Identification of an endogenous peptide-ligand for the benzodiazepine receptor

The pharmacology of human anxiety

The human pharmacology of anxiety disorders, including panic disorder, is detailed. The major theories center around the role of benzodiazepine receptor, noradrenergic and serotonergic dysfunction. The contribution that challenge tests with lactate, hyper- and hypocapnia, beta- and alpha-2-adrenoceptor agonists, peptides, pentylenetetrazol, and caffeine make to our understanding of the biological basis of anxiety and these major theories are described and discussed.

GABA-B receptor activation and conflict behaviour

Baclofen and oxazepam enhance extinction of conflict behaviour in the Geller-Seifter test while baclofen and diazepam release punished behaviour in Vogel's conflict test. In order to investigate the possibility that the effect of the selective GABA-B receptor agonist baclofen is mediated indirectly via the GABA-A/benzodiazepine receptor complex, the effect of pretreatment of rats with baclofen on [3H]-diazepam binding to washed and unwashed cortical and cerebellar membranes of rats has been studied. Baclofen pretreatment increased Bmax in washed cerebellar membranes when bicuculline was present in the incubation mixture. No effect was seen in cortical membranes. The present results render it unlikely that the effect of baclofen on extinction of conflict behaviour and punished drinking is mediated via the GABA-A/benzodiazepine receptor complex.

Demonstration and purification of an endogenous benzodiazepine from the mammalian brain with a monoclonal antibody to benzodiazepines

Four hybridoma lines secreting monoclonal antibodies to benzodiazepines were produced after BALB/c mice were immunized with a benzodiazepine-bovine serum albumin conjugate. The monoclonal antibodies were purified from ascites fluids, and their binding affinities for benzodiazepines and other benzodiazepine receptor ligands were determined. These antibodies have very high binding affinities for diazepam, flunitrazepam, Ro5-4864, Ro5-3453, Ro11-6896, and Ro5-3438 (the Kd values are in the 10(-9) M range). However, these antibodies have very low affinities for the benzodiazepine receptor inverse agonists (beta-carbolines) and antagonists (Ro15-1788 and CGS-8216). One of the monoclonal antibodies (21-7F9) has been used to demonstrate the existence of benzodiazepine-like molecules in the brain and for the purification of these molecules. Immunocytochemical experiments show that these molecules are neuronal and not glial and that they are ubiquitously distributed throughout the brain. Immunoblots indicate the presence of benzodiazepine-like epitopes in several brain peptides. An endogenous substance that binds to the central-type benzodiazepine receptor with agonist properties has been purified to homogeneity from the bovine brain. The purification consisted on immunoaffinity chromatography on immobilized monoclonal anti-benzodiazepine antibody followed by gel filtration on Sephadex G-25 and two reverse phase HPLCs. The purified substance has a small molecular weight and its activity is protease resistant. The endogenous substance blocks the binding of agonists, inverse agonists and antagonists to the central-type benzodiazepine receptor but it does not inhibit the binding of Ro5-4864 to the peripheral-type benzodiazepine receptor. The neurotransmitter gamma-aminobutyric acid increases the affinity of the benzodiazepine receptor for the purified substance. Preliminary evidence indicates that the purified substance is a benzodiazepine with a molecular structure that is identical or very close to N-desmethyldiazepam.

Isolation and identification in bovine cerebral cortex of n-butyl beta-carboline-3-carboxylate, a potent benzodiazepine binding inhibitor

A substance having benzodiazepine-binding inhibitory activity has been extracted from 18 kg of gray matter of bovine cerebral cortex and purified to homogeneity. This substance inhibits competitively [3H]flunitrazepam and ethyl beta-[3H]carboline-3-carboxylate binding with high affinity (Ki, 3 nM), but it is inactive upon 3H-labeled Ro 5-4864, [3H]quinuclidinyl benzylate, [3H]prazosin, [3H]clonidine, [3H]dihydroalprenolol, and upon high-affinity [3H]muscimol binding. This inhibitor has been identified as n-butyl beta-carboline-3-carboxylate (beta-CCB) by fast atom bombardment mass spectroscopy (Mr, 268) and electron bombardment fragmentography, ultraviolet and fluorescence spectra, coelution in HPLC with standard beta-CCB, and by the exact correspondence in Ki with beta-CCB on the displacement of [3H]flunitrazepam binding. The possible artificial formation of beta-CCB has been discarded by a series of control experiments including addition of tryptophan to the starting homogenate, extraction from liver, isolation and purification by an alternative procedure avoiding organic solvents, and by the impossibility of making beta-CCB from beta-carboline-3-carboxylic acid or its methyl ester in the conditions of our extraction and purification procedures.

