Erling N. Petersen

Evaluation of the β-carboline ZK 93 426 as a benzodiazepine receptor antagonist

We describe here biochemical and pharmacological effects of the β-carboline ZK 93426, a new and potent benzodiazepine (BZ) receptor antagonist. ZK 93426 was compared with Ro 15-1788 and CGS 8216, two compounds previously described as BZ receptor antagonists. Certain effects of ZK 93426, Ro 15-1788 and CGS 8216 were quite similar (e.g., 3H-FNM displacement, “GABA ratio”, “photo-shift”). In most pharmacological tests ZK 93426 and Ro 15-1788 lacked overt effects; Ro 15-1788 was a weak agonist in some paradigms, while ZK 93426 exhibited a potent proconflict effect but also a weak anticonvulsant effect. This interesting finding with ZK 93426 suggests that BZ receptor ligands may possess differential efficacy at BZ receptor subtypes. In contrast, CGS 8216 exhibited potent proconvulsant effects in several paradigms in addition to proconflict and pentylenetetrazol generalizing effects. ZK 93426, Ro 15-1788 and CGS 8216 were almost equally potent as antagonists of the effects of BZ receptor agonists, such as diazepam and lorazepam. However, ZK 93426 was the most potent inhibitor of the convulsions produced by the BZ receptor inverse agonist DMCM.

Audiogenic seizures in DBA/2 mice discriminate sensitively between low efficacy benzodazepine receptor agonists and inverse agonists

In experiments with audiogenic seizures in DBA/2 mice, we observed that several socalled benzodiazepine receptor antagonists exhibited either anticonvulsive (Ro 15-1788, PrCC) or proconvulsive (FG 7142, beta-CCE, CGS 8216) effects at high receptor occupancy (17-85%), as compared to benzodiazepines and DMCM which had anticonvulsive and proconvulsive actions, respectively, at very low receptor occupancy (less than 10%). Sensitive distinction between benzodiazepine receptor ligands with low anticonvulsive efficacy (partial agonists) and ligands with low proconvulsive, and maybe anxiogenic, efficacy (partial inverse agonists) can thus be obtained in sound seizure susceptible mice.

Proconflict effect of benzodiazepine receptor inverse agonists and other inhibitors of GABA function.

GABA seems to be a neurotransmitter with great impact on conflict behaviour in rats. We studied the effects of different types of GABA function inhibitors on conflict behaviour in rats. Among these inhibitors, the benzodiazepine (BZ) receptor inverse agonists are a new type of compound downregulating GABA-mediated functions allosterically. The most effective proconflict inducing compounds were pentylenetetrazol and the three BZ inverse agonists beta-CCM, beta-CCE and ZK 90886. The BZ receptor inverse agonists, FG 7142, DMCM and CGS 8216, the GABA antagonist bicuculline and the GABA synthesis inhibitor isoniazid were moderately active. Only a weak effect was seen with just subconvulsive doses of picrotoxin, a chloride channel inhibitor. These results show that the mode of GABA function inhibition determines the degree to which proconflict action is elicited and that proconflict effects and proconvulsant or convulsant effects may be separated. Evidence is presented that proconflict action in rats is predictive of an anxiogenic action in man.

ZK 91296, a partial agonist at benzodiazepine receptors

ZK 91296 (ethyl 5-benzyloxy-4-methoxymethyl-β-carboline-3-carboxylate) is a potent and selective ligand for benzodiazepine (BZ) receptors. Biochemical investigations indicate that ZK 91296 may be a partial agonist at BZ receptors. Such partial agonism may explain to some extent why ZK 91296 needs higher BZ receptor occupancy than diazepam for the same effect against chemical convulsants and for behavioural effects. The lack of sedatiye effects, and the very potent inhibition of reflex epilepsy, spontaneous epilepsy and DMCM-induced seizures suggest, furthermore, that ZK 91296 may possess pharmacological selectivity for a particular type of BZ receptor interaction, perhaps including topographic as well as receptor subtype differentiation.

DMCM: a potent convulsive benzodiazepine receptor ligand.

DMCM (methyl 6,7-dimethoxy-4-ethyl-beta-carboline-3-carboxylate) is a very potent convulsant with high affinity for specific benzodiazepine binding sites. A number of compounds were compared for their ability to prevent seizures induced by DMCM and pentylenetetrazol. DMCM seizures were antagonized by benzodiazepine (BZ) receptor antagonists, such as Ro 15-1788, CGS 8216 and several beta-carboline-3-carboxylates, which all fail to inhibit pentylenetetrazol seizures. The benzodiazepines diazepam, clonazepam and lorazepam as well as valproate, ethosuximid, phenobarbital, primidone, diphenylhydantoin and carbamazepine antagonized both DMCM and pentylenetetrazol. Muscimol and gamma-vinyl-GABA did not inhibit DMCM seizures whereas THIP showed a weak and selective effect against DMCM. Valproate showed a relatively potent (60 mg/kg i.p.) and competitive antagonism of short duration. Baclofen antagonized DMCM at 3 mg/kg. Valproate and baclofen were at least 5 times more potent against DMCM-induced than against pentylenetetrazol-induced seizures. DMCM most probably induces the seizures by selective impairment of the functions mediated by the GABA/BZ receptor-chloride channel complex (inverse agonism) and therefore differs from GABA receptor blockers.

