The glycine site on the NMDA receptor: structure-activity relationships and therapeutic potential.
Explore the source record for details and available documents.
Biomedical subjects
Publications and source records attributed to P D Leeson.
Explore the source record for details and available documents.
Explore the source record for details and available documents.
1. The anticonvulsant and behavioural effects of the glycine/NMDA receptor partial agonist, L-687,414 (R(+)-cis-beta-methyl-3-amino-1-hydroxypyrrolid-2-one) have been investigated in rodents. 2. L-687,414 dose-dependently antagonized seizures induced by N-methyl-D,L- aspartic acid (NMDLA, ED50 = 19.7 mg kg-1), pentylenetetrazol (PTZ, ED50 = 13.0 mg kg-1) and electroshock (ED50 = 26.1 mg kg-1) when given intravenously 15 min before test, in male Swiss Webster mice but was most potent against audiogenic seizures induced by a 120 dB bell in DBA/2 mice (ED50 = 5.1 mg kg-1, i.p., 30 min before test). 3. L-687,414 also induced impairments of performance in a rotarod test in both Swiss Webster and DBA/2 mice and the ratio [rotarod MED:anticonvulsant ED50] varied between 0.9 and 5, depending on the convulsant used. 4. Similar behaviours to those seen after administration of the non-competitive NMDA receptor antagonist, MK-801 (head weaving, body rolling, hyperlocomotion) were seen in the mouse after giving L-687,414, although the peak effect occurred at a dose (100 mg kg-1) which was 5-20 times the anticonvulsant ED50S, depending on the convulsant used. Unlike MK-801, however, doses of L-687,414 that were behaviourally stimulant did not increase dopamine turnover in the nucleus accumbens. 5. Consistent with the interaction of L-687,414 with the glycine/NMDA receptor, the anticonvulsant, ataxic and motor stimulant effects of the compound were significantly attenuated by the glycine/NMDA receptor agonist, D-serine (10-100 micrograms per mouse, i.c.v.). 6. The results show that L-687,414 is a potent, orally active anticonvulsant with a more benign pharmacological profile than antagonists acting at the ion channel of the NMDA receptor complex. The compound is a useful tool with which to probe the functional role of the glycine co-agonist site in vivo.
Most full antagonists at the glycine site of the NMDA receptor contain a carboxylic acid, which we believe to be detrimental to penetration of the blood-brain barrier. By consideration of a pharmacophore, novel antagonists at this site have been designed in which the anionic functionality is a vinylogous acid, in the form of a 4-hydroxyquinolin-2(1H)-one. In this series, a 3-substituent is necessary for binding, and correct manipulation of this group leads to compounds such as the 3-(3-hydroxyphenyl)propargyl ester 24 (L-701,273), with an IC50 for displacement of [3H]-L-689,560 binding of 0.17 microM and Kb against NMDA in the cortical slice of 1.39 microM. Compounds were tested for their ability to prevent audiogenic seizure in DBA/2 mice; the most potent compound in this series is the cyclopropyl ketone 42 (L-701,252), with an ED50 of 4.1 mg/kg ip. A model is proposed for binding to the glycine site, in which an important interaction is of a putative receptor cation with the pi-system of the 3-substituent.
3,4-Dihydro-2(1H)-quinolones, evolved from 2-carboxy-1,2,3,4,- tetrahydroquinolines and 3-carboxy-4-hydroxy-2(1H)-quinolones, have been synthesized and evaluated in vitro for antagonist activity at the glycine site on the NMDA receptor and for AMPA [(RS)-alpha-amino-3- hydroxy-5-methyl-4-isoxazolepropionic acid] antagonist activity. Generally poor potency at the glycine site is observed when a variety of electron-withdrawing substituents are attached to the 3-position of 3,4-dihydro-2(1H)-quinolones. The analogues 5-9 (IC50 values > 100 microM, Table I) exist largely in the 3,4-dipseudoaxial conformation (as evidenced by 1H NMR spectra), whereas the 3-cyano derivative (10, IC50 = 12.0 microM) has a relatively high population of the 3-pseudoequatorial conformer. The 3-nitro analogue (4, IC50 = 1.32 microM) has a pKa approximately 5 and thus exists at physiological pH as an anion with the nitro group planar to the quinolone ring. The general requirement of acidity for high affinity binding at the glycine/NMDA site is supported with the good activity of the other 3-nitro derivatives (13-21), all of which are deprotonated at physiological pH. The 3-nitro-3,4-dihydro-2(1H)-quinolones and 2-carboxy-1,2,3,4-tetrahydroquinolines show quite different structure-activity relationships at the 4-position. The unselective excitatory amino acid activity of 21 is comparable with 6,7-dichloro-quinoxaline-2,3-dione and 6,7-dichloroquinoxalic acid and this suggests similarities in their modes of binding to excitatory amino acid receptors. The broad spectrum excitatory amino acid antagonist activity of the 4-unsubstituted analogue 21 (KbNMDA = 6.7 microM, KbAMPA = 9.2 microM) and the glycine/NMDA selectivity of the other 3-nitro derivatives allows the proposal of a model for AMPA receptor binding which differs from the glycine binding pharmacophore in that there is bulk intolerance adjacent to the 4-position. Compound 21 (L-698,544) is active (ED50 = 13.2 mg/kg) in the DBA/2 mouse anticonvulsant model and is the most potent combined glycine/NMDA-AMPA antagonist yet reported, in vivo, and may prove to be a useful pharmacological tool.