Purification and characterization of tribulin, and endogenous inhibitor of monoamine oxidase and of benzodiazepine receptor binding

A low molecular weight fraction of human urine (less than 500 daltons) which both inhibits monoamine oxidase and benzodiazepine binding to central and peripheral receptors has been purified by ethyl acetate extractions, HPLC and thin layer chromatography. This material extracted equally well at acid and basic pH and was insoluble in heptane. It competitively inhibited binding of 3H-clonazepam, a central benzodiazepine receptor agonist and, in addition, displaced 3H-Ro 5-4864, a specific peripheral benzodiazepine receptor ligand, from its binding sites. It showed no GABA shift with the benzodiazepine receptor antagonist, Ro-15 1788. MAO A and B were inhibited approximately equipotently and the material competitively inhibited tyramine oxidation by rat liver. It was stable on boiling and is unlikely to be a peptide.

Demonstration of benzodiazepine-like molecules in the mammalian brain with a monoclonal antibody to benzodiazepines

An anti-benzodiazepine monoclonal antibody has been used to demonstrate the existence of benzodiazepine-like molecules in the brain. Immunocytochemical experiments show that these molecules are neuronal and not glial and that they are ubiquitously distributed throughout the brain. Immunoblots indicate the presence of benzodiazepine-like epitopes in several brain peptides. Small benzodiazepine-like molecules were isolated from the brain soluble fraction by immunoaffinity chromatography. They block the binding of agonists, inverse agonists, and antagonists to the neuronal-type benzodiazepine receptor. The neurotransmitter gamma-aminobutyric acid increases the affinity of the benzodiazepine receptor for the purified endogenous molecules. The results indicate that the immunoaffinity-purified molecules behave like the neuronal-type benzodiazepine receptor agonists. The purified molecules, however, do not inhibit the binding of tritiated Ro 5-4864 to the "peripheral-type" benzodiazepine receptor. The results demonstrate the existence of benzodiazepine-like molecules in the brain that bind to the benzodiazepine receptor. These molecules are different from the endogenous benzodiazepine receptor ligands reported by others.

Benzodiazepines modulate the A2 adenosine binding sites on 108CC15 neuroblastoma X glioma hybrid cells

We have demonstrated high affinity diazepam binding sites of the Ro5-4864 benzodiazepine receptor subtype on 108CC15 neuroblastoma X glioma hybrid cells. These cells were previously shown to have purinoceptors of the A2 adenosine subtype and we have now found that [3H]-adenosine can be displaced from this binding site by the benzodiazepines and related compounds that can also bind to the Ro5-4864 site. Diazepam was found to have no intrinsic activity at the A2-receptor as measured by the stimulation of adenosine 3':5'-cyclic monophosphate (cyclic AMP) production in this cell line. At concentrations sufficient to compete for the A2-receptor, diazepam was shown to facilitate, by approximately 2 fold, the stimulation of cyclic AMP by adenosine. These effects are not due to inhibition of adenosine uptake or phosphodiesterase activity, but are probably a consequence of modulation of the coupling of the A2-receptor to cyclic AMP production in this hybrid cell line.

Endogenous benzodiazepine ligands in human cerebrospinal fluid

Human cerebrospinal fluid was chromatographed on Bio-Gel P-4. Fractions containing material with molecular weights less than 4000 Dalton were pooled and further fractionated by high pressure liquid chromatography on an UltroPack TSK column G 2000 SW. At least three peaks, which were free of salt and GABA, were shown to displace (3H)-diazepam in the receptor-binding assay. Two of these peaks inhibited diazepam-binding competitively as shown by Lineweaver-Burke and displacement analysis. Their activity could be enhanced by the addition of GABA to the assay mixture. Incubation of these two peaks with various enzymes indicated that at least part of the activity of the second peak is due to a peptide.

