The Peripheral-Type Benzodiazepine Receptor Is Present in Astrocytes but Is Not a Primary Site of Action for Convulsants/Anticonvulsants

Abstract: High-affinity binding sites for [³H]PK 11195 and [³H]Ro 5-4864 with the properties of the peripheral-type benzodiazepine receptor were detected in primary cultures of both mouse neocortical and cerebellar astrocytes. The binding sites were enriched in mitochondrial fractions on differential centrifugation. An 18-kDa polypeptide was specifically photolabelled in cerebellar astrocytes by [³H]-PK 14105, a photolabel for the peripheral-type benzodiazepine receptor. However, this polypeptide did not show any reactivity with an antiserum previously raised against the corresponding polypeptide from rat adrenal gland. Various anticonvulsant and convulsant agents were tested for their ability and potency at inhibiting [³H]Ro 5-4864 binding to neocortical astrocytes. Many of these compounds, previously reported to be inhibitors of diazepam binding to neocortical astrocytes, proved ineffective in this study. No correlation was observed between convulsant/anticonvulsant potency and ability to inhibit [³H]Ro 5-4864 binding to the peripheral-type benzodiazepine receptor in these cells. Thus, whereas some convulsants and anticonvulsants might interact with this astrocytic receptor, such a system has no validity as a general screening method for these agents.     

Differential mediation of the anticonvulsant effects of carbamazepine and diazepam

Possible mechanisms of action of carbamazepine and diazepam on amygdala-kindled seizures were studied using compounds acting at the central and "peripheral-type" benzodiazepine binding sites. Ro-15-1788, a selective antagonist at the central benzodiazepine site, blocked the anticonvulsant effect of diazepam, but not of carbamazepine. In contrast, Ro5-4864, which acts at the "peripheral-type" benzodiazepine site, blocked the anticonvulsant effect of carbamazepine, but not of diazepam. The effect of Ro5-4864 was itself reversed by PK-11195, a compound that displaces Ro5-4864 binding in vitro and in vivo. These data indicate that the anticonvulsant effects of carbamazepine and diazepam on amygdala-kindled seizures are differentially mediated and suggest that the "peripheral-type" benzodiazepine binding site is functionally involved in the anticonvulsant effect of carbamazepine.     

Flunitrazepam Binding to Intact and Homogenized Astrocytes and Neurons in Primary Cultures

[3H]Flunitrazepam binds to intact and homogenized mouse astrocytes and neurons in primary cultures. In intact cells, the binding is to a single, high-affinity, saturable population of benzodiazepine binding sites with a KD of 7 nM and Bmax of 6,033 fmol/mg protein in astrocytic cells and a KD of 5 nM and Bmax of 924 fmol/mg protein in neurons. After homogenization, the Bmax values decrease drastically in both cell types, but most in astrocytes. The temperature and time dependency are different for the two cell types, with a faster association and dissociation in astrocytes than in neurons and a greater temperature sensitivity in the astrocytes. Moreover, flunitrazepam binding sites on neuronal and astrocytic cells have different pharmacological profiles. In intact astrocytic cells, Ro 5-4864 (Ki = 4 nM) is the most potent displacing compound, followed by diazepam (Ki = 6 nM) and clonazepam (Ki = 600 nM). In intact neurons, the relative order of potency of these three compounds is different: diazepam (Ki = 7 nM) is the most potent, followed by clonazepam (Ki = 26 nM) and Ro 5-4864, which has little effect. After homogenization the potency of diazepam decreases. We conclude that both neuronal and astrocytic cells possess high-affinity [3H]flunitrazepam binding sites. The pharmacological profile and kinetic characteristics differ between the two cell types and are further altered by homogenization.     