The glycine site on the NMDA receptor complex has generated an enormous amount of interest since it was first described five years ago. In this review by John Kemp and Paul Leeson the structure-activity relationships of agonists, partial agonists and antagonists acting at the glycine site are reviewed, along with what has been learned from studies with these compounds about the role of this site in physiological and pathological processes. Studies with prototype antagonists, and low-efficacy partial agonists that penetrate the brain, show that these compounds possess anticonvulsant and neuroprotective properties but lack some of the side-effects of other types of NMDA antagonists.
2-Carboxy-1,2,3,4-tetrahydroquinoline derivatives, derived from kynurenic acid, have been synthesized and evaluated for in vitro antagonist activity at the glycine site on the NMDA receptor. 2,3-Dihydrokynurenic acids show reduced potency relative to the parent lead compounds (Table I) possibly as a result of conformational effects. Removal of the 4-oxo group results in further reduced potency, but introduction of a cis-carboxymethyl group to the 4-position restores antagonist activity (Tables III and IV). Replacement of the keto group of 5,7-dichloro-2,3-dihydrokynurenic acid with other alternative H-bonding groups, for example cis- and trans-benzyloxycarbonyl and cis- and trans-carboxamido (Table V), gives comparable activity, but there is negligible stereoselectivity. A significant increase in potency and stereoselectivity is seen within the 4-acetate series (Table VI). The trans-4-acetic acid is significantly more potent than the corresponding lead kynurenic acid and has 100-fold greater affinity than the cis isomer. The results are consistent with a requirement in binding for a pseudoequatorially placed 2-carboxylate and clearly demonstrate the importance for binding of a correctly positioned hydrogen-bond-accepting group at the 4-position. The high-affinity binding of an anionic group in the 4-substituent binding pocket suggests that the glycine site and the neurotransmitter recognition (NMDA) site may have some features in common.
trans-2-Carboxy-5,7-dichloro-4-amidotetrahydroquinolines, evolved from the lead 5,7-dichlorokynurenic acid, have been synthesized and tested for in vitro antagonist activity at the glycine site on the N-methyl-D-aspartate (NMDA) receptor. Optimization of the 4-substituent has provided antagonists having nanomolar affinity, including the urea trans-2-carboxy-5,7-dichloro-4[[(phenylamino)carbonyl]amino]-1,2,3, 4-tetrahydroquinoline (35; IC50 = 7.4 nM vs [3H]glycine binding; Kb = 130 nM for block of NMDA responses in the rat cortical slice), which is one of the most potent NMDA antagonists yet found. The absolute stereochemical requirements for binding were found to be 2S,4R, showing that, in common with other glycine-site NMDA receptor ligands, the unnatural configuration at the alpha-amino acid center is required. The preferred conformation of the trans-2,4-disubstituted tetrahydroquinoline system, as shown by X-ray crystallography and 1H NMR studies, places the 2-carboxyl pseudoequatorial and the 4-substituent pseudoaxial. Modifications of the 4-amide show that bulky substituents are tolerated and reveal the critical importance for activity of correct positioning of the carbonyl group. The high affinity of trans-2-carboxy-5,7-dichloro-4-[1-(3-phenyl-2-oxoimidazolidinyl)]- 1,2,3,4-tetrahydroquinoline (55; IC50 = 6 nM) suggests that the Z,Z conformer of the phenyl urea moiety in 35 is recognized by the receptor. Molecular modeling studies show that the 4-carbonyl groups of the kynurenic acids, the tetrahydroquinolines, and related antagonists based on N-(chlorophenyl)glycine, can interact with a single putative H-bond donor on the receptor. The results allow the establishment of a three-dimensional pharmacophore of the glycine receptor antagonist site, incorporating a newly defined bulk tolerance/hydrophobic region.
