters is reported, and the (potential) therapeutic applications of the compounds are discussed.

5-HT(3) receptor antagonists and anxiety; a preclinical and clinical review.

The present paper reviews the evidence for anxiolytic activity of 5-HT(3) receptor antagonists in animal models of anxiety and in clinical trials in humans. Compared to the established anxiolytics (benzodiazepine receptor agonists and, to a lesser extent, 5-HT(1A) receptor agonists) 5-HT(3) receptor antagonists display a different anxiolytic profile. They are anxiolytic in a limited number of animal anxiety models. If active, they often are very potent and display bell-shaped dose response curves, whereas the ratio between therapeutic activity and side effects appears remarkably large. 5-HT(3) receptor antagonists remain active after chronic dosing and no indications for tolerance, dependence or rebound effects were found, which seems to make these drugs an attractive alternative to the benzodiazepines. However, the large body of animal data indicating a complete lack of psychotropic activity of 5-HT(3) receptor antagonists weakens the prediction of anxiolytic activity in these drugs. Human data are also controversial; some investigators have reported positive effects in anxiety disorders (panic disorder, GAD), others did not. It can be concluded that 5-HT(3) receptor antagonists do not represent a breakthrough in the treatment of various anxiety disorders, as initially suggested.

The 5-HT1A receptor and its ligands: structure and function.

An overview is presented on progress made in research on 5-HT1A receptors and their ligands since their discovery in 1983. Molecular biology has offered new tools, for example cloned 5-HT1A receptors, their mutants and chimeras to study structure and function. Many compounds, belonging to different chemical classes, display high affinity and selectivity for 5-HT1A receptors. The majority of these compounds are agonists or partial agonists, full antagonists are still scarce. Agonists and partial agonists are active in various animal models of anxiety and depression. Partial receptor agonists have been proven to be effective in general anxiety disorder and depression in man. Potential therapeutic applications for 5-HT1A receptor antagonists are evaluated, for example, in cognition disorders.

Functional characteristics of a series of N4-substituted 1-(2,3-dihydro-1,4-benzodioxin-5-yl)piperazines as 5-HT1A receptor ligands. Structure-activity relationships.

The agonistic/antagonistic profile of a series of 10 N4-substituted 1-(2,3-dihydro-1,4-benzodioxin-5-yl)piperazines is evaluated in the in vitro adenylyl cyclase assay. The profile is severely affected by the characteristics of the N4-substituents ranging from full agonism (benzamidoethyl derivative 1), mixed agonism/antagonism (phthalimidobutyl derivative 7) to predominantly antagonism (saccharinbutyl derivate 9). A novel full antagonist 10, as potent as WAY 100635, is obtained by substitution of Cl at C-7 of the benzodioxinyl moiety in 9.

Study of the interaction between aryloxypropanolamines and Asn386 in helix VII of the human 5-hydroxytryptamine1A receptor.

We studied the stereoselective interaction between aryloxypropanolamines and the human 5-hydroxytryptamine1A (5-HT1A) receptor. R- and S-enantiomers of propranolol, penbutolol, and alprenolol were investigated for their ability to bind to human 5-HT1A wild-type and Asn386Val mutant receptors. Asn386 seemed to act as a chiral discriminator. Although both aryloxypropanol enantiomers displayed lower affinity for the mutant receptors, the affinities for the S-enantiomers were more affected. Receptor affinities of other structurally unrelated 5-HT1A ligands were not decreased by the mutation of Asn386 to valine. In addition, a series of analogues of propranolol with structural variation in the oxypropanolamine moiety was synthesized, and affinities for wild-type and Asn386Val mutant 5-HT1A receptors were determined. Both the hydroxyl and the ether oxygen atoms of the oxypropanol moiety seem to be required for binding at wild-type 5-HT1A receptors. The hydroxyl group of propranolol probably directly interacts with Asn386. The ether oxygen atom may be important for steric reasons but can also be involved in a direct interaction with Asn386. These findings are in agreement with the interactions of aryloxypropanolamines with Asn386 in rat 5-HT1A receptors that we previously proposed. The loss of affinity for propranolol by the Asn386Val mutation could be regained by replacement of the hydroxyl group of the ligand by a methoxy group. This modification of the propranolol structure has no effect on the affinity of both enantiomers for the wild-type 5-HT1A receptor, which provides an alternative hypothesis for the interaction of Asn386 with the oxypropanol oxygen atoms. According to this novel hypothesis, the oxypropanol oxygen atoms may both act as hydrogen bond acceptors from the NH2 group of Asn386.