Molecular aspects of the action of benzodiazepine and non-benzodiazepine anxiolytics: a hypothetical allosteric model of the benzodiazepine receptor complex

The availability of radiolabeled benzodiazepines has resulted in the identification of high affinity receptors in the central nervous system for this class of psychotherapeutic agent which are linked to recognition sites for the inhibitory neurotransmitter, GABA. Evaluation of new, synthetic compounds in the benzodiazepine radioligand binding assay has resulted in the identification of nine classes of non-benzodiazepine putative anxiolytic agents, some of which may be more anxioselective than the benzodiazepines. At least three and possibly five subclasses of benzodiazepine receptor have been identified in mammalian tissues using radioligand binding assays. The possibility exists that one of these receptor subclasses may mediate the anxiolytic effects of the benzodiazepines while the remainder may be involved in the mediation of the sedative, ataxic and anticonvulsant properties associated with benzodiazepine-like agents. Several endogenous ligands for the benzodiazepine receptor(s) have been postulated. These include various proteins and peptides, purines and the beta-carbolines. This latter group, which competitively antagonizes the pharmacological and biochemical effects of the benzodiazepines, has the highest affinity for the benzodiazepine receptor of all compounds thus far examined; however, none of these compounds has been conclusively identified as the endogenous ligand akin to the enkephalins and endorphins at the opiate receptor. The majority of available evidence would indicate that the endogenous ligand for the benzodiazepine receptor(s) is an antagonist of the benzodiazepines and other putative anxiolytic agents.

On a brain polypeptide functioning as a putative effector for the recognition sites of benzodiazepine and beta-carboline derivatives

Stimulation of benzodiazepine recognition sites by various ligands can elicit opposite types of responses such as proconvulsant and anticonvulsant or proconflict and anticonflict actions. The study of the pharmacological profile of various ligands makes it possible to distinguish three classes of compounds: (1) those that elicit anticonflict responses in the Vogel test, inhibit the convulsions due to an impairment of GABAergic transmission and increase the Bmax of the high affinity recognition site for GABA; (2) those that displace benzodiazepines from specific binding sites, facilitate convulsions due to impairment of GABAergic mechanisms, elicit proconflict responses in Vogel's test and inhibit the facilitation by benzodiazepines of GABA binding and (3) those that displace benzodiazepines and beta-carboline-derivatives from specific binding sites, antagonize the anticonflict, the proconflict, the anticonvulsant and the proconvulsant actions of the two preceding groups of substances and in very large doses elicit a small proconvulsant action. Examples of the latter are an imidazobenzodiazepine (RO 15-1788) and a pyrazolquinolinone derivative (CGS 8216). The nomenclature for these three classes of drugs should be kept flexible until the action of the endogenous ligand that functions as the physiological effector of the benzodiazepine/beta-carboline recognition site is known. A putative ligand for this site (DBI = diazepam binding inhibitor) has been isolated and purified to homogeneity. It includes 104 amino acid residues, the sequence of the last 45 amino acids has been determined. This compound elicits a proconflict action, displaces beta-carboline derivatives more than anxiolytic benzodiazepines, but its high molecular weight and relative low affinity for the binding sites in brain might suggest that it is a precursor, rather than the putative effector for benzodiazepine and/or beta-carboline recognition sites.

New endogenous benzodiazepine receptor ligand in human urine: identity with endogenous monoamine oxidase inhibitor?

Normal human urine contains both monoamine oxidase-inhibiting and benzodiazepine receptor-binding material. Each was extracted into ethyl acetate at pH 1 and subjected to high performance liquid chromatography: they ran similarly, showing three major peaks. The correlation coefficient between the pattern of MAO inhibition and inhibition of 3H-flunitrazepam binding to benzodiazepine receptors in the second half of the elution process was 0.78 (p less than 0.001): most UV-absorbing material present was eluted earlier in the run. These results are compatible with, although they do not prove, the hypothesis that the endogenous MAO inhibitor, previously shown to be increased in stress, is also an endogenous inhibitor of 3H-flunitrazepam binding to the benzodiazepine receptor. This material is different from other putative endogenous ligands: it migrates more rapidly than the potent but artefactual beta-carboline-3-carboxylic acid ethyl ester previously isolated from human urine; nor can the effect we have identified derive from harmane, inosine, hypoxanthine or nicotinamide which fail to extract into ethyl acetate at pH 1.

Isolation, characterization, and purification to homogeneity of an endogenous polypeptide with agonistic action on benzodiazepine receptors

