A 17beta-derivative of allopregnanolone is a neurosteroid antagonist at a cerebellar subpopulation of GABA A receptors with nanomolar affinity.

BACKGROUND AND PURPOSE: High-affinity, subtype-selective antagonists of the neurosteroid binding sites of GABA(A) receptors are not available. We have characterized an allopregnanolone derivative as an antagonist of cerebellar GABA(A) receptors with nanomolar affinity. EXPERIMENTAL APPROACH: Receptor binding and electrophysiological methods were used for the allosteric modulation of cerebellar GABA(A) receptors by an allopregnanolone derivative, (20R)-17beta-(1-hydroxy-2,3-butadienyl)-5alpha-androstane-3alpha-ol (HBAO). GABA(A) receptors of rat cerebellar membranes were labelled with the chloride channel blocker [(3)H]ethynylbicycloorthobenzoate (EBOB). The ionophore function of GABA(A) receptors was studied by whole-cell patch clamp electrophysiology in cultured rat cerebellar granule and cortical cells. KEY RESULTS: Partial displacement of cerebellar [(3)H]EBOB binding by nanomolar HBAO was attenuated by 0.1 mM furosemide, an antagonist of alpha(6) and beta(2-3) subunit-containing GABA(A) receptors. Displacement curves of HBAO were reshaped by 30 nM GABA and shifted to the right. However, the micromolar potency of full displacement by allopregnanolone was not affected by 0.1 mM furosemide or 30 nM GABA. The nanomolar, but not the micromolar phase of displacement of [(3)H]EBOB binding by GABA was attenuated by 100 nM HBAO. Submicromolar HBAO did not affect [(3)H]EBOB binding to cortical and hippocampal GABA(A) receptors. HBAO up to 1 microM did not affect chloride currents elicited by 0.3-10 microM GABA, while it abolished potentiation by 1 microM allopregnanolone with nanomolar potency in cerebellar but not in cortical cells. Furosemide attenuated cerebellar inhibition by 100 nM HBAO. CONCLUSIONS AND IMPLICATIONS: HBAO is a selective antagonist of allopregnanolone, a major endogenous positive modulator via neurosteroid sites of cerebellar (probably alpha(6)beta(2-3)delta) GABA(A) receptors.

Bicarbonate and thiocyanate ions affect the gating of gamma-aminobutyric acid(A) receptors in cultured rat cortical cells.

The ionophore function of gamma-aminobutyric acid(A) (GABA(A)) receptors was studied by whole-cell patch clamp electrophysiology in primary cultures of rat cerebral cortex. Chloride ions were replaced for SCN(-) (thiocyanate) and HCO(-3) (bicarbonate) ions. The EC(50) values of the GABA(A) agonist muscimol (HCO(-3)>Cl(-)>SCN(-)) varied parallel with the free energies of dehydration of the anions, while the inhibition constants of the GABA(A) antagonist bicuculline methiodide were not affected. These findings might be relevant in the interpretation of the contribution of HCO(-3) versus Cl(-) currents to the pharmacological differences between depolarizing and hyperpolarizing GABA responses.

Subunit-specific modulation of glycine receptors by neurosteroids.

The effects of pregnene and androstane steroids were studied on recombinant human glycine receptors (GlyRs) by whole-cell voltage-clamp electrophysiology. The 3beta-sulphates of pregnenolone (PREGS) and dehydroepiandrosterone (DHEAS) inhibited GlyR currents with K(I) values of 2-20 microM for different (alpha(1), alpha(2), alpha(4) and beta) GlyR subunits. PREGS resulted in a parallel shift of the response curve of glycine for alpha(1) GlyRs. The inhibitory potencies of DHEAS relative to PREGS were decreased in transition from embryonic alpha(2) towards adult alpha(1)beta GlyRs. A decreased potency of DHEAS for alpha(4) versus alpha(2) GlyRs represents the first pharmacological difference reported between these subunits. A negative charge at C3 is required for GlyR antagonism but androsterone sulphate epimers at C3 inhibited without stereoselectivity. Some point mutations of alpha(1) GlyRs with characteristic functional consequences did not significantly affect the inhibitory potency of PREGS. Progesterone selectively inhibited alpha(2) GlyRs, while PREG and its acetic ester potentiated alpha(1) GlyRs. Coexpression of the alpha subunits with the beta subunit eliminated the enhancing effects of PREG and attenuated the inhibitory potencies of the neurosteroids. Based on these data we propose that neurosteroids might modulate perinatal GlyR activity and thereby influence neuronal development.