5-HT1A-versus D2-receptor selectivity of flesinoxan and analogous N4-substituted N1-arylpiperazines.

We investigated the structural requirements for high 5-HT1A affinity of the agonist flesinoxan and its selectivity versus D2 receptors. For this purpose a series of arylpiperazine congeners of flesinoxan were synthesized and evaluated for their ability to displace [3H]-8-OH-DPAT and [3H]spiperone from their specific binding sites in rat frontal cortex homogenates and rat striatum, respectively. Variations were made in the N4-substituent and the arylpiperazine region. Effects of N4-substitution in the investigated compounds appeared to be quite similar for 5-HT1A- and D2-receptor affinity. Lipophilicity at a distance of four carbon atoms from the piperazine N4 atom seems to be the main contributing factor to affinity for both receptors. Our data show that the amide group in the flesinoxan N4-substituent is unlikely to interact with the 5-HT1A receptor but, instead, acts as a spacer. In contrast to the structure-affinity relationships (SARs) of the N4-substituents, selectivity for 5-HT1A versus D2 receptors was gained by the arylpiperazine substitution pattern of flesinoxan. Restriction of flexibility of the N4-(benzoylamino)ethyl substituent and its effect on 5-HT1A-receptor affinity and activity were also studied. Our data show that in the bioactive conformation, the N4-[(p-fluorobenzoyl)amino]ethyl substituent is probably directed anti-periplanar relative to the HN4 atom. These results were used to dock flesinoxan (1) and two of its congeners (27 and 33) into a model of the 5-HT1A receptor that we previously reported. Amino acid residues surrounding the N4-[(p-fluorobenzoyl)amino]ethyl substituent of flesinoxan and its congeners are also present in D2 receptors. In contrast, several residues that contact the benzodioxane moiety differ from those in D2 receptors. These observations from the 3D model agree with the 5-HT1A SAR data and probably account for the selectivity of flesinoxan versus D2 receptors.

A series of N4-imidoethyl derivatives of 1-(2,3-dihydro-1,4-benzodioxin-5-yl)piperazine as 5-HT1A receptor ligands: synthesis and structure-affinity relationships.

A series of unsubstituted and substituted succinimido, maleimido, and glutarimidoethyl derivatives of eltoprazine (3) was synthesized and tested for affinity for the 5-HT1A receptor in rat brain homogenates. The unsubstituted compounds have a moderate affinity for the receptor, while the affinity considerably increases by substitution at or enlargement of these cyclic ring systems. A good correlation was found between the inhibition constant Ki (expressed as pKi) and the lipophilicity (clogP). No correlation was observed between the pKi or pKi+ (local inhibition constant) and the basicity of the N4-nitrogen atom.

N4-unsubstituted N1-arylpiperazines as high-affinity 5-HT1A receptor ligands.

In order to explore the structural requirements for high 5-HT1A affinity, a series of aryl-substituted N1-phenylpiperazines were synthesized and evaluated for their ability to displace [3H]-8-OH-DPAT from its specific binding sites in rat frontal cortex homogenates. We found 2-methoxy substitution to be favorable, while 4-methoxy substitution was detrimental for 5-HT1A affinity. Substitution with annelated rings at the 2,3-positions was highly favorable for all investigated compounds, with the exception of a pyrrole ring. All other substitutions, except fluoro, in this class of heterobicyclic phenylpiperazines decreased affinity in the order: ortho > para > meta. The loss of affinity in the ortho and para positions is probably due to steric factors: the substituents either cause steric hindrance with the receptor or prevent the compound from adopting the appropriate conformation for binding to the 5-HT1A receptor. Conformational analysis combined with structure-affinity relationships (SAR) indicates that our arylpiperazines may bind at the 5-HT1A receptor in a nearly coplanar conformation. Observed interactions of the compounds in our 5-HT1A receptor model appeared to be in agreement with SAR data. The aromatic part of the arylpiperazine moiety has pi-pi interactions with the aromatic residues Trp161 and Phe362 in helices IV and VI, respectively. The positively charged protonated basic nitrogen forms a hydrogen bond with the negatively charged Asp116 in helix III. The ammonium-aspartate complex is surrounded by aromatic residues Trp358 and Phe361 in helix VI. A lipophilic pocket is formed by Phe362, Leu366 (both helix VI), and the methyl group of Thr200 (helix V). In agreement with the model, addition of a methyl substituent to the structure of the benzodioxine analogue 12 in this region, yielding 13, is favorable for 5-HT1A receptor affinity. Unfavorable positions for substitution with bulky groups, like the 3- and 4-positions in the benzofuran compound 14, are explained by steric hindrance with the backbone atoms of helix V. Thus, we were able to rationalize the 5-HT1A SAR of existing N1-phenylpiperazines, as well as a series of newly synthesized bicyclic heteroarylpiperazines, in terms of receptor-ligand interactions. Several of these N4-unsubstituted compounds had affinities in the low-nanomolar range.