Bidirectional effects of β-carbolines and benzodiazepines on cognitive processes

Abstract Experiments with benzodiazepine receptor ligands in two paradigms involving cognitive processing were performed in order to test whether the concept of bidirectional effects of benzodiazepine receptor ligands could also be applied to cognitive functions. Benzodiazepine receptor agonists like chlordiazepoxide, lorazepam, ZK 93423 and ZK 91296 induced amnesia in a passive avoidance paradigm. Mice treated with the benzodiazepine receptor antagonist, ZK 93426, reached a learning criterion after fewer foot-shocks than saline treated mice both in naive animals and in scopolamine pre-treated animals. Furthermore, ZK 93426, attenuated the amnesic effect of corneal electroshock. The inverse agonists FG 7142 and DMCM decreased the detrimental effect of scopolamine on retrieval. In a signal detection paradigm, chlordiazepoxide impaired signal detection. In aged rats ZK 93426, ZK 90886 and FG 7142 had no effect on signal detection but ZK 93426 and FG 7142 attenuated the impairment of signal detection induced by scopolamine. These effects of benzodiazepine receptor ligands may reflect changes in arousal/vigilance, suggesting that BZ inverse agonists may have useful properties in enhancing vigilance.

Does the reversal of the anticonflict effect of phenobarbital by β-CCE and FG 7142 indicate benzodiazepine receptor-mediated anxiogenic properties?

In mice and rats, the high affinity ligand for brain benzodiazepine (BZ) receptors beta-CCE, and the more stable congener FG 7142, failed to exert anticonflict activity in conflict situations but instead reversed the anticonflict effect of lorazepam. In contrast to Ro 15-1788, beta-CCE and FG 7142 also antagonized the anticonflict effect of phenobarbital in rats. This effect suggests that beta-CCE and FG 7142 may produce anxiety by either inducing a conformational change in the BZ receptors which is directly opposite to that induced by the benzodiazepines, or binding to a particular subclass of BZ receptors.

A water lick conflict paradigm using drug experienced rats

A modified water-lick conflict paradigm is described, using trained rats for up to 35 weekly test sessions under 48 h of water deprivation. The rats rapidly became maximally suppressed by the punishment. This suppression was attenuated by the anxiolytics lorazepam, diazepam, phenobarbital, and meprobamate. The potentially anxiolytic drug CL 218872 and the anticonvulsant drug valproate sodium were also active. The antiserotonin drugs methysergide, cyproheptadine, cinanserin and parachlorophenylalanine were all inactive, as were several other distinct classes of psychotropic drugs including propranolol, clonidine, THIP, theophylline, chlorpromazine, paroxetine and ethanol. The paradigm proved reliable, reproducible and useful for large scale investigations. Furthermore, it may provide means for detailed neuropharmacological and anatomical studies.

Modulation of anxiety by β-carbolines and other benzodiazepine receptor ligands: Relationship of pharmacological to biochemical measures of efficacy

Several beta-carbolines and other benzodiazepines (BZ) receptor ligands have been investigated for anxiolytic or anxiogenic action in 4 unrelated animal models of anxiety using rats. The substances could be grouped into essentially 2 groups. The first, anxiolytics, exhibited antipunishment activity in a lick-suppression test, antagonised the discriminative stimulus provided by pentylenetetrazol, resembled chlordiazepoxide (CDP) in a drug discrimination test, and reduced the rise in plasma corticosterone levels following swim stress. Such substances included several benzodiazepines, the beta-carboline ZK 93 423, and the triazolapyridazine CL 218 872. A subgroup of anxiolytics were active in only some of these tests. They included two beta-carbolines, ZK 91 296 and ZK 95 962, and the pyrazoloquinoline CGS 9896, and these 3 substances were also distinguishable in not producing rate-decreasing effects in any of the 3 operant tests. The second group were anxiogenic in that they produced a discriminative stimulus resembling that of PTZ, they antagonised the CDP cue, exhibited propunishment effects in the lick-suppression test, and themselves caused increases in plasma corticosterone in otherwise unstressed animals. Such substances included the beta-carbolines DMCM, FG 7142 and ZK 90 886, and the pyrazoloquinoline CGS 8216. Two substances, Ro 15-1788 and ZK 93 426 had little or only weak activity in any test. The classification of these substances into anxiolytics or anxiogenics could be predicted qualitatively both by their ability to enhance (anxiolytics) or decrease the binding of 35S-TBPS to rat brain membranes and by whether their own binding was increased (anxiolytics) by adding the GABA agonist muscimol to the in vitro incubation medium. For the limited number of substances for which full data was available, there was also a quantitative relationship between the degree of enhancement of 35S-TBPS binding by a substance and its potency in the CDP cue test when such potency was expressed as numbers of BZ receptors occupied at the ED50 value in the pharmacological test. Furthermore, for the anxiolytics, activity in the CDP cue correlated significantly with potency in 2 other tests. Otherwise, surprisingly weak correlations existed between potencies in the different tests. In particular, the beta-carboline ZK 95 962 was highly potent in antagonising the PTZ cue but inactive in both a conflict test and in protecting against stress. These results are discussed in terms of differences in the neuropharmacologies of the 4 tests and in selectivity of the BZ receptor ligands for subtypes of BZ receptor.

Evidence that the anticonflict effect of midazolam in amygdala is mediated by the specific benzodiazepine receptors.

The benzodiazepine midazolam produced an anticonflict effect in rats measured in a water lick paradigm following local injection into the basolateral and lateral complex of the amygdala. This effect of midazolam seems to involve specific benzodiazepine receptors, since the systemic injection of benzodiazepine antagonists Ro 15-1788, ZK 93426, FG 7142 and CGS 8216 produced strong antagonism of the effect of midazolam in doses not affecting the non-punished drinking behaviour.