Entropy as the predominant driving force of binding to human recombinant alpha(x)beta(3)gamma(2) GABA(A) receptors.

In order to study the correlation of the thermodynamic driving forces of binding with the efficacies of displacing ligands, the specific binding of [3H]SR 95531 [2-(3-carboxypropyl)3-amino-6-p-methoxyphenylpyridazinium bromide], a GABA(A) receptor antagonist, was studied in cell lines stably expressing human alpha(1)beta(3)gamma(2) and alpha(2)beta(3)gamma(2) GABA(A) receptors. Displacing potencies for the agonists with different efficacies (muscimol, 4,5,6,7-tetrahydroisoxazolo[5,4-c]pyridin-3-ol (THIP) and piperidine-4-sulfonic acid) and for antagonists (SR 95531 and 5-(4-piperidyl)isothiazol-3-ol) were determined at 0 degrees C, 20 degrees C and 37 degrees C. Displacing potencies were temperature-nearly independent for alpha(1)beta(3)gamma(2) receptors. At alpha(2)beta(3)gamma(2), receptor binding of the antagonists was exothermic, endothermic for the agonists THIP and piperidine-4-sulfonic acid and isothermic for muscimol. The free energy increments of displacement for the binding of the antagonist [3H]SR 95531 versus the agonist [3H]muscimol approach saturation as a function of the efficacies of the displacers only for alpha(1)beta(3)gamma(2) receptors. This suggests that, for binding to alpha(1)beta(3)gamma(2) GABA(A) receptors, displacement is an efficacy-dependent interaction predominantly driven by entropic increases.

Allosteric modulators affect the efficacy of partial agonists for recombinant GABA(A) receptors.

Different alpha subunits of human gamma-aminobutyric acid type A (GABA(A)) receptors were transiently expressed together with beta(3) and gamma(2) subunits in Xenopus oocytes to examine the interactions of various GABA(A) agonists and representative allosteric modulators. Chloride currents elicited by agonists were measured using two electrode voltage clamp electrophysiology. Where compounds behaved as full agonists, i.e. GABA on all subtypes and 4,5,6, 7-tetrahydroisoxazolo [5,4-c]pyridin-3-ol (THIP) on alpha2beta(3)gamma(2) GABA(A) receptors, agonist concentration-response curves were shifted to the left by the benzodiazepine full agonist chlordiazepoxide and the anticonvulsant loreclezole, or to the right by the inverse agonist 6, 7-dimethoxy-4-ethyl-beta-carboline-3-carboxylic acid methyl ester (DMCM), with no effect on the maximal currents (I(max)). In contrast, maximal responses for different partial GABA(A) agonists on all benzodiazepine-sensitive alpha(x)beta(3)gamma(2) GABA(A) receptors were enhanced by chlordiazepoxide. I(max) values for piperidine-4-sulphonic acid (P4S) on alpha(1)beta(3)gamma(2), THIP on alpha(3)beta(3)gamma(2), and 5-(4-piperidyl)isothiazol-3-ol (thio-4-PIOL) on alpha(2)beta(3)gamma(2) and alpha(5)beta(3)gamma(2) GABA(A) receptors were increased by chlordiazepoxide, while that for P4S on alpha(1)beta(3)gamma(2) receptors was decreased by DMCM. The I(max) values for partial agonists were also enhanced by pentobarbitone, the neurosteroid allopregnanolone and loreclezole irrespective of receptor subtype or the nature of the partial agonist. In the light of models of ligand-gated ion channel receptor activation we suggest two possible mechanisms of action for the effects of allosteric modulators on partial agonist receptor activation: either selective modulation of agonist affinity for the open/closed state, or direct modulation of the gating process itself.