Structure-affinity relationship studies on 5-HT1A receptor ligands. 2. Heterobicyclic phenylpiperazines with N4-aralkyl substituents.

Structure-affinity relationship (SAR) studies for the 5-HT1A receptor site are presented for two series of heterobicyclic phenylpiperazines with N4-aralkyl substituents: 4-aralkyl derivatives of 1-(2,3-dihydro-1,4-benzodioxin-5-yl)piperazine (3) and 1-(benzo[b]furan-7-yl)piperazine (4). Their affinities for 5-HT1A receptors range from 0.15 to 28 nM and thus emphasize the importance of N4-substitution. By combining the SAR of these N4-aralkyl series with the recently published SAR of the N4-alkyl-substituted phenylpiperazines, the nature of the interaction of the N4-substituted phenylpiperazines and the 5-HT1A receptor was further examined using comparative molecular field analysis (CoMFA). To discriminate between two postulated hypotheses, CoMFA models were built and validated utilizing cross-validation, bootstrapping, and randomizing techniques. The model based on a N4-substituent alignment in which all N4-substituents are equally oriented in space was selected for further evaluation. According to the CoMFA/PLS analysis, the steric and electrostatic field properties contribute in a 98:2 ratio to the affinity found for the 5-HT1A receptor. Increasing steric bulk was found to be positively as well as negatively related to affinity depending on the distance of the bulk's center from the N4-nitrogen. The location of these steric CoMFA contour levels are well defined in space when the defined alignment rules are followed. Because CoMFA does not take hydrogen bonding into account, this could indicate that the contribution of the amide function (its ability to interact through hydrogen bonding), as present in the N4-substituents, to affinity is of minor importance.

A model of the serotonin 5-HT1A receptor: agonist and antagonist binding sites.

We built a model for the 5-HT1A receptor, using the 3D-structure of bacteriorhodopsin as a structural template. With the use of site-directed mutagenesis data, several potent 5-HT1A agonists, belonging to five different structural classes, and an aryloxypropanolamine antagonist, were docked into the receptor model. After docking, the surrounding of the ligands appeared to be in full agreement with previously reported SAR-data of 5-HT1A ligands. In this study, for the first time, an explanation for 5-HT SAR results is given in terms of interactions between ligands and amino acid residues. Also the selectivity of 8-OH-DPAT for the 5-HT1A receptor is accounted for. In our model the agonists and the antagonist interact with different residues on several helices. They all interact with the essential aspartic acid on helix III, that is known to bind all amines to receptors for biogenic amines. This partial overlap of the binding sites accounts for the antagonism of the class of aryloxypropanolamines and for the deviating SAR of this class of compounds when compared to agonists.

Development of high-affinity 5-HT3 receptor antagonists. Structure-affinity relationships of novel 1,7-annelated indole derivatives.

On the basis of the structures of ondansetron and GR 65,630, its ring-opened C-linked methylimidazole analogue, novel 1,7-annelated indole derivatives were synthesized as potential 5-HT3 antagonists. Receptor binding studies show that all compounds display a high affinity for the 5-HT3 receptors. In both series annelation results in compounds being 7 and 4 times more potent than the references ondansetron and GR 65,630, respectively. Similar to ondansetron, the 1,7-annelated indoles show little stereoselectivity. The (-)-isomers are only slightly more potent than the (+)-isomers. The receptor binding profile of l-10-[(2-methyl-1H-imidazol-1-yl)methyl]-5,6,8,9,10,11-hexahydro-4H-pyri do [3,2,1-jk]carbazol-11-one hydrochloride (24b) (INN cilansetron) shows that the compound displays, besides a high affinity for 5-HT3 receptors (Ki = 0.19 nM), a weak affinity for sigma-receptors (Ki = 340 nM), muscarine M1 receptors (Ki = 910 nM), and 5-HT4 receptors (Ki = 960 nM) and no affinity (Ki > or = 5000 nM) for all the other receptor types tested (n = 37). The new compounds fit the proposed necessary chemical template for binding: a heteroaromatic ring system, a coplanar carbonyl group, and a nitrogen center at well-defined distances. The enhanced potency of the annelated 1,7-indole derivatives indicates that the extra ring provides a favorable hydrophobic area for interaction with the 5-HT3 receptor site. In vivo cilansetron is more potent and induces less central side effects than ondansetron. At present cilansetron is in clinical trials.