Effects of benzodiazepines on single unit activity in the substantia nigra pars reticulata

Intravenous administration of two benzodiazepines, flurazepam and diazepam, had an inhibitory effect on the firing rates of neurons of the substantia nigra pars reticulata, a brain region with an identified GABAergic innervation. Diazepam was more potent than flurazepam. Bicuculline and picrotoxin, two drugs which block GABAergic transmission, and caffeine and theophylline, two methylxanthines which inhibit benzodiazepine binding, all reversed the inhibition produced by diazepam. The action of theophylline was less consistent than that of caffeine. Similarly, Ro 15–1788, an imidazodiazepine which putatively functions as a specific benzodiazepine antagonist, reversed the diazepam-induced inhibition. These findings are consistent with previous reports which suggest that the benzodiazepines may act through a GABAergic mechanism. In a separate group of experiments, caffeine or Ro 15–1788 was administered alone. While caffeine excited all reticulata cells tested. Ro 15–1788, the more specific benzodiazepine antagonist, generally had little excitatory effect. These results suggest: 1) that cells of the substantia nigra pars reticulata may not receive a substantial, tonic inhibition mediated by an endogenous benzodiazepine-like substance; and 2) that the methylxanthines may increase reticulata cell firing, at least in part, through mechanisms unrelated to the blockade of benzodiazepine receptors.     

Modulation of benzodiazepine by lysine and pipecolic acid on pentylenetetrazol-induced seizures

L-lysine, an essential amino acid for man and animals, and its metabolite pipecolic acid (PA) have been studied for their effects on pentylenetetrazol (PTZ)-induced seizures in mice. L-Lysine or L-PA i.p. significantly increased clonic and tonic latencies in a dose-dependent manner against 90 mg/kg PTZ-induced seizures. L-Lysine but not L-PA enhanced the anticonvulsant effect of diazepam (DZ) (0.2 mg/kg). L-PA (0.1 mmol/kg) i.c.v. showed a slight decrease in clonic latency; it did not enhance the antiseizure activity of DZ; it caused seizures at 0.6 mmol/kg. D-PA (0.1 mmol/kg) i.c.v. displayed an opposite effect compared to its L-isomer. The anticonvulsant effect of L-lysine in terms of increase in seizure latency and survival was even more amplified when tested with a submaximal PTZ concentration (65 mg/kg). L-Lysine showed an enhancement of specific 3H-flunitrazepam (FZ) binding to mouse brain membranes both in vitro and in vivo. The possibility of L-lysine acting as a modulator for the GABA/benzodiazepine receptors was demonstrated. Since L-PA showed enhancement of 3H-FZ binding only in vitro but not in vivo, the anticonvulsant effect of L-PA may not be linked to the GABA/benzodiazepine receptor.     

Putative endogenous ligands for the benzodiazepine receptor.

The recent discovery of pharmacologically relevant, high affinity, stereospecific binding sites for the benzodiazepines in the central nervous system (CNS) has rekindled investigations concerning the mechanism of action of these drugs. It has become increasingly clear that elucidation of benzodiazepine action will provide new and important insights into the neurochemical substances of seizure activity, centrally mediated muscle relaxation and anxiety, three major actions of this class of drugs.The existence of a functional receptor for the benzodiazepines, compounds not present $\text{in vivo}$, suggests that endogenous substances exist that serve as natural substrates for this receptor. Furthermore, the characterization of endogenous benzodiazepine receptor ligands affords an opportunity to determine the neurochemical mechanisms underlying the pharmacologic and behavioral effects manifested by the benzodiazepines.Using receptor binding methodology to assay tissue extracts for [³H] diazepam binding inhibitory activity, putative endogenous ligands for the benzodiazepine receptor have been isolated and identified as the purine nucleosides. Compounds such as inosine and hypoxanthine exhibit competitive inhibition of [³H] diazepam binding. The low affinity purinergic inhibition of diazepam binding is consistent with their $\text{in vivo}$ concentrations. Distinct structure-activity relationships exist for the purines with subtle structural alterations having marked effects on diazepam binding inhibitory potency. The methylxanthine stimulants, caffeine, theophylline, and theobromine, also competitively inhibit diazepam binding, suggesting that some of their actions may be mediated by the benzodiazepine receptor.The purines also have “benzodiazepine-like” pharmacologic properties, since they have been shown to antagonize pentylenetetrazol induced seizures in mice in a dose dependent manner. Neurophysiologic studies have also shown that iontophoresis of inosine on cultured mouse primary neurons produce neurotransmitter like effects. Furthermore, these effects are similar to those observed with flurazepam, a finding that provides additional evidence for the “benzodiazepine-like” properties of the purines.The preliminary studies outlined below indicate that the purines are good candidates as putative endogenous ligands for the benzodiazepine receptor and provide a foundation for future studies that concern the homeostatic mediation of seizure activity and anxiety.     