Selective blocking effects of tropisetron and atropine on recombinant glycine receptors.

Some serotonin 5-HT3 receptor ligands of tropeine structure have been recently shown to modulate ionophore function and binding of glycine receptors. This led us to study the effects of the tropeines tropisetron and atropine on recombinant human glycine receptors transiently expressed in Xenopus oocytes by using whole-cell voltage-clamp electrophysiology. Glycine currents were inhibited by atropine in an apparently competitive manner and with considerable selectivity of the tropeines for alpha2 versus alpha1 subunits. Coexpression of beta with alpha subunits and replacement of the N-terminal region of the alpha1 subunits by the corresponding beta segment resulted in similar increases in the inhibitory potencies. Our data suggest common sites of the tropeines for inhibition on the N-terminal region of glycine receptors. The point mutations R271K and R271L of the alpha1 subunit decreased, whereas a T112A substitution increased, the inhibition constants (Ki) of the tropeines. These changes in the Ki values of the tropeines were associated with opposite changes in the EC50 of glycine. Selectivities for the tropeines versus glycine (EC50/Ki) varied within three orders of magnitude. These results, when expressed in terms of free energy changes, can be interpreted according to a two-state receptor model.

Bimodal action of furosemide on convulsant [3H]EBOB binding to cerebellar and cortical GABA(A) receptors.

Picrotoxinin-sensitive binding of a convulsant 4'-ethynyl-4-n[2,3-3H2]propyl-bicycloorthobenzoate ([3H]EBOB) to gamma-aminobutyric acid type A (GABA(A)) receptors was characterized in rat cerebrocortical and cerebellar membranes. The non-penetrating organic anions, furosemide and niflumate, in spite of their structural similarities, exerted differential effects on [3H]EBOB binding. Furosemide, a loop diuretic and a specific antagonist of a cerebellar GABA(A) receptor population, and GABA decreased the inhibitory potencies of each other in the cerebellum, while enhanced them in the cortex. The inhibitory potencies of niflumate, an anti-inflammatory and a chloride channel blocker. and GABA were enhanced by each other both in the cerebellum and cortex. Removal of chloride ions did not modify the effects of furosemide on [3H]EBOB binding. Furosemide antagonized the inhibition of cerebellar [3H]EBOB binding by a low pentobarbital concentration (0.1 mM), but enhanced the inhibition by a high concentration (0.5 mM). The results indicate that [3H]EBOB binding can be used to detect the known pharmacological features of the cerebellar granule cell-specific 16 subunit-containing GABA(A) receptors. The data extends the properties of furosemide antagonism of this receptor subtype to chloride insensitivity and interactions with barbiturate sites.

Differential effects of thiocyanate on the binding thermodynamics of bicuculline methiodide versus SR95531 (gabazine) to the gamma-aminobutyric acid receptor-ionophore complex.

The temperature dependence of the binding of [3H]SR 95531 (Gabazine), an antagonist of gamma-aminobutyric acid (GABA(A)) receptors, was studied in synaptosomal membranes of rat brain in the presence of 50 mM KSCN. The displacing potencies of the antagonists bicuculline methiodide and Gabazine were determined at five temperatures between 0 degrees and 37 degrees. Van't Hoff plots of the displacing potencies were analyzed by linear regression in the presence and absence of thiocyanate. Thiocyanate hardly affected the exothermic ionic binding interaction of gabazine. In contrast, thiocyanate strongly potentiated the binding of bicuculline methiodide and deprived it of its exothermic nature. The enhanced binding of bicuculline methiodide in the presence of chaotropic SCN- ions might be reconciled with "entropic trapping" in a sterically constrained hydrophobic binding pocket.