Structure-affinity relationship studies on 5-HT1A receptor ligands. 1. Heterobicyclic phenylpiperazines with N4-alkyl substituents.

Structure-affinity relationship (SAR) studies for 5-HT1A receptor site are presented for two series of heterobicyclic phenylpiperazines with N4-alkyl substituents: 4-alkyl derivatives of 1-(2,3-dihydro-1,4-benzodioxin-5-yl)piperazine (3) and 1-(benzo[b]furan-7-yl)piperazine (4). The linear and branched hydrocarbon chain derivatives up to n-decyl were synthesized and evaluated for their ability to displace [3H]-2-(di-n-propylamino)-8-hydroxytetralin from its specific binding sites in rat frontal cortex homogenates. All compounds displayed a nanomolar affinity for the 5-HT1A receptor. In both series the N-ethyl and N-n-propyl substituted derivatives have similar affinities, being slightly but statistically significantly less active than the N-methyl-substituted derivatives. Elongation of the hydrocarbon chain increases the affinity for the central 5-HT1A receptor site, reaching a local maximum for the N-n-hexyl-substituted phenylpiperazines 23 (Ki = 0.50 nM) and 39 (Ki = 0.54 nM). Assuming that the arylpiperazine derivatives at the 5-HT1A binding site are in the ionic state, ionization constants were determined in order to evaluate the use of the local inhibition constant, Ki+, as a more convenient parameter to study the structure-affinity relationships. However, the Ki+ could not account for the specific N4-substituent effects found.

Hybrid cholecystokinin-A antagonists based on molecular modeling of lorglumide and L-364,718.

A series of novel nonpeptide cholecystokinin-A (CCK-A) antagonists have been synthesized. Designed on the basis of the structural homology between lorglumide and L-364,718, as investigated with molecular modeling, these compounds constitute a link between the N-acylglutamic acid and 3-amino-5-phenyl-1,4-benzodiazepin-2-one derived antagonists. The prepared compounds were tested in vitro as antagonists of the binding of [3H]-(+/-)-L-364,718 and [3H]-CCK-8(S) to rat pancreas and guinea pig brain membranes, respectively. All compounds proved to be selective for the (peripheral) CCK-A receptor, the most potent analogue, 6, having a Ki value of 90 nM. The structure-activity profile of the series of hybrid compounds relates closest to that of the N-acylglutamic acid derived antagonists.

1H-imidazo[4,5-c]quinolin-4-amines: novel non-xanthine adenosine antagonists.

On the basis of a model we recently developed for the antagonist binding site of the adenosine A1 receptor (J. Med. Chem. 1990, 33, 1708-1713), it was predicted that 1H-imidazo[4,5-c]quinolin-4-amines would be antagonists of the A1 receptor. Furthermore, it was expected that certain hydrophobic substitutions at the 2- and 4-positions would enhance affinity. Here, we report on the synthesis and the adenosine A1 and A2 receptor affinity of substituted 1H-imidazo[4,5-c]quinolin-4-amines. Some of these compounds have nanomolar affinity for the A1 receptor. The structure-activity relationships (SAR) of these compounds are discussed in relation to SAR for other adenosine receptor ligands. The 1H-imidazo[4,5-c]quinolin-4-amines constitute a novel class of non-xanthine adenosine antagonists.

The concept of selectivity in 5-HT receptor research.

Since the demonstration that serotonin (5-hydroxytryptamine, 5-HT) interacts with different (sub)types of membrane receptors, several compounds have been proposed as potent and selective ligands for one of these 5-HT subtypes. Unfortunately, specific and highly selective ligands (selectivity ratios greater than or equal to 1000) for the majority of 5-HT subtypes are still lacking. A few compounds are selective (ratios greater than or equal to 100), but most of the reputed 'selective' tools display affinities for other 5-HT subtypes and/or other (neuro-) transmitter receptors. Mainly due to different interpretations of the concept of selectivity, many of these nonselective compounds are still used to characterize 5-HT receptors. In this paper, we present the affinities (obtained by radioligand binding studies) of the most selective tools known today for each of the 5-HT subtypes and discuss the structure-activity relationships of some interesting series. The potential use of several of these selective ligands as pharmacological tools and therapeutics will be briefly reviewed.