A brain polypeptide termed diazepam-binding inhibitor (DBI) and thought to be chemically and functionally related to the endogenous effector of the benzodiazepine recognition site was purified to homogeneity. This peptide gives a single band of protein on NaDodSO4 and acidic urea gel electrophoresis. A single UV-absorbing peak was obtained by HPLC using three different columns and solvent systems. DBI has a molecular mass of approximately equal to 11,000 daltons. Carboxyl-terminus analysis shows that tyrosine is the only residue while the amino-terminus was blocked. Cyanogen bromide treatment of DBI yields three polypeptide fragments, and the sequences of two of them have been determined for a total of 45 amino acids. DBI is a competitive inhibitor for the binding of [3H]diazepam, [3H]flunitrazepam, beta-[3H]carboline propyl esters, and 3H-labeled Ro 15-1788. The Ki for [3H]-diazepam and beta-[3H]carboline binding were 4 and 1 microM, respectively. Doses of DBI that inhibited [3H]diazepam binding by greater than 50% fail to change [3H]etorphine, gamma-amino[3H]butyric acid, [3H]-quinuclidinyl benzilate, [3H]dihydroalprenolol, [3H]adenosine, and [3H]imipramine binding tested at their respective Kd values. DBI injected intraventricularly at doses of 5-10 nmol completely reversed the anticonflict action of diazepam on unpunished drinking and, similar to the anxiety-inducing beta-carboline derivative FG 7142 (beta-carboline-3-carboxylic acid methyl ester), facilitated the shock-induced suppression of drinking by lowering the threshold for this response.

Baseline exploratory activity predicts anxiolytic responsiveness to diazepam in five mouse strains

Inbred mouse strains showing variability in spontaneous exploratory behaviors displayed differential responsiveness to diazepam in an anxiety-related exploration model. C57B1/6J, BALB/cJ, Swiss Webster/NIH, Swiss Webster/Harlan Sprague-Dawley, and CF-1 mice demonstrated a significant correlation between baseline exploratory activity and maximal percentage increase in exploratory behavior induced by diazepam. No correlation was seen between those behavioral responses and the characteristics of brain benzodiazepine binding sites in the different strains. Anti-anxiety responsiveness appears to be a function of some genetically-determined substrate for spontaneous exploratory behaviors which may have multiple neurochemical substrates.

[3H]Diazepam displacing activity in human cerebrospinal fluid

Pooled human cerebrospinal fluid was separated by Sephadex G-50 chromatography. The presence of three peaks, A, B and C, was demonstrated by monitoring absorbance at 254 and 280 nm. All peaks showed [3H]diazepam displacing activity in the membrane receptor test. Peak B was further separated on Bio-Gel P-4. At least two major fractions free of salt and GABA in the molecular weight range of approximately 700--3600 were shown to displace [3H]diazepam in the receptor test. This activity was enhanced by a factor of 3 in the presence of 10 microM-GABA.

The GABA postsynaptic membrane receptor-ionophore complex. Site of action of convulsant and anticonvulsant drugs

The function of the inhibitory neurotransmitter, gamma-aminobutyric acid (GABA), has been implicated in the mode of action of many drugs which excite or depress the central nervous system. Many convulsant agents appear to block GABA action whereas anticonvulsants enhance GABA action. Some of these drug effects involve altered GABA-mediated synaptic transmission at the level of GABA biosynthesis, release from nerve endings, uptake into cells, and metabolic degradation. A greater number of agents of diverse classes appear to affect GABA action at the postsynaptic membrane, as determined from both electrophysiological and biochemical studies. The recently developed in vitro radioactive receptor binding assays have led to a wealth of new information about GABA action and its alteration by drugs. GABA inhibitory transmission involves the regulation, by GABA binding to its receptor site, of chloride ion channels. In this GABA receptor-ionophore system, other drug receptor sites, one for benzodiazepines and one for barbiturates/picrotoxinin (and related agents) appear to form a multicomponent complex. In this complex, the drugs binding to any of the three receptor categories are visualized to have an effect on GABA-associated chloride channel regulation. Available evidence suggests that the complex mediates many of the actions of numerous excitatory and depressant drugs showing a variety of pharmacological effects.

An endogenous ligand to the benzodiazepine receptor: preliminary evaluation of its bioactivity

Chromatographic separation of aqueous brain extracts yields a peptide containing fraction which competitively inhibits 3H-diazepam binding to its receptor. An intracerebral-ventricular injection of this isolated fraction results in altered responses in pharmacological and behavioral tests which are similar to those observed when diazepam is administered in the same fashion. The most pronounced effect was obtained in the conflict test. Changes observed in other tests, such as blocking pentylenetetrazole convulsions, altering motility or reducing hyperthermia, were also consistent with the actions of diazepam. At the dose used, neither diazepam nor the brain extract altered muscular co-ordination in two ataxia evaluations. Thus, the animals' performance in the other paradigms would not be adversely influenced by immobilization side-effects. The results reported here support the notion that an endogenous factor does exist in brain which can act like the benzodiazepine drugs when tested for bioactivity in animal studies.