Bidirectional allosteric modulation of strychnine-sensitive glycine receptors by tropeines and 5-HT3 serotonin receptor ligands.

Specific binding of [3H]strychnine was studied on membranes prepared from rat spinal cord. Several antagonists and agonists of 5-HT3 receptors and tropane derivatives displaced [3H]strychnine binding with micromolar potencies. In the presence of 10 microM glycine a high affinity (nanomolar) component of displacement was also observed for the tropeines zatosetron, bemesetron and tropisetron. The displacing potency of glycine was also enhanced by these agents which are therefore termed glycine-positive. In contrast, atropine, SR 57227A, m-chlorophenylbiguanide, metoclopramide and granisetron are termed glycine-negative, because they decreased the displacing potency of glycine while glycine decreased the displacing potencies of atropine and metoclopramide. The dissociation of [3H]strychnine binding was accelerated in the presence of m-chlorophenylbiguanide, SR 57227A, atropine and zatosetron with a concentration dependence (EC50 values and Hill slopes) similar to their displacing effects. This demonstrates that the displacement of strychnine binding is associated with allosteric interactions between different binding sites. Structure-activity analysis revealed that the tropeine structure is essential for high affinity binding, and its substitutions (in scopolamine and cocaine) or its replacement (in ondansetron and metoclopramide) strongly decrease the potency and/or efficacy of allosteric modulation. High affinity modulatory sites for tropeines appear to be associated with the potentiation of ionophore function, but distinct from the low affinity channel blocking sites as well as from the binding sites of strychnine and glycine.

From kinetics and thermodynamics of GABAA receptor binding to ionophore function.

The time- and temperature-dependencies of binding are overviewed for the benzodiazepine, GABA and convulsant binding sites of the GABAA receptor-ionophore complex. Kinetic separation of the dissociation phases of a beta-carboline inverse agonist demonstrated the heterogeneity of its binding sites. The kinetics and thermodynamics of benzodiazepine binding alone do not correlate with ionophore function. The majority of the data suggest that agonist- and antagonist-preferring conformations exist for GABAA receptors. The high affinity binding of GABAA antagonists (SR 95531 and bicuculline) corresponds to the (super) low affinity binding of GABA. The correlation between the thermodynamic parameters of binding and efficacies common for GABAA and glycine receptor agonists and antagonists supports the functional similarities of these anionophore complexes. Binding kinetics of the bicyclic cage convulsants show several correlations with ionophore function because the convulsant sites are most intimately coupled to the ion channels. Kinetic interactions of the convulsant sites with the binding sites of benzodiazepines, GABA and central depressants have revealed several pharmacologically relevant allosteric GABAergic modulatory effects. Arrhenius analysis, Hammond's postulate and transition state theory were applied for the dissociation of convulsants. A kinetic model of interconvertible multiaffinity states of the convulsant sites shows correlations with the functional states of the GABAA ionophore.

Distinct thermodynamic parameters of serotonin 5-HT3 agonists and antagonists to displace [3H]granisetron binding.

Specific binding of [3H]granisetron was examined to serotonin 5-HT3 receptors in synaptosomal membranes of rat cerebral cortex between 1 and 37 degrees C. Displacing potencies were determined for 5-HT3 antagonists (granisetron, ondansetron, tropisetron, and d-tubocurarine) and agonists (5-hydroxytryptamine, 2-methyl-5-hydroxytryptamine, phenylbiguanide, m-chlorophenylbiguanide, and SR 57227A). Displacing potencies of the agonists decreased with decreasing temperature. In contrast, displacing potencies of all antagonists increased with decreasing temperature, whereas those of tropisetron and d-tubocurarine passed a maximum. Scatchard analysis of [3H]granisetron binding resulted in KD values lower than the IC50 values of granisetron and a decreasing number of binding sites at higher temperatures. It can be reconciled with temperature-dependent agonist and antagonist states of 5-HT3 receptors. A semiquantitative thermodynamic analysis was based on displacing potencies. The distinct patterns for the signs of entropy, enthalpy, and heat capacity changes on binding can be reconciled with ionic interactions for agonists and hydrophobic interactions for antagonists. The distinctive differences in these thermodynamic parameters exceed those for GABAA and glycine receptor-ionophore complexes.