RO5-4864 inhibits the binding of [35S]t-butylbicyclophosphorothionate to rat brain membranes

RO5-4864, a 1,4-benzodiazepine, has recently been shown to possess anticonvulsant, convulsant and anxiogenic properties and to inhibit Ca++-calmodulin-stimulated membrane phosphorylation. RO5-4864 inhibited the binding of [35S]t-butylbicyclophosphorothionate (TBPT) to cerebral cortex, cerebellar and hippocampus membranes, with an IC50 value of approximately 20 microM. TBPT binds apparently to the picrotoxinin site of the benzodiazepine-GABA receptor-ionophore complex and appears to be a site of action for several classes of convulsant, depressant and anxiolytic drugs that modulate GABAergic transmission. RO5-4864 inhibited [35S] TBPT binding in cerebral cortex, apparently competitively. Antagonists of GABA and central benzodiazepine sites did not interfere with the ability of RO5-4864 to inhibit [35S] TBPT binding. The properties of RO5-4864 to inhibit TBPT binding are similar to other convulsants and GABA antagonists (except bicuculline) which inhibit TBPT binding. These results suggest that RO5-4864 interacts with the TBPT binding sites of the oligomeric GABA receptor complex.     

Benzodiazepine Binding Characteristics of Embryonic Rat Brain Neurons Grown in Culture

The binding of [3H]diazepam to cell homogenates of embryonic rat brain neurons grown in culture was examined. Under the conditions used to prepare and maintain these neurons, only a single, saturable, high-affinity binding site was observed. The binding of [3H]diazepam was potently inhibited by the CNS-specific benzodiazepine clonazepam (Ki = 0.56 +/- 0.08 nM) but was not affected by the peripheral-type receptor ligand Ro5-4864. The KD for [3H]diazepam bound specifically to cell homogenates was 2.64 +/- 0.24 nM, and the Bmax was 952 +/- 43 fmol/mg of protein. [3H]Diazepam binding to cell membranes washed three times was stimulated dose-dependently by gamma-aminobutyric acid (GABA), reaching 112 +/- 7.5% above control values at 10(-4) M. The rank order for potency of drug binding to the benzodiazepine receptor site in cultured neurons was clonazepam greater than diazepam greater than beta-carboline-3-carboxylate ethyl ester greater than Ro15-1788 greater than CL218,872 much greater than Ro5-4864. The binding characteristics of this site are very similar to those of the Type II benzodiazepine receptors present in rat brain. These data demonstrate that part, if not all, of the benzodiazepine-GABA-chloride ionophore receptor complex is being expressed by cultured embryonic rat brain neurons in the absence of accompanying glial cells and suggest that these cultures may serve as a model system for the study of Type II benzodiazepine receptor function.     

Brain benzodiazepine levels following intravenous administration of [34]diazepam: relationship to the potentiation of purinergic depression of central nervous system neurons

Levels of [3H]benzodiazepine were measured in rat cerebral cortex following intravenous injection of [3H]diazepam using a dose and time schedule reported to elicit a marked potentiation of the depressant effects of iontophoretically applied 5'-AMP to rat cerebral cortical neurons. The levels of [3H]benzodiazepine obtained strongly suggest (i) that blockade of adenosine uptake as a mechanism for this potentiation is not consistent with the potency of diazepam as an inhibitor of adenosine uptake in vitro, and (ii) that a potentiative interaction of adenosine and diazepam may reflect the binding of these compounds to benzodiazepine receptors.     

Evidence for the pharmacological relevance of benzodiazepine receptors to anticonvulsant activity

Each of a series of benzodiazepines was found to be effective in preventing convulsions evoked by intermittent photic stimulation of epileptic chickens. There was a high correlation between the anticonvulsant potencies (mean effective dosages) and the affinity of the agents for the putative benzodiazepine receptor as measured by displacement of [3H]diazepam from binding sites on chicken synaptosomal membranes. This correlation in a genetic model of epilepsy provides further evidence that benzodiazepines exert their anticonvulsant effects by interacting with the benzodiazepine receptor.     

