Octodecaneuropeptide Structure

Figure 19.6 The chemical structure of the imidazopyridine and benzodiazepine (BZ^ receptor ligand, zolpidem, and the cyclopyrrolone, zopiclone

Whether simple augmentation of GABAa receptor function accounts for the anti-anxiety effects of these compounds remains equivocal. If this was the case then other agents that augment GABAergic transmission such as inhibitors of GABA uptake (e.g. vigabatrin) or metabolism (e.g. tiagabine) should also have anti-anxiety effects. Indeed, there are reports of their anti-anxiety effects in patients receiving these treatments for relief of epilepsy. There is also some supporting evidence from preclinical studies but the behavioural effects of these drugs in animal models are less robust than are those of the benzodiazepines. It remains to be seen whether this is because they are just less effective anti-anxiety agents than the benzodiazepines or whether existing preclinical models show a bias that detects preferentially the anti-anxiety effects of benzodiazepines.


In the 1980s, a further binding site for benzodiazepines was identified and, because it was first discovered in the rat adrenal gland, the term 'peripheral benzodiazepine receptor' was coined. This is regrettably confusing because this receptor has now been found in the brain also (Awad and Gavish 1987). These benzodiazepine receptors differ from those described above in a number of important respects, not least because they do not affect, nor are they affected by, GABA binding. They also have their own specific ligands: the isoquinolone, PK 11195, and the benzodiazepine, Ro 4864, neither of which binds to the GABAA receptor. Moreover, the benzodiazepine, clonazepam, which is a high-affinity, partial agonist ligand for the benzodiazepine domain on the GABAA receptor, does not bind to the 'peripheral' receptor.

Another difference is that these peripheral benzodiazepine receptors are located mainly intraneuronally, on mitochondrial outer membranes, rather than on the plasma membrane. In the brain, they are associated with glial cells but in the periphery they are found in a range of tissues, including mast cells and platelets. Their function is still a matter of intense debate but one possibility is that they regulate cholesterol uptake and, secondary to this, the synthesis of neurosteroids (Do Rego et al. 1998). Since neuro-steroids also have a binding domain on the GABAA receptor, there might be some indirect functional coupling between these two types of receptors after all. To some extent, the possibility of such an interaction is supported by evidence that the density of peripheral benzodiazepine receptors differs in inbred strains of rats which are distinguished by their behavioural reactivity ('fearfulness') to novel stimuli (Drugan et al. 1987). Other possible, albeit controversial, functions of these receptors are reviewed in Doble and Martin (1996).


Among the early indications that the benzodiazepines were not the only compounds to bind to the benzodiazepine receptor were findings that emerged from a search for an endogenous ligand for this receptor site. This effort produced a non-benzodiazepine ligand, ethyl-;6-carboline-3-carboxylate (¿-CCE), and this pointed the way to a whole family of compounds that are high-affinity ligands for this receptor. However, not all turned out to share the properties of the protypical benzodiazepines (anti-anxiety, anticonvulsant, etc.). Some, including ¿-CCE itself, had the opposite effects in animals: i.e. they induced anxiety and reduced seizure threshold and some, such as 3-carbomethoxy-4-ethyl-6,7-dimethoxy-^-carboline (DMCM), caused overt seizures.

These new recruits to the activity spectrum were named 'inverse agonists' and subsequent studies confirmed that they reduce the affinity of GABA for its binding site on the GABAa receptor and attenuate the GABAa receptor-mediated increase in Cl" conductance (Fig. 19.5).


The rich portfolio of compounds that bind to the benzodiazepine receptor includes many compounds which, despite not being benzodiazepines, share the properties of the prototypical benzodiazepines, chlordiazepoxide and diazepam. However, all these groups of compounds, including the benzodiazepines themselves, span the activity spectrum: from full inverse agonist to full agonist. In between these extremes are compounds which have either partial agonist or partial inverse agonist activity and some are antagonists (Fig. 19.7). This spectrum of actions reflects the overall effects of these drugs on native receptors and is usually assessed in whole animals. However, the synthesis of receptors comprising different combinations of subunits has shown that the activity of these drugs depends greatly on subunit composition. For instance, GABAa receptors have been characterised to which diazepam does not bind at all (see Chapter 11).