Thermodynamics and kinetics of t-butylbicyclophosphorothionate binding differentiate convulsant and depressant barbiturate stereoisomers acting via GABAA ionophores.

The temperature dependence of [35S]-t-butylbicyclophosphorothionate (TBPS) binding to the convulsant sites of the GABAA receptor complex was studied in membrane preparations of rat forebrain. Although specific [35S]TBPS binding was maximal around 20 degrees C, the rate constants of dissociation decreased monotonously between 37 degrees C and 2 degrees C. The displacing potencies of the convulsant S(+) enantiomer of 1-methyl-5-phenyl-5-propyl-barbituric acid (MPPB) (IC50 = 1250 +/- 30 microM) and the depressant R(-) MPPB (IC50 = 310 +/- 5 microM) did not show significant changes between 19 degrees C and 37 degrees C. Therefore barbiturate binding seems to be driven by entropic, rather than enthalpic changes. An excess of MPPB enantiomers elicited accelerated and polyphasic dissociations of [35S]TBPS as compared to the monophasic dissociation by TBPS. Arrhenius analysis was applied to the measurable initial rate constants of dissociation. Arrhenius plots were linear between 2 degrees C and 37 degrees C. Activation parameters were similar when [35S] TBPS dissociation was triggered by the convulsants TBPS and S(+) MPPB. It can be attributed to similar conformations of the closed ionophore complex. In contrast, the depressant R(-) MPPB strongly decreased the activation energy of TBPS dissociation from the open ionophore ternary complex. In whole-cell patch-clamp experiments R(-) MPPB, but not S(+) MPPB, elicited chloride currents in rat primary cortical cultures with an EC50 value of 560 +/- 30 microM and a Hill coefficient of 2.9 +/- 0.2. These currents were similar to those elicited by GABA and blocked by TBPS. A kinetic scheme is proposed for the dissociation of TBPS and to explain the different effects of MPPB enantiomers. Submillimolar R(-) MPPB is supposed to bind to (about three) barbiturate sites on GABAA-ionophores and to open them in a cooperative manner to result in a decreased activation energy for accelerated displacement of convulsant binding.

Diazepam biphasically modulates [3H]TBOB binding to the convulsant site of the GABAA receptor complex.

Interactions of GABA, bicuculline methochloride and diazepam with [3H]TBOB binding to rat brain membranes were evaluated in vitro. GABA displaced [3H]TBOB binding with and IC50 of 4 microM and a slope factor near unity. The competitive GABA antagonist bicuculline methochloride shifted the displacement curve of GABA parallelly to the right, indicating that the interaction of GABA with [3H]TBOB binding is of an allosteric nature. In the presence of GABA, diazepam displaced the binding of [3H]TBOB according to a two-site model: a high affinity site with an IC50 of about 50 nM and a lower affinity site with an IC50 of about 30 microM. Bicuculline methochloride abolished the nanomolar displacement by diazepam and increased the micromolar IC50 value. These results indicate that the interaction of the high affinity diazepam site with the [3H]TBOB binding site is totally GABA dependent and that the low affinity effect of diazepam on [3H]TBOB binding is at least partially GABA dependent. It is likely that the low affinity potency of diazepam to displace [3H]TBOB binding has physiological relevance.

Common modes of action of gamma-butyrolactones and pentylenetetrazol on the GABAA receptor-ionophore complex.