Effects of denzimol on benzodiazepine receptors in the CNS: Relationship between the enhancement of diazepam activity and benzodiazepine binding sites induced by denzimol in the rat

Denzimol, a new anticonvulsant drug with a pharmacological profile similar to that of phenytoin, enhances the ataxic and antimetrazol activity of diazepam in rats without affecting its activity against picrotoxin-induced seizures. $\text{In vivo}$ and $\text{ex vivo}$ denzimol enhances the binding of ³H-flunitrazepam in cortex and in hippocampus but not in cerebellum.The possibility of this increase in the number of benzodiazepine binding sites contributing in some way to enhancement of the depressive and anticonvulsant activity of diazepam is discussed.     

Convulsant/depressant site of action at the allosteric benzodiazepine-GABA receptor-ionophore complex

Several classes of centrally acting convulsant, depressant, anticonvulsant and anxiolytic drugs modulate GABAergic transmission. The postsynaptic receptor with which these drugs interact is an allosteric complex with distinct binding sites for GABA, benzodiazepines, picrotoxinin and related compounds. Convulsants which inhibit GABA transmission (except bicuculline) inhibit competitively the binding of dihydropicrotoxinin (DHP) or t-butylbicyclophosphorothionate (TBPT) to the picrotoxinin site and prevent the allosteric enhancing effect of depressant drugs on GABA and benzodiazepine binding. Depressant drugs give a mixed inhibition of TBPT binding. The possible topography of the picrotoxinin site and its relationship to convulsant/depressant drug action at the benzodiazepine-GABA receptor-ionophore complex is discussed.     

Avermectin Bla Modulation of γ-Aminobutyric Acid/Benzodiazepine Receptor Binding in Mammalian Brain

The anthelminthic natural product avermectin B1a (AVM) modulates the binding of gamma-aminobutyric acid (GABA) and benzodiazepine (BZ) receptor ligands to membrane homogenates of mammalian brain. The potent (EC50 = 40 nM) enhancement by AVM of [3H]diazepam binding to rat or bovine brain membranes resembled that of barbiturates and pyrazolopyridines in being inhibited (partially) by the convulsants picrotoxin, bicuculline, and strychnine, and by the anticonvulsants phenobarbital and chlormethiazole. The maximal effect of AVM was not increased by pentobarbital or etazolate. However, AVM affected BZ receptor subpopulations or conformational states in a manner different from pentobarbital. Further, unlike pentobarbital and etazolate, AVM did not inhibit allosterically the binding of the BZ receptor inverse agonist [3H]beta-carboline-3-carboxylate methyl ester, nor did it inhibit, but rather enhanced, the binding of the cage convulsant [35S]t-butyl bicyclophosphorothionate to picrotoxin receptor sites. AVM at submicromolar concentrations had the opposite effect of pentobarbital and etazolate on GABA receptor binding, decreasing by half the high-affinity binding of [3H]GABA and related agonist ligands, and increasing by over twofold the binding of the antagonist [3H]bicuculline methochloride, an effect that was potentiated by picrotoxin. AVM also reversed the enhancement of GABA agonists and inhibition of GABA antagonist binding by barbiturates and pyrazolopyridines. These overall effects of AVM are unique and require the presence of another separate drug receptor site on the GABA/BZ receptor complex.     

Actions of pentobarbitone and derivatives with modified 5-butyl substituents on GABA and diazepam binding to rat brain synaptosomal membranes

The effects of a variety of factors known to influence the enhancement of GABA binding by diazepam, were studied upon pentobarbitone stimulation of GABA binding to washed synaptosomal membranes prepared from whole rat brains. The differential kinetics of, and effects of temperature, chloride ions, a benzodiazepine receptor antagonist (Ro15-1788) and picrotoxinin upon pentobarbitone and diazepam enhancement of GABA binding, suggest that these drugs exert their actions upon GABA binding at different loci. The degree of enhancement of diazepam binding and of high affinity GABA binding in chloride-containing media at 25 degrees C by members of a series of twelve side chain methyl substituted and/or unsaturated derivatives of 5-butyl-5-ethyl-barbituric acid (pentobarbitone analogs) correlated significantly. For the sedative members of the series, enhancement of high affinity GABA binding correlated with their anaesthetic but not their anticonvulsant activities. It appears likely that the anaesthetic and anticonvulsant activities of barbiturates arise from different molecular actions.     