Explore the source record for details and available documents.
The glycine site on the N-methyl-D-aspartate (NMDA) subtype of receptors for the excitatory neurotransmitter glutamate is a potential target for the development of neuroprotective drugs. We report here two chemical series of glycine site antagonists derived from kynurenic acid (KYNA), with greatly improved potency and selectivity. Disubstitution with chlorine or bromine in the 5- and 7-positions of KYNA increased affinity for [3H]glycine binding sites in rat cortex/hippocampus P2 membranes, with a parallel increase of potency for antagonism of NMDA-evoked responses in the rat cortical wedge preparation. The optimal compound was 5-I,7-Cl-KYNA, with an IC50 for [3H]glycine binding of 29 nM and an apparent Kb in the cortical wedge preparation of 0.41 microM. Reduction of the right-hand ring of 5,7-diCl-KYNA reduced affinity by 10-fold, but this was restored by substitution in the 4-position with the trans-phenylamide and further improved in the trans-benzylamide. The optimal compound was the transphenylurea (L-689,560), with an IC50 of 7.4 nM and an apparent Kb of 0.13 microM. Both series of compounds displayed a high degree of selectivity for the glycine site, having IC50 values of greater than 10 microM versus radioligand binding to the glutamate recognition sites of NMDA, alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionate (AMPA), and kainate receptors and the strychnine-sensitive glycine receptor. Selectivity versus AMPA receptor-mediated responses was also apparent in the rat cortical wedge and in patch-clamp recordings of cortical neurons in culture. Experiments using [3H]dizocilpine (MK-801) binding indicated that 5,7-diBr-KYNA, 5,7-diCl-KYNA, 5-I,7-Cl-KYNA, and L-689,560 all behaved as full antagonists and were competitive with glycine. Patch-clamp recordings of cortical neurons in culture also indicated that NMDA-induced currents were antagonized by competition for the glycine site, and gave no evidence for partial agonist activity. pKi values for 5,7-diBr-KYNA and L-689,560 in these experiments were 7.2 and 7.98, respectively, similar to the affinities of these compounds in the glycine binding assay. The high affinity and selectivity of these new derivatives make them useful tools to investigate the function of the glycine site on the NMDA receptor.
The binding characteristics of [3H]L-689,560 [(+/-)-4-(trans)-2-carboxy-5,7-dichloro-4-phenylaminocarbonylamino -1,2,3,4- tetrahydroquinoline], a selective antagonist for the glycine site on the N-methyl-D-aspartate receptor, have been evaluated using rat cortex/hippocampus P2 membranes. Specific [3H]L-689,560 binding was saturable, having a Kd of 2.97 nM and a Bmax of 4.15 pmol/mg of protein. The Bmax value was not significantly different from that obtained for [3H]glycine in the same membrane preparation, and L-689,560 and glycine were found to be mutually competitive. The specific binding of [3H]L-689,560 (1 nM) represented 96 +/- 0.02% (four experiments) of total binding. Association experiments at 4 degrees revealed that [3H]L-689,560 reached equilibrium in 120 min, with a t1/2 of 40 min. The dissociation of [3H]L-689,560 was slow at 4 degrees (t1/2 = 118 min), allowing the use of filtration to separate free from bound radioactivity. Both association and dissociation curves were best fitted by a double-exponential function, suggesting the presence of two components. Comparison of IC50 values obtained using [3H]glycine and [3H]L-689,560 binding for 21 glycine site ligands (including agonists, partial agonists, and antagonists, with affinities spanning 5 orders of magnitude) showed a 1:1 correlation, with a correlation coefficient of 0.97. This suggests that efficacy does not have a large influence on the affinity of glycine site ligands when an agonist or antagonist radioligand is used. Ligands for other amino acid recognition sites did not directly inhibit [3H]L-689,560 binding. [3H]L-689,560 is an improved radioligand for the glycine site that will facilitate further investigations of its properties.
In an investigation of novel potential bile acid sequestrants, the affinities of the sodium salts of the glycine and taurine conjugates of naturally occurring bile acids (cholate, deoxycholate, chenodeoxycholate and lithocholate) for several cationic ammonium bile acid derivatives have been investigated by measurements of the extent to which the derivatives are able to precipitate the bile acids. This is roughly proportional to the lipophilicity of the interacting species. Thus, amino and ammonium derivatives of cholic acid do not precipitate taurocholate or glycocholate to any great extent, whereas ammonium derivatives of deoxycholate and lithocholate are much more effective. To complement the precipitation measurements, high resolution 13C-NMR has been applied to investigate the weaker interactions between the ammonium cholate derivative and glycocholate, glycodeoxycholate and glycochenodeoxycholate. Addition of either of the latter two bile acids to the cationic ammonium compound results in considerable broadening of the 13C resonances of both species, indicating the formation of relatively rigid structures. In addition, we have used T2 relaxation enhancement induced by spin-labelled fatty acids to examine the mechanism of interaction with bile acids of amphiphilic anions, which might compete with bile acids for sites on bile acid sequestrants. Low concentrations of 16-DOXY L-Stearate dramatically broaden the 13C-NMR resonances of deoxycholate carbons 19, 18 and 7 in particular, while 5-DOXY L-Stearate exerts much less specific effects. These results have been incorporated into a snapshot model of bile acid-fatty acid interactions.