The first antagonist to be developed was the (imidazo)benzodiazepine, flumazenil. This compound blocks the actions of both agonists and inverse agonists in vitro. It will

Figure 19.7 The activity spectrum for different generic groups of compounds that bind to the 'benzodiazepine' domain on the GABAa receptor

also block the effects of these agents in vivo, but there is some argument over whether it is a true antagonist: i.e. whether it really lacks any intrinsic activity (as would be required of an antagonist) or whether it is merely a weak partial agonist. The subunit composition of the GABAa receptor could be one confounding factor in resolving this question. For instance, flumazenil has been reported to augment the action of GABA at cloned receptors comprising a4 fi2 y2 subunits. Apparent effects of the antagonist in vivo could also depend on whether there is any tonic activation of the benzodiazepine receptor by an endogenous ligand. Flumazenil is available in the clinic for intravenous infusion to reverse benzodiazepine-induced sedation (e.g. in the post-anaesthetic context) or coma (after overdose). However, because it has a half-life of only 1 h in humans, it is only of realistic benefit in reversing the actions of agonist benzodiazepines with a short half-life, such as midazolam.

The potential benefits of benzodiazepine partial agonists are as non-sedative, anti-anxiety agents. Because of their low efficacy, it was predicted that a partial agonist should not induce sedation even if their receptor occupancy exceeds that normally required for an anti-anxiety effect when using a full agonist. One such compound, bretazenil, has been developed but failed to reach the clinic because it displayed some sedative activity and, more problematic, there were end-of-dose rebound effects that were undoubtedly exacerbated by its short half-life. Currently, the partial agonist, abecarnil (a ^-carboline), is undergoing clinical trials. For the current status of the development of partial agonists and other promising benzodiazepine receptor ligands see Cheetham and Heal (2000).

Even benzodiazepine inverse agonists might yet find some useful applications such as in the relief of cognitive deficits (which are increased by benzodiazepine full agonists) (Abe, Takeyama and Yoshimura 1998). With the rapidly expanding understanding of different combinations of subunits that comprise the GABAA receptor, it is hoped to develop compounds that target specific subunit combinations and improve cognitive function in dementia but which lack any proconvulsant or anxiogenic actions.


The discovery of the opioid receptor, followed by isolation of endogenous opioids, provided the impetus for a search for an endogenous ligand for the established benzodiazepine receptor. Although many candidates have emerged (De Robertis et al. 1988; Table 19.4), most are present in the CNS at concentrations far too low for them to be feasible endogenous modulators of GABAa receptor function. However, three candidates have been given prominent attention, albeit for different reasons, and are worthy of mention.

Table 19.4 Putative endogenous ligands for the benzodiazepine binding domain on the GABAA receptor

^-carbolines (0-CCB) Desmethyldiazepam

'Endozepines' (unknown chemical structure) Peptides (nepenthin, octodecaneuropeptide) Purines (inosine, hypoxanthine, guanosine, nicotinamide) Thromboxane A2

The first, ¿S-CCE, was the product of an arduous attempt to isolate an endogenous ligand from human urine. Although subsequently found to be an artefact of the extraction process, this compound turned out to be a ligand for the benzodiazepine receptor, nonetheless, and was the first inverse agonist to be identified. The anxiogenic effects in humans of its more stable congener, FG 7142, are described graphically in a report by Dorow et al. (1983). ¿S-Carbolines are realistic candidates for an endogenous ligand because they can be synthesised in the brain (Han and Dryhurst 1996) but, although other members of this group of compounds have at various times been suggested to fulfil the role of an endogenous ligand, none has been confirmed as such.