Benzodiazepine Binding to Cultured Human Pituitary Cells

Benzodiazepine receptors were investigated in a cell line of human pituitary cells (18-54,SF) grown in serum-free medium. Preparations of 18-54,SF whole cells and cell membranes were shown to possess saturable [3H]diazepam binding sites. Membrane sites were found to have a KD of 20 nM for diazepam while whole cells possessed a twofold higher value. The KD values determined from Rosenthal, Hill, and kinetic analyses were consistent for each preparation. Whole-cell binding of [3H]diazepam was observed to be more stable than binding to membranes at higher temperatures (37 degrees C) and when longer incubation times (60 min) were employed at 4 degrees C. The rank order potency of various benzodiazepines to inhibit [3H]diazepam binding to whole cells and membranes was Ro 5-4864, flunitrazepam, diazepam, and clonazepam. Representatives of other drug classes did not inhibit this benzodiazepine binding. When 18-54,SF cells were grown for 24 h with 100 nM diazepam and then extensively washed membranes prepared, the KD for diazepam increased to 38 nM whereas the Bmax was unchanged when compared with untreated controls. Overall, these findings indicate that pituitary cells possess a peripheral-type benzodiazepine receptor and that the whole cell receptor differs quantitatively when compared with the membrane receptor.     

GABA-benzodiazepine interactions: physiological, pharmacological and developmental aspects

Many of the pharmacological actions of the benzodiazepines can be attributed to their actions on gamma-aminobutyric acid (GABA) systms in the brain. Electrophysiological studies on dorsal raphe neurons indicate that the benzodiazepines act postsynaptically to potentiate GABAergic inhibition in this midbrain nucleus. Direct binding studies have shown that both in vitro and in vivo binding of [3H]diazepam to a specific high affinity benzodiazepine binding site in cerebral cortical tissue are enhanced by the direct in vitro addition of GABA and GABA agonists or by pretreatment of animals with GABA analogs and agents that elevate GABA levels in brain. Ontogenic development of [3H]diazepam binding in brain parallels the development of the sodium-independent [3H]GABA binding. The ability of GABA to enhance benzodiazepine binding is present throughout development and inversely related to age. These data suggest that there is a functionally significant interaction between the benzodiazepines and GABA throughout development and at maturity. A model is proposed to relate these interactions to conformational changes in a benzodiazepine/GABA/Cl- ionophore complex.     

beta-Carboline-3-carboxylate-t-butyl ester: a selective BZ1 benzodiazepine receptor antagonist

The effectiveness of beta-carboline-3-carboxylate-t-butyl ester (beta CCtB) in antagonizing the anticonvulsant, ataxic and antipunishment effects of diazepam were evaluated. In mice, beta CCtB at doses of 3 and 10 mg/kg produced a dose-related antagonism of the anticonvulsant effects of diazepam against pentylenetetrazole (80 mg/kg). A dose of 30 mg/kg of beta CCtB did not produce a further shift in the diazepam dose-effect curve, apparently because beta CCtB failed to block the muscle-relaxant effects of diazepam. Further, beta CCtB (30 mg/kg) failed to antagonize the ataxic effects of diazepam in an inverted screen test. Rats responded under a multiple schedule where in one component every twentieth response (FR20) resulted in water presentation (unpunished component) and in another component every twentieth response (FR20) resulted in both shock and water presentation (punished component). Diazepam p.o. (0.1 to 10 mg/kg) first increased and then decreased rates in the punished component but only decreased rates in the unpunished component. beta CCtB had no effect on response rates when administered alone, but antagonized the rate-increasing effects of diazepam in the punished component. beta CCtB did not alter the rate-decreasing effects of diazepam in either component. Thus, beta CCtB selectively antagonized the effects of diazepam on punished behavior as well as the anticonvulsant effects of diazepam, but beta CCtB failed to antagonize the rate-decreasing and ataxic effects of diazepam. These results are consistent with the interpretation that beta CCtB is a selective BZ1 benzodiazepine receptor antagonist.     