2-Phenylpyrroles as conformationally restricted benzamide analogues. A new class of potential antipsychotics. 2.

A series of 2-phenylpyrrole Mannich bases was synthesized and screened in pharmacological models for antipsychotic activity and extrapyramidal effects. Structure modifications of 5-(4-fluorophenyl)-2-[[4-(2-methoxyphenyl)-1-piperazinyl]methyl]pyrrole (1), the prototype of a new class of sodium-independent atypical dopamine D-2 antagonists, resulted in 2-[[4-(7-benzofuranyl)-1-piperazinyl]methyl]-5-(4-fluorophenyl)pyrrole (15), which was an even more potent and selective D-2 antagonist than the parent compound. The excellent oral activity in the apomorphine-induced climbing behavior and the conditioned avoidance response tests and the absence of catalepsy make this compound particularly promising as a potential antipsychotic with a low propensity to induce acute extrapyramidal side effects.

2-Phenylpyrroles as conformationally restricted benzamide analogues. A new class of potential antipsychotics. 1.

2-Phenylpyrroles were synthesized as conformationally restricted analogues of the substituted benzamide sultopride and the butyrophenones haloperidol and fluanisone. Dopamine antagonistic activity is maintained if the 2-phenylpyrrole side chain is linked to the pharmacophoric N-ethylpyrrolidine moiety of sultopride or to the 4-substituted piperazine moiety of fluanisone but is lost if the 2-phenylpyrrole is combined with the 4-substituted piperidine moiety of haloperidol. The 2-phenylpyrrole analogue 1 of sultopride is in vitro 0.25 and in vivo 3 times as potent as the parent compound. Its binding to the dopamine D-2 receptors is, in analogy to the substituted benzamides, strongly sodium-dependent. The 2-(4-fluorophenyl)pyrrole analogue 5 of fluanisone is superior in vitro as well as in vivo to the corresponding benzamide 7 and the butyrophenone fluanisone. The increase in activity is not only due to a higher affinity for the D-2 receptors but also to an enhanced oral absorption (ratio po/ip = 4.5 vs 40 for the benzamide and 60 for fluanisone). Compound 5 is further characterized by a high selectivity for the D-2 receptors, in contrast to the benzamide and butyrophenone analogues (ratio D-2/alpha 1 = 60, 2.0, and 0.3, respectively). The binding to the D-2 receptors has little dependence on sodium. The 2-phenylpyrrole 5 shares with the benzamide 7 a low potential to induce catalepsy, which is in contrast to haloperidol. So, 5-(4-fluorophenyl)-2-[[4-(2-methoxyphenyl)-1-piperazinyl]methyl]pyrrole (5) is the prototype of a new class of sodium-independent dopamine D-2 antagonists, which may be particularly useful as potential antipsychotics with a low propensity to induce acute extrapyramidal side effects.

Opiate receptor interaction of compounds derived from or structurally related to fentanyl.

The opiate receptor affinity of compounds derived from or structurally related to fentanyl (1) was determined by in vitro receptor binding assays. The relatively high affinity of fentanyl (3 times morphine) was hardly influenced by the introduction of a 2-CH3, 2-OCH3, or a 2-Cl substituent into the anilino phenyl and was moderately reduced by 2-C2H5, 2-OC2H5, and 2,6-(CH3)2 substitution in this ring. Removal of the n-propionyl g in vitro receptor binding assays. The relatively high affinity of fentanyl (3 times morphine) was hardly influenced by the introduction of a 2-CH3, 2-OCH3, or a 2-Cl substituent into the anilino phenyl and was moderately reduced by 2-C2H5, 2-OC2H5, and 2,6-(CH3)2 substitution in this ring. Removal of the n-propionyl g in vitro receptor binding assays. The relatively high affinity of fentanyl (3 times morphine) was hardly influenced by the introduction of a 2-CH3, 2-OCH3, or a 2-Cl substituent into the anilino phenyl and was moderately reduced by 2-C2H5, 2-OC2H5, and 2,6-(CH3)2 substitution in this ring. Removal of the n-propionyl group of the 2-OCH3 derivative, fixation of the anilino phenyl in fentanyl to the propionyl group or the piperidine ring, and replacement of the amide N by C all caused a sharp decline of receptor affinity. Examination of molecular models seemed to indicate that optimal opiate receptor interaction of fentanyl and its derivatives requires a virtually perpendicular position of the anilino phenyl with respect to the amide function.