Anticonvulsant doses of inosine result in brain levels sufficient to inhibit [3H] diazepam binding

Several purines have been shown to be competitive inhibitors of [3H] diazepam binding. Inosine has also been shown to have benzodiazepine-like neurophysiologic, pharmacologic and behavioral effects, and to partially inhibit caffeine-induced seizures in mice. Using presumptive therapeutic doses of inosine, levels were determined in mouse brain at various times following injection. Inosine and hypoxanthine concentrations in brain increased several fold following inosine administration, indicating that inosine permeated the blood-brain barrier. The levels of inosine and hypoxanthine attained in brain were sufficient to inhibit by more than 50% the GABA-stimulated [3H] diazepam binding. These data suggest that the anticonvulsant properties of inosine are related to its interaction with the benzodiazepine receptor.

Endogenous modulator of benzodiazepine binding in rat cortex

Benzodiazepine binding sites, solubilized with 1% digitonin, were used to study specific [3H]flunitrazepam ([3H]FNP) binding. Specific binding increased nonlinearly with increasing amounts of digitonin extract in the assay. Specific binding was increased, and the relationship to amount of extract became linear, in the presence of 2% polyethylene glycol 6000 (PEG). Heat treatment destroyed binding activity of the extract, but not ability to inhibit [3H]FNP binding. Kinetic analysis showed inhibition to be noncompetitive. The inhibitory activity was sensitive to trypsin. Extracts of repeatedly frozen, thawed, and washed membrane preparations still possessed inhibitory activity. It is suggested that digitonin solubilizes a membrane protein that inhibits benzodiazepine binding. PEG apparently removes this substance from the binding sites.

beta-Carboline-3-carboxylic acid ethyl ester antagonizes diazepam activity

Analogous to the progression of events in the opiate receptor-enkaphalin area, the first reports that benzodiazepines have selective and specific high-affinity binding sites in brain have stimulated a search for the endogenous 'ligand' or substance that might normally act at these sites. Braestrup and co-workers have extracted from human urine a gamma-fraction (ref. 10) which they have recently identified as beta-carboline-3-carboxylic acid ethyl ester (beta CEE). They reported that this substance is extremely potent in displacing 3H-diazepam from brain binding sites and proposed that a beta-carboline-3-carboxylic acid derivative might, in part, be the endogenous ligand for the brain benzodiazepine receptor. We have examined several synthetically derived beta-carboline-3-carboxylic acid analogues and now present data obtained from testing only the beta CEE described by Braestrup et al. In addition to confirming these workers' observation that this compound is a potent displacer of 3H-diazepam from brain tissue, our pharmacological data indicate that beta CEE has activity that is opposite to, rather than similar to, that of diazepam.

Endogenous peptide(s) inhibiting [3H]cocaine binding in mouse brain

The supernatant fraction of centrifuged homogenate of brain tissue contains material that inhibits the saturable binding of [3H]cocaine to crude mouse brain membranes. This material was subjected to heat treatment to remove protein; further purification was achieved by filtering through an Amicon UM-10 membrane ultrafilter and gel filtration of the ultrafiltrate on Sephadex G-25. Sensitivity to acid hydrolysis and peptidase action indicates that the inhibitory activity resides in peptide material with low molecular weight. The partially purified inhibitor has similar effects to that of cocaine on the specific binding of various ligands to opiate and nonopiate receptors in mouse brain membranes.

1-Methyl-beta-carboline (harmane), a potent endogenous inhibitor of benzodiazepine receptor binding

The interaction of several beta-carbolines with specific [3H]-flunitrazepam binding to benzodiazepine receptors in rat brain membranes was investigated. Out of the investigated compounds, harmane and norharmane were the most potent inhibitors of specific [3H]-flunitrazepam binding, with IC50-values in the micromolar range. All other derivatives, including harmine, harmaline, and several tetrahydroderivatives were at least ten times less potent. Harmane has been previously found in rat brain and human urine, so it is the most potent endogenous inhibitor of specific [3H]-flunitrazepam binding known so far, with a several fold higher affinity for the benzodiazepine receptor than inosine and hypoxanthine. Thus, we suggest that harmane or other related beta-carbolines could be potential candidates as endogenous ligands of the benzodiazepine receptor.

Anticonvulsant effects of careulein and cholecystokinin octapeptide, compared with those of diazepam

Caerulein and the C-terminal octapeptide of cholecystokinin (CCK-8), after subcutaneous administration to mice, both delayed the onset and retarded the development of toxic effects of convulsants such as strychnine, pentetrazol, bicuculline, and picrotoxin. They also increased the seizure threshold doses of intravenously infused pentetrazol and picrotoxin. In this regard, both peptides were at least equipotent with diazepam.