Another, more recent, candidate is an endogenous propeptide, 'diazepam binding inhibitor' (also known as Acyl-CoA Binding Protein (DBI/ACBP)), which yields 'octodecaneuropeptide' ('ODN') and 'triakontatetraneuropeptide' ('TTN') (Costa and Guidotti 1991). Both these peptides are neuroactive and ODN turns out to have inverse agonist activity at GABAA receptors both in vivo and in vitro and to have marked effects on behaviour (e.g. Reddy and Kulkarni 1998). However, there is scepticism as to whether the brain can manufacture sufficient peptide to regulate the ubiquitous GABAA receptor on a moment-to-moment basis. Currently, the binding of TTN to the peripheral benzodiazepine site, and its effect on neurosteroid synthesis, is attracting greater interest (Do Rego et al. 1998).

Finally, the presence in human post-mortem brain tissue of the active metabolite of diazepam, desmethyldiazepam, raised some curiosity and frank alarm (Sangameswaran et al. 1986). At the time of its discovery in the brain it was thought that there was no enzyme system capable of producing such halogenated compounds and that its presence in the brain reflected dietary intake from an environment contaminated by overuse of its parent compound. However, its discovery in stored brain tissue which had been obtained before the synthesis of the benzodiazepines allayed these fears. It is now thought possible that some benzodiazepines, including desmethyl-diazepam, occur naturally and that they are taken in as part of a normal diet (Table 19.5).

Although, by analogy with the opioids, one would expect there to be an endogenous ligand for the widely distributed benzodiazepine receptor, its existence remains uncertain and we must be alert to the possibility that any such ligand(s) could have either agonist or inverse agonist activity.

Table 19.5 Examples of plants containing ligands for benzodiazepine receptors

Plant source

Valeriana officinalis

Hypericum perforatum L. Hypericaceae

(St John's Wort) Matricaria recutita L. Passiflor coeruleus L. Wheat grain Potato

Karmelitter Geist

Active agent(s)

Hydroxypinoresinol (a lignan) Unknown

5,7,4'-trihydroxyflavone (apigenin) Chrysin

Diazepam, desmethyldiazepam, lormetazepam Diazepam, desmethyldiazepam, lormetazepam Amentoflavon


The undisputed efficacy of benzodiazepines in relief of anxiety led to the question of whether this disorder could arise from abnormal concentrations in the brain of an endogenous ligand or a malfunction of the benzodiazepine/GABA receptor system. An important study, aimed at distinguishing between these possibilities, has been carried out in humans (Nutt et al. 1990) and was based on the premise that anxiety could be caused by either:

(1) Inadequate activity of an endogenous ligand which is a benzodiazepine receptor agonist and suppresses anxiety. In this case, the administration of the antagonist, flumazenil, should induce anxiety in normal subjects and exacerbate anxiety in anxious patients.

(2) Excessive activity of an endogenous ligand which is a benzodiazepine receptor inverse agonist and induces anxiety. In this case, the administration of flumazenil should relieve anxiety in anxious patients and have no, or sedative, effects in healthy subjects.

(3) Dysfunction of the GABAA receptor complex such that the effects of all benzo-diazepine receptor ligands are shifted in the direction of inverse agonism. In this case, flumazenil (which normally has zero efficacy) should induce anxiety in anxious patients but have no effects in healthy subjects because they have normal receptors.

To distinguish between these possibilities, flumazenil was administered to panic patients and control subjects. The results of the experiment were consistent with the third possibility: flumazenil induced panic attacks in 8 of 10 patients but not in control subjects (Fig. 19.8). Unfortunately, the change(s) in the benzodiazepine receptor or its coupling to the rest of the GABAA receptor are unknown, as are the stimuli that could explain this functional change. Recent studies suggest that the binding of [nC]flumazenil is abnormally low in panic patients (Malizia et al. 1998), although this finding does not relate in any obvious way to the 'GABAA receptor shift' hypothesis. However, this is the only tested theory so far to connect panic anxiety directly with a disorder of the GABAA receptor. The receptor shift theory could also explain why benzodiazepines are ineffective in treating panic disorder but, because these drugs do effectively relieve generalised anxiety, it seems that the theory might explain the origin of the former, but not the latter disorder, and that they have different causes.

Continue reading here: Monoamines In Anxiety

Was this article helpful?

0 0