The effects of pentylenetetrazol and bicyclic gamma-butyrolactones of similar stereostructures were studied on the convulsant and benzodiazepine binding sites and chloride ionophore activity of the gamma-aminobutyric acid (GABAA) receptor-complex. Bicyclic gamma-butyrolactones displayed millimolar IC50 values and low stereoselectivities on [35S]t-butylbicyclophosphorothionate (TBPS) binding to the convulsant sites in synaptosomal membranes of rat forebrains. Ring saturation of bicyclic gamma-butyrolactones decreased their IC50 values by one order of magnitude. The IC50 values of saturated bicyclic gamma-butyrolactones and pentylenetetrazol were increased by GABA versus its antagonist R 5135 (3 alpha-hydroxy-16-imino-5 beta,17-aza-androstan-11-one). A bicyclic gamma-butyrolactone and pentylenetetrazol accelerated the dissociation of [35S]TBPS, displaced [3H]flumazenil binding in two phases and blocked the muscimol-elicited chloride currents in patch-clamped cortical neurones in culture in a similar manner. These similar effects on binding and ionophore function support their common modes of action on the GABAA receptor-ionophore complex.

Thermodynamics of gamma-aminobutyric acid type A receptor binding differentiate agonists from antagonists.

Specific binding of the gamma-aminobutyric acid (GABA)A antagonist [3H]SR 95531 to synaptosomal membranes of rat whole brain was examined between 0 degrees and 37 degrees. Scatchard analysis revealed two (high and low affinity) populations of [3H]SR 95531 binding sites. The Kd values increased with increasing temperature. Ki values for GABAA agonists and antagonists were determined from the displacement of [3H]SR 95531 binding at a low concentration (1.8 nM) of [3H]SR 95531, which binds predominantly to high affinity sites. For most compounds van't Hoff plots (--In Ki, i.e., In Ka, versus 1/T) were linear between 0 degrees and 37 degrees. Curvilinear van't Hoff plots for the antagonists R 5135 and bicuculline methiodide can be attributed to their hydrophobic binding interactions. The enthalpy changes of binding (delta H degrees) were positive for the agonists (muscimol, isoguvacine, GABA, 4,5,6,7-tetrahydroisoxazolo[4,5-c]pyridin-3-ol hydrochloride, and imidazole-4-acetic acid) and negative for the antagonists (pitrazepin, bicuculline methiodide, R 5135, SR 95531, and SR 95103). Separation of the enthalpic and entropic components of the Gibbs free energy changes of binding (delta G degrees) revealed that binding of the antagonists is driven by both the enthalpic and entropic terms, whereas that of the agonists is driven entirely by entropy changes. A plot of the entropic term (-T delta S degrees) versus the enthalpic term (delta H degrees) showed separate patterns for GABAA agonists and antagonists, with the partial agonists [5-(4-piperidyl)isoxazol-3-ol, imidazole-4-acetic acid, and 4,5,6,7-tetrahydroisoxazolo[4,5-c]pyridin-3-ol hydrochloride] between them. It is proposed that the entropic term is partly determined by a transition from antagonist to agonist conformation of the GABAA binding sites.

Interconvertible kinetic states of t-butylbicycloorthobenzoate binding sites of the gamma-aminobutyric acidA ionophores.

The kinetics of t-[3H]butylbicycloorthobenzoate (TBOB) binding to the convulsant sites of the gamma-aminobutyric acidA (GABAA) receptor-ionophore complex were examined in synaptosomal membrane preparations of rat brain. On and off rates of TBOB binding were accelerated by 1 microM GABA and decelerated by 1 microM bicuculline methochloride, a GABAA antagonist. The presence of GABA and bicuculline methochloride created rapid and slow phases of dissociation, respectively. The three groups of rate constants distinguished for the dissociation of 4 nM and 30 nM [3H]TBOB represent multi-affinity states of the convulsant sites depending on the presence of GABA or bicuculline methochloride. Apparent association rate constants do not obey the equation k(app) = k(off) + k(on) [TBOB] without assuming interconvertibility of the kinetic states during binding. Avermectin B1a (AVM B1a), a chloride channel opening agent, accelerated the association and dissociation of TBOB and resulted in a biphasic effect on TBOB binding, i.e., enhancement at low concentrations (EC50 7.8 nM) followed by displacement at high concentrations (IC50 6.3 microM) of AVM B1a. AVM B1a resulted in similar biphasic effects on t-[35S]butylbicyclophosphorothionate binding. DIDS, an isothiocyanatostilbene derivative with irreversible anion channel blocking effect, selectively inhibited basal [3H]TBOB binding (IC50 125 microM DIDS) leaving the enhancement by AVM B1a unaffected.

