Benzodiazepine interactions with GABA receptors

Neuroscience Letters, 47 (1984) 201-206

201

Elsevier Scientific Publishers Ireland Ltd.

NSL 02779

BENZODIAZEPINE INTERACTIONS WITH GABA RECEPTORS

WILLY HAEFELY

Pharmaceutical Research Department, F. Hoffmann-La Roche, Ltd., CH-4002 Basel (Switzerland) (Received March 26th, 1984; Accepted April 2nd, 1984)

Key words: benzodiazepines - ~,-aminohutyric acid - chloride channel - synaptic inhibition - inverse agonists - GABA potentiation - GABA receptor - benzodiazepine receptor

Benzodiazepines (BZs) produce most, if not all, of their pharmacological actions by specifically enhancing the effects of endogenous and exogenous GABA that are mediated by GABAA receptors. This potentiation consists in an increase of the apparent affinity of GABA for increasing chloride conductance without increase in its efficacy and, on the single chloride channel level, seems to be the result of an increased probability of channel opening events. Recent studies indicate that the BZR is a site on the GABA receptor (GABA-R)-chloride channel complex, through which the gain of the signal transducer function of the latter is allosterically modulated. The unique feature of this drug receptor, which is located on a neurotransmitter receptor-gated ion channel, is its specific interaction with three classes of ligands. Agonists, competitive antagonists and inverse agonists at BZR, respectively increase, do not alter and reduce the gain of the GABA-R function. Compounds have been found that cover a whole spectrum of transition between the two extremes (partial agonists, partial inverse agonists) and for which interesting therapeutic applications can be foreseen.

There is at present hardly a group of psychotropic drugs more suitable than benzodiazepines (BZs) to attract and satisfy the interest of investigators both in basic pharmacology and in clinical neurosciences. The reason for this is two-fold. On the one hand, BZs are the most widely used psychotropic drugs; after more than twenty years of experience their main effects in man are now well known and their exceptionally high safety is generally accepted. On the other hand, BZs have been shown to interact with the central nervous system (CNS) in a very specific manner, namely by enhancing most effects of the major inhibitory neurotransmitter, GABA. They do it by an apparently novel and very exciting molecular mechanism which involves the allosteric modulation of the receptor-effector function of GABA receptors (GABA-Rs). The interesting profile of therapeutic activity of BZs, their safety, and their specific site of interaction with one of the most vital, chemical interneuronal communication systems of the CNS make these drugs an almost ideal tool both in basic research to understand more deeply the function of neurotransmitter receptors and, in clinical neurosciences, to investigate the role of GABA in physiological and pathological conditions. Benzodiazepines and GABA effects. BZs have been shown in the past decade to 0304-3940/84/$ 03.00 © 1984 Elsevier Scientific Publishers Ireland Ltd.

202 enhance GABAergic transmission and the effect of exogenous GABA in all major regions of the CNS investigated so far [2]. This GABA potentiation has been found in elecrophysiological experiments (a) by stimulating distinct GABAergic pathways and recording the response of GABA responsive target cells (single cells or cell population), e.g. the presynaptic inhibitory pathway in the spinal cord or recurrent inhibitory pathways in cortex, hippocampus and hypothalamus, (b) by studying the effect of exogenous GABA on single cells in vitro as well as in situ, and (c) by recording the spontaneous activity of neurones known to be GABAergically innervated. In biochemical experiments, the level of cyclic guanosine 3¢5 '-phosphoric acid (cGMP) in the cerebellar cortex has been used as an indirect parameter reflecting the balance between excitatory (mostly glutamatergic) and inhibitory GABAergic inputs to Purkinje cells. A consistent potentiation of GABAergic transmission by BZs is obtained in some circuits, e.g. of presynaptic inhibition in the spinal cord and of recurrent inhibition in several regions. A potentiation was not observed in the striatonigral pathway, unless GABAergic transmission was experimentally depressed. Irregular potentiation of spontaneous and evoked inhibitory postsynaptic potentials (IPSPs) in several brain areas is an observation made by several investigators. Possible explanations for these differences will be provided below. Insight into the molecular mechanism, by which BZs enhance the effects of GABA on the neuronal membrane conductance, has been obtained using three approaches (see ref. 2). The classic pharmacological approach was to study dose-response curves for the conductance-increasing effect of exogenous GABA on cultured spinal cord cells or on primary afferent endings under the effect of BZs. BZs were found to shift the GABA dose-response curve to the left without affecting its maximum. In phamacological terms this effect of BZs is an increase of the apparent affinity of GABA without any change in its apparent efficacy. The shift of the GABA dose-response curve by BZs was only two- to three-fold. The second approach was the study of elementary chloride channel characteristics in cultured spinal cord neurones using noise analysis. BZs were found to enhance the probability of single-channel opening events in response to GABA, to have inconsistent or no effect on single channel half-life and to fail to affect single-channel conductance, total number of channels and driving force for chloride flux. Results obtained by both approaches could be accounted for by a BZ-induced increase of the affinity of GABA-Rs for GABA or of the coupling between GABA binding and channel gating. The biochemical approach made use of the radioligand technique; under critical experimental conditions BZs were reported to enhance the binding of GABA to specific binding sites in neuronal cell membranes, thus supporting the former conclusion drawn from electrophysiological studies (see ref. 3). The benzodiazepine receptor. The specific high-affinity binding sites for BZs are very sensitive to agents interacting with GABAA-RS and chloride channels: BZ-receptor (BZR) agonists bind with higher affinity in the presence of GABA-R agonists