Effect of some potent adenosine uptake inhibitors on benzodiazepine binding in the CNS

A series of nucleoside transport inhibitors has been tested for their ability to displace [³H]diazepam binding to CNS membranes. No correlation between their potency as [³H]adenosine uptake blockers and as inhibitors of [³H]diazepam binding was found, either in rat or guinea-pig brain tissue. Dipyridamole, a potent adenosine transport inhibitor interacted strongly (K_i = 54 nM) with peripheral-type benzodiazepine binding sites (“acceptor sites”) and was 4–5 fold weaker in displacing [³H]methylclonazepam and [³H]Ro15-1788, ligands selective for the specific central benzodiazepine “receptor”. Unlike the benzodiazepines, dipyridamole had no anticonvulsant action against metrazole-induced convulsions in mice. Ro5-4864, a benzodiazepine which selectively interacts with the peripheral-type benzodiazepine binding site, was approximately equipotent with diazepam in inhibiting [³H]adenosine uptake in brain tissue. These results do not support the idea of a very close link between high-affinity central binding sites for clinically-active benzodiazepines and the adenosine uptake site. The possibility of a connection between benzodiazepine “acceptor” sites and the membrane nucleoside transporter is discussed.     

"Peripheral-Type"Binding Sites for Benzodiazepines in Brain: Relationship to the Convulsant Actions of Ro 5–4864

Previous studies have shown that Ro 5-4864 is a potent convulsant and increases the firing rate of substantia nigra zona reticulata neurons. The pharmacologic profile of compounds that antagonize these actions suggested that the effects of Ro 5-4864 were not mediated by "brain-type" benzodiazepine receptors. We examined a number of compounds that are structurally related to Ro 5-4864 for their capacities to displace [3H]Ro 5-4864 from "peripheral-type" binding sites and their potencies as convulsants (or as antagonists of Ro 5-4864-induced convulsions). It was observed that compounds such as KW 3600 (the N-desmethyl analog of Ro 5-4864), which have very low affinities for "peripheral-type" sites, are convulsants with a potency nearly equal to that of Ro 5-4864. In contrast, compounds such as Ro 5-6900 and PK 11195, which bind with very high affinities to "peripheral-type" binding sites, are neither convulsants nor do they antagonize the convulsant actions of Ro 5-4864. Within a series of compounds that are structurally related to Ro 5-4864 there is a good correlation (r = 0.93; p less than 0.01) between their potencies as convulsants and their capacities to displace [35S]t-butylbicyclophosphorothionate from sites that may be associated with the chloride ionophore. Thus, it appears that occupation of "peripheral-type" binding sites by high-affinity ligands may not be directly involved in the convulsant actions of Ro 5-4864 and related compounds.     

The F-loop of the GABA A receptor gamma2 subunit contributes to benzodiazepine modulation

GABA(A) receptors can be modulated by benzodiazepines, although these compounds do not directly activate or inhibit the receptors. The prototypic benzodiazepine, diazepam, potentiates responses to GABA in GABA(A) receptors that contain a gamma subunit. Here we have used mutagenesis, radioligand binding, voltage clamp electrophysiology, and homology modeling to probe the role of the F-loop residues Asp(192)-Arg(197) in the GABA(A) receptor gamma(2) subunit in diazepam potentiation of the GABA response. Substitution of all of these residues with Ala and/or a residue with similar chemical properties to the wild type residue decreased the level of diazepam potentiation, and one mutation (D192A) resulted in its complete ablation. None of the mutations changed the GABA EC(50) or the [(3)H]flumazenil binding affinity, suggesting they do not affect GABA or benzodiazepine binding characteristics; we therefore propose that they are involved in the diazepam-mediated conformational change that results in an increased response to GABA. Homology models of the receptor binding pocket in agonist-bound and unbound states suggest that the F-loop is flexible and has different orientations in the two states. Considering our data in relation to these models, we find that the F-loop residues could contribute to hydrogen bond networks and hydrophobic interactions with neighboring residues that change during receptor activation.