Partial and full agonists/inverse agonists affect [35S]TBPS binding at different occupancies of central benzodiazepine receptors.

Concentration-dependent effects of benzodiazepine receptor ligands were examined on nonequilibrium binding of t-butylbicyclophosphoro[35S]thionate (TBPS, 20 min of incubation at 25 degrees C) to synaptosomal membranes of rat cerebral cortex. Benzodiazepine receptor occupancies were calculated from the displacing potencies of the ligands determined for [3H]flumazenil binding under identical conditions. Greater maximal enhancing (i.e. accelerating) effects of the full agonists diazepam and flunitrazepam on [35S]TBPS binding were reached at lower occupancies of benzodiazepine receptors than the smaller enhancing effects of the partial agonists bretazenil and the beta-carboline ZK 91296. Similarly, the maximal decreasing effect of the full inverse agonist methyl 6,7-dimethoxy-4-ethyl-beta-carboline-3-carboxylate (DMCM) on TBPS binding was reached at lower occupancy than that of the partial inverse agonist FG 7142. Half-maximal effects on TBPS binding corresponded to about 20-30% occupancies for the full agonists and DMCM, while for partial agonists and FG 7142 they exceeded 60-80% occupancies. Different (concave versus convex) shapes of the occupancy-effect curves can also differentiate partial from full agonists and inverse agonists. The results suggest that different pharmacological efficacies of benzodiazepine receptor ligands are associated with differences in coupling between benzodiazepine and convulsant binding sites to modulate the chloride ionophores.

Kinetic cooperative effect of glycine receptor agonists and antagonists on the dissociation of strychnine binding.

The dissociation of [3H]strychnine binding was studied in synaptosomal membranes of rat spinal cord. Dissociation elicited by 100-fold dilution was accelerated by completely displacing concentrations of glycinergic agents. The rank order of acceleration was iso-THAZ < strychnine approximately taurine < R 5135 < avermectin b1a approximately beta-alanine << glycine (THAZ, 5,6,7,8-tetrahydro-4H-isoxazolo-(3,4-d)azepin-3-ol). The accelerating effects were correlated with the changes of entropy (r = 0.93) but not with the changes of free energies (r = 0.06) of their binding. Half-maximal acceleration was elicited by 58 microM glycine. The accelerating effects of glycine and beta-alanine were attenuated by the antagonists.

Modification of benzodiazepine receptors supports the distinctive role of histidine residues.