Derivatives of the nonselective excitatory amino acid antagonist kynurenic acid (4-oxo-1,4-dihydroquinoline-2-carboxylic acid, 1) have been synthesized and evaluated for in vitro antagonist activity at the excitatory amino acid receptors sensitive to N-methyl-D-aspartic acid (NMDA), quisqualic acid (QUIS or AMPA), and kainic acid (KA). Introduction of substituents at the 5-, 7-, and 5,7-positions resulted in analogues having selective NMDA antagonist action, as a result of blockade of the glycine modulatory (or coagonist) site on the NMDA receptor. Regression analysis suggested a requirement for optimally sized, hydrophobic 5- and 7-substituents, with bulk tolerance being greater at the 5-position. Optimization led to the 5-iodo-7-chloro derivative (53), which is the most potent and selective glycine/NMDA antagonist to date (IC50 vs [3H]glycine binding, 32 nM; IC50's for other excitatory amino acid receptor sites, greater than 100 microM). Substitution of 1 at the 6-position resulted in compounds having selective non-NMDA antagonism and 8-substituted compounds were inactive at all receptors. The retention of glycine/NMDA antagonist activity in heterocyclic ring modified analogues, such as the oxanilide 69 and the 2-carboxybenzimidazole 70, suggests that the 4-oxo tautomer of 1 and its derivatives is required for activity. Structurally related quinoxaline-2,3-diones are also glycine/NMDA antagonists, but are not selective and are less potent than the 1 derivatives, and additionally show different structure-activity requirements for aromatic ring substitution. On the basis of these results, a model accounting for glycine receptor binding of the 1 derived antagonists is proposed, comprising (a) size-limited, hydrophobic binding of the benzene ring, (b) hydrogen-bond acceptance by the 4-oxo group, (c) hydrogen-bond donation by the 1-amino group, and (d) a Coulombic attraction of the 2-carboxylate. The model can also account for the binding of quinoxaline-2,3-diones, quinoxalic acids, and 2-carboxybenzimidazoles.
Displacement of [3H]MK-801 (dizocilpine, 1) binding to rat brain membranes has been used to evaluate the affinities of novel dibenzocycloalkenimines related to 1 for the ion channel binding site (also known as the phencyclidine or PCP receptor) on the N-methyl-D-aspartate (NMDA) subtype of excitory amino acid receptor. In common with many other agents having actions in the central nervous system, these compounds contain a hydrophobic aromatic moiety and a basic nitrogen atom. The conformational rigidity of these ligands provides a unique opportunity to evaluate the importance of specific geometrical properties that influence active-site recognition, in particular the role of the nitrogen atom in hydrogen-bonding interactions. The relative affinities (IC50s) of hydrocarbon-substituted analogues of 1 and ring homologated cyclooctenimines illustrate the importance of size-limited hydrophobic binding of both aryl rings and of the quaternary C-5 methyl group. Analysis of the binding of a series of the 10 available structurally rigid dibenzoazabicyclo[x.y.z]alkanes, by using molecular modeling techniques, uncovered a highly significant correlation between affinity and a proposed ligand-active site hydrogen bonding vector (r = 0.950, p less than 0.001). These results are used to generate a pharmacophore of the MK-801 recognition site/PCP receptor, which accounts for the binding of all of the known ligands.