203

\

~r"11~. ,"

tl/ t

1

/ j/ i .X\ }\'q I

,'

3

Fig. 1. Hypothetical model of the GABA-R-BZR-chloride channel complex (above view from outside of the membrane, below section along the membrane axis). The complex is thought to be formed by four monomeric peptide units, labelled l to 4. These units contain at least three domains with different functions: C is the anion channel-forming part, G the GABA-binding domain and B the BZ-binding site. The binding sites for so-called channel agents, GABA-R ligands and the three classes of BZR ligands on the three domains are indicated on monomer 2. On monomer 3 is indicated that GABA-R activation results in the opening of the anion channel, and that agents interacting with the C and B domains can affect alosterically this gating process. On the fourth monomer are indicated by arrows the multiple bidirectional interactions between the three domains.

a n d o f some agents s u p p o s e d to act o n the chloride c h a n n e l , e.g. b a r b i t u r a t e s . These very consistent interactions, together with the pieces o f evidence m e n t i o n e d above, strongly suggest that BZs act o n a site that is f u n c t i o n a l l y a n d physically very int i m a t e l y c o u p l e d to G A B A - R s a n d G A B A - R - g a t e d chloride channels. Fig. 1 is a schematic a n d hypothetical m o d e l o f the s u p r a m o l e c u l a r complex c o m p r i s i n g three distinct f u n c t i o n s : G A B A r e c o g n i t i o n a n d b i n d i n g , BZ r e c o g n i t i o n a n d b i n d i n g , a n d

204

anion translocation. The anion channel seems, in addition, to bind a variety of agents that are able to affect the anion translocation function, such as channelblocking convulsants (prototype: picrotoxinin) and channel-facilitating a n d / o r activating drugs (prototypes: various barbiturates). Most recent findings in favour of the model are the copurification of proteins displaying the capacity to bind BZ as well as GABA and channel convulsants, and the synthesis and membrane incorporation of a functional GABA-gated and BZ-sensitive chloride channel by frog oocytes injected with messenger ribonucleic acid (mRNA) from chicken brain [5]. While all BZ binding sites of the central high-affinity type appear to be coupled with GABA-R-chloride channel complexes, the reverse does not seem to hold true. GABAB-Rs do not seem to be modulated by BZs; whether all GABAA-RS of vertebrates also contain a BZ binding site remains to be shown. GABA-Rs of invertebrates, at least in the membrane-bound form, do not contain a BZ binding site. BZRs have recently been found to bind with high affinity a variety of agents that are chemically distinct from BZs, e.g. agents with similar effects asthe classical BZs, such as cyclopyrrolones (zopiclone, suriclone), triazolopyridazines (CL 218,872),/3carbolines (ZK 91296) and pyrazoloquinolinones (CGS 9895) (see ref. 3). Benzodiazepine antagonists. Compounds with selective competitive antagonistic activity have been found among imidazobenzodiazepinones. The most intensively studied antagonist in animals as well as in man is Ro 15-1788 [1]. This compound is highly selective for BZRs, meaning that no effects mediated by an action on other receptors have been found in relevant doses. Its purity as an antagonist is very high, i.e. in most studies agonistic effects have not been revealed. Under certain conditions, however, a weak agonistic component of Ro 15-1788 has been found, which is virtually restricted to a slight anticonvulsant activity. This partial agonistic activity, though very weak, has to be considered when the drug is used to reveal the presence of a putative endogenous ligand of BZRs. BZ antagonists have also been found in pyrazoloquinolinones; CGS 8216 is a potent ligand of BZRs but also blocks adenosine receptors. Various esters of /3carboline-3-carboxylate have more or less pure BZ antagonistic acitivty; the best known is the ethyl ester of/3-carboline-3-carboxylate (/3-CCE). Inverse agonists at benzodiazepine receptors. A most exciting discovery has been the finding that agents with high affinity to BZRs may produce effects that are the exact opposite of those produced by the classical BZ agonists, e.g. convulsions, anxiety, and sleep suppression. Prominent representatives of this group of agents are found among/3-carboline derivatives, the most typical one being methyl-6,7-dimethyl-4-ethyl-/3-carboline-3-carboxylate (DMCM). The effects of these agents are blocked by both BZ agonists and antagonists, showing that their effect is, indeed, mediated by BZRs. These inverse agonists obviously depress the receptor-effector function of the GABA-R-chloride channel complex. Thus, it appears that we are faced with the first example of a pharmacological receptor which can mediate two opposite effects. How this is brought about is not yet clear. Inverse agonists may