The effect of selective protein modifying reagents was examined on benzodiazepine (BZ) receptors in synaptosomal membrane preparations of rat whole brain and cerebellum. The potency of diethyl pyrocarbonate, a histidine modifying reagent, to inactivate BZ receptor binding, correlated with the rank order of agonist-inverse agonist efficacies of BZ ligands, the binding of the partial inverse agonist [3H]Ro 15-4513 was inactivated least. Diethyl pyrocarbonate slightly enhanced the displacing potency of Ro 15-4513 and enhanced its binding in low concentrations (1-2 mM). Diazepam-sensitive and -insensitive components of [3H]Ro 15-4513 binding were separated in cerebellum. Diethyl pyrocarbonate inactivated the diazepam-sensitive component with a potency (IC50 = 1.8 mM) similar to that on the binding of other benzodiazepines, while the diazepam-insensitive component was resistant to diethyl pyrocarbonate. Tetranitromethane and 2,3-butanedione (diacetyl), reagents specific for tyrosine and arginine residues respectively, exerted concentration-dependent partial inactivation of [3H]Ro 15-4513 binding. The diazepam-insensitive component of cerebellar Ro 15-4513 binding was more sensitive to inactivation by diacetyl but less sensitive to inactivation by tetranitromethane. These findings are consistent with a distinctive role of histidine-101 in alpha 1, alpha 2, alpha 3 and alpha 5 subunits of the gamma-aminobutyric acidA receptor complex and the His is replaced by an arginine residue in the alpha 6 subunit of the diazepam-insensitive cerebellar benzodiazepine receptors. The only other point of the protein sequence where histidine residues conserved in alpha 1, alpha 2, alpha 3 and alpha 5 subunits are replaced in alpha 6 is tyrosine-214 but this residue does not appear to contribute to benzodiazepine binding.

Central benzodiazepine receptors: in vitro efficacies and potencies of 3-substituted 1,4-benzodiazepine stereoisomers.

3-Acyloxy-, 3-methoxy-, and 3-alkyl-substituted derivatives of the benzodiazepine (BZ) agonist desmethyl-diazepam (DMD) were resolved, and the stereochemical properties of binding to central BZ receptors were investigated in synaptosomal membrane preparations of rat brain. Decreasing potency and stereoselectivity of 3-methyl, 3-ethyl, and 3-isopropyl derivatives in displacement of [3H]diazepam binding can be attributed to differential susceptibilities for steric hindrance of 3-axial versus 3-equatorial substituents of the binding conformation M. Chirality in the alpha-methyl-beta-phenyl-propionic acyl moiety of oxazepam, the 3-OH-derivative of DMD, was noncritical in binding, whereas the beta-phenyl substituent selectively increased the binding of the 3S-stereoisomer. Changing the pH from 7.4 to 5.6 significantly increased the IC50 of (3R)-oxazepam acetate but not those of (3R)-methyl-DMD and diazepam. Binding data led to a steric model of the BZ binding site with the postulation of an additional hydrogen-bond-donating moiety, probably histidine in the "ceiling" of the receptor cavity, that binds the 3-carbonyloxy groups and hinders the 3-alkyl ones. In vitro efficacies of 3-substituted BZs were estimated by allosteric binding interactions within the gamma-aminobutyric acidA (GABAA) receptor-ionophore complex. Non-equilibrium enhancement of t-butyl-bicyclophosphoro[35S]thionate binding by the BZ agonist oxazepam was stereoselectively antagonized by (3S)-oxazepam-(S)-alpha-methyl-beta-phenyl-propionate, suggesting a mixed agonist-antagonist character. GABA enhanced the [3H]diazepam-displacing potencies of the 3S-enantiomers of the acetate, hemisuccinate, and (S)-alpha-methyl-beta-phenyl-propionate esters of oxazepam by a factor of about 1.5-1.6, whereas the GABA shifts for 3R-esters were about 1.2. UV affinity labeling with flunitrazepam resulted in a significantly smaller decrease in the displacing potency of (3R)-oxazepam acetate than in that of the 3S-enantiomer. GABA shifts of successively 3-methylated DMD derivatives were also compared. The GABA shifts of DMD and its (3S)-methyl and 3,3-dimethyl derivatives were all characteristic of full agonists (2.4-2.7), whereas that of (3R)-methyl-DMD was 1.5. The 3-methoxy enantiomers of DMD displayed stereoselectivity and GABA shift values intermediate between those of 3-methyl and 3-acetoxy derivatives. These allosteric interactions suggest that 3-carbonyloxy derivatives in general, as well as (3R)-BZ enantiomers bound with axial 3-alkyl and 3-alkyloxy groups, decrease the agonist efficacies of 1,4-BZs to modulate the GABAA receptor complex.