The antagonist effect of (+/-)-3-amino-1-hydroxypyrrolid-2-one (HA-966) at the N-methyl-D-aspartate (NMDA) receptor occurs through a selective interaction with the glycine modulatory site within the receptor complex. When the enantiomers of (+/-)-HA-966 were resolved, the (R)-(+)-enantiomer was found to be a selective glycine/NMDA receptor antagonist, a property that accounts for its anticonvulsant activity in vivo. In contrast, the (S)-(-)-enantiomer was only weakly active as an NMDA-receptor antagonist, but nevertheless it possessed a marked sedative and muscle relaxant action in vivo. In radioligand binding experiments, (+)-HA-966 inhibited strychnine-insensitive [3H]glycine binding to rat cerebral cortex synaptic membranes with an IC50 of 12.5 microM, whereas (-)-HA-966 had an IC50 value of 339 microM. In electrophysiological experiments, (+)-HA-966 selectively antagonized NMDA receptor responses in rat cortical slices, whereas the (-)-enantiomer was much weaker. On cultured cortical neurones (+)-HA-966 inhibited glycine-potentiated NMDA responses with an IC50 = 13 microM compared with (-)-HA-966, which has an IC50 = 708 microM. In agreement with findings with racemic HA-966, even high concentrations of (+)-HA-966 did not completely inhibit NMDA responses, suggesting that (+)-HA-966 is a low-efficacy partial agonist. (+)-HA-966 produced parallel shifts to the right of the glycine concentration curve for potentiation of NMDA responses, resulting in an estimated pKb = 5.6. In mice, (+)-HA-966 antagonized sound and N-methyl-DL-aspartic acid (NMDLA)-induced seizures with ED50 values of 52.6 mg/kg of body weight (i.p.) and 900 mg/kg (i.v.), respectively. The coadministration of D-serine dose-dependently (10-100 micrograms into the cerebral ventricles per mouse) antagonized the anticonvulsant effect of a submaximal dose of (+)-HA-966 (100 micrograms administered directly into the cerebral ventricles) against NMDLA-induced seizures. The sedative/ataxic effect of racemic HA-966 was mainly attributable to the (-)-enantiomer, which was greater than 25-fold more potent than the (+)-enantiomer. It is suggested that, as in the case of the sedative gamma-butyrolactone, disruption of striatal dopaminergic mechanisms may be responsible for this action.
Explore the source record for details and available documents.
Interactions between prealbumin and several thyromimetic compounds have been examined by proton magnetic resonance spectroscopy. One equivalent of thyroxine (T4) or reverse triiodothyronine (rT3) selectively broadens a number of protein signals, while addition of a second equivalent induces much less widespread changes. One equivalent of triiodothyronine (T3), however, produces much less dramatic changes, and effects comparable with T4 and rT3 are only apparent when a second equivalent binds. The broadening is ascribed to immobilization of flexible residues. The non-halogenated analogue 3,5-dimethyl-3'-isopropylthyronine induces qualitatively different changes suggesting incomplete entry into the thyromimetic binding channel. The fluorinated analogue SK&F 95049 (3,5-bis-thiotrifluoromethyl-3'-isopropylthyronine) induces very similar changes to T3. A fluorine-19 NMR signal with a half-height line width of approximately 150 Hz can be observed from the bound ligand. Finally, a spin-labeled T4 analogue, with a nitroxyl on the alanyl moiety, induces changes identical to those induced by T4 itself, and additionally broadens some signals corresponding to residues at the opening of the ligand binding channel. The natural tryptophan fluorescence of the protein is shown to be a sensitive indicator of binding. The possible influence of the dynamic restrictions induced by binding the first molecule of T4 or rT3 on the protein's affinity for a second hormone is discussed. It is suggested that the first interaction confers rigidity on the second site and reduces its ability to flex open and accommodate a second thyromimetic, which results in the marked negative co-operativity associated with the occupancy of this site.
Introduction of specific arylmethyl groups at the 3'-position of the thyroid hormone 3,3',5-triiodo-L-thyronine (T3), and its known hormonally active derivatives, gives liver-selective, cardiac-sparing thyromimetics, with potential utility as plasma cholesterol lowering agents. Selectivity-conferring 3'-substituents include substituted benzyl, e.g. p-hydroxybenzyl, and heterocyclic methyl, e.g. 2-oxo-1,2-dihydropyrid-5-ylmethyl and 6-oxo-1,6-dihydropyridazin-3-ylmethyl. Correlations between in vivo and in vitro receptor binding affinities show that liver/heart selectivity does not depend on receptor recognition but on penetration or access to receptors in vivo. QSAR studies of the binding data of a series of 20 3'-arylmethyl T3 analogues show that electronegative groups at the para position increase both receptor binding and selectivity in vivo. However, increasing 3'-arylmethyl hydrophobicity increases receptor binding but reduces selectivity. Substitution at ortho and meta positions reduces both binding and selectivity. Replacement of the 3,5-iodo groups by halogen or methyl maintains selectivity, with 3,5-dibromo analogues in particular having increased potency combined with oral bioavailability. Diphenyl thioether derivatives also have improved potency but are less orally active. At the 1-position, the D enantiomer retains selectivity, but removal of the alpha-amino group to give a propionic acid results in loss of selective thyromimetic activity.