205 induce a conformation of the receptor which is different from that induced by agonists and which reduces the bias of the GABA-R-chloride channel coupling; together with the conformation induced by agonists (which increases the bias) and the inactive or neutral conformation stabilized by a pure antagonist, three different conformational and functional states of the BZR would exist [4]. Alternatively, the BZR may exist normally in two interconvertible conformational states that are in equilibrium. Agonists may preferably bind to and stabilize one state thereby shifting the equilibrium in its favour (increase of the bias, positive efficacy). Inverse agonists would bind preferably to and stabilize the other state, shifting the equilibrium in the direction of this state (decrease of the bias, negative efficacy). Antagonists would be unable to distinguish between the two states and, correspondingly, would not alter the equilibrium. Inverse agonists with low negative efficacy could be useful in situations of abnormally reduced arousal and attention because, in contrast to GABA-R blockers, they should allow a limited reduction of the bias with a reduced danger of a too massive depression of the GABA-R-chloride channel function. GABA potentiation and benzodiazepine actions. Although BZs are in principle able to enhance the effect of GABA on all central neurones investigated so far, one should not conclude that in the presence of these drugs the efficiency of all GABAergic synapses is uniformly augmented. The potentiating effect of BZs depends on the actual GABA concentration in a given synapse; since the drugs produce a parallel shift of the GABA dose-response curve, they will be ineffective in synapses in which the GABA concentration is sufficiently high to activate all functional GABA-Rs that are present in the subsynaptic membrane, but they will be effective when the synaptic GABA concentration is submaximal. BZs differ fundamentally from GABA agonists by not activating GABA-Rs directly but by facilitating the effect of synaptically released GABA. The effect of BZs is not a general depression of neuronal activity; the existence of pathways with GABA neurones in series can, indeed, result in disinhibition. Moreover, depolarization by GABA of terminals, such as primary afferent endings, is enhanced by BZs and may result in the increased generation of action potentials. Subtypes of BZRs have been proposed to explain different profiles of activity of more recent BZR ligands. Alternative explanations should be checked, e.g. the possibility that partial agonistic property combined with varying receptor reserves in the different neuronal pathways may enable anxiolytic or anticonvulsant activity to be maintained and other actions, such as sedation and physical dependence liability, reduced. The role of GABA in functions and states of the CNS that are affected by BZs is becoming increasingly appreciated. Of particular interest is the generation of many forms of epilepsy by GABA-R of chloride-channel blockers as well as by inverse agonists at BZRs. Experimental anxiety, sleep disturbances and increased muscle tone are also produced by these agents. Depression of GABAergic transmission decreases the firing threshold of neurones in the brainstem reticular formation and

206

in the cerebral cortex and increases their receptive fields. Since the effect of BZs has been found to be most pronounced under conditions of experimentally reduced GABAergic transmission, it is not too surprising that these drugs at appropriate doses have beneficial effects on, for example, anxiety or insomnia without greatly disturbing normal functions. 1 Haefely, W., Antagonists of benzodiazepines, L'Enc6phale, 9 (1983) 143B-150B. 2 Haefely, W. and Polc, P., Electrophysiological studies on the interaction of anxiolytic drugs with GABAergic mechanisms. J.B. Malick, S.J. Enna and H.I. Yamamura (Eds.), Anxiolytics: Neurochemical, Behavioral and Clinical Perspectives, Raven Press, New York, 1983, pp. 113-145. 3 Haefely, W., Kyburz, E., Gerecke, M. and M6hler, H., Recent advances in the molecular pharmacology of benzodiazepine receptors and in the structure-activity relationships of their agonists and antagonists, Advanc. Drug Res., in press. 4 Polc, P., Bonetti, E.P., Schaffner, R. and Haefely, W., A three-state model of the benzodiazepine receptor explains the interactions between the benzodiazepine antagonist Ro 15-1788, benzodiazepine tranquilizers, t3-carbolines and phenobarbitone, Arch. Pharmacol., 321 (1982) 260-264. 5 Smart, T.G., Constanti, A., Bilbe, G., Brown, D.A. and Barnard, E.A., Synthesis of functional chick brain GABA-benzodiazepine-barbiturate/receptor complex in mRNA-injected Xenopus oocytes, Neurosci. Lett., 40 (1983) 55-59.