Interactions between the benzodiazepines, methylxanthines, and adenosine

From cAMP to adenosine: an illuminating shift of focus

In a remarkable career, straddling five decades, John Phillis pursued with fierce determination and exceptional energy the main goal of his scientific life, to throw light on the chemical agents that control brain function. Starting in Australia, he settled in North America, first in Canada, then in the USA, where his long tenure at Wayne State brought his career to its culmination.

Adenosinergic mechanisms in anticonvulsant action of diazepam and sodium valproate

The effects of adenosine receptor agonists and antagonists were studied in pentylenetetrazole (PTZ)-induced seizures in rats. Animals were pretreated with the non-specific adenosine receptor antagonist, theophylline (50 and 100 mg/kg, i.p.), or the specific A(1) adenosine receptor antagonist, 8-cyclopentyl-1,3-dipropylxanthine (DPCPX), in a dose of 1 mg/kg, i.p., followed by 100% anticonvulsant doses of diazepam (4 mg/kg)/sodium valproate (300 mg/kg, i.p.). Subsequently, they were challenged with convulsant doses of PTZ i.e. 60 mg/kg, i.p. It was seen that while DPCPX could not reverse the protection of both the antiepileptic drugs, theophylline significantly reversed this protection, as assessed by percent incidence of seizures and change in latency parameters. In another set of experiments, the rats were pretreated with a combination of subanticonvulsant doses of adenosine (500 mg/kg) or specific adenosine A(1) receptor agonist, cyclopentyladenosine (CPA) and diazepam (0.5 and 1 mg/kg)/sodium valproate (150 mg/kg), prior to PTZ challenge. We observed a decrease in incidence and increase in latency of seizures following either combination. The protection observed was independent of the hypothermic and hypotensive effects of adenosine and CPA. These results indicate that though A(1) agonist enhances the protection of diazepam and sodium valproate, a direct involvement of adenosine A(1) receptor in anticonvulsant action of these drugs is doubtful.

S-adenosyl-L-homocysteine and 5'-methylthioadenosine inhibit binding of [3H]flunitrazepam to rat brain membranes

The effect of inhibitors of transmethylation reactions, S-adenosylhomocysteine and 5'-methylthioadenosine, on [3H]flunitrazepam-specific binding to the rat brain membranes has been investigated. Both S-adenosylhomocysteine and 5'-methylthioadenosine are able to inhibit binding with Ki values of 7.9 microM and 15.8 microM respectively. These compounds therefore may be candidate endogenous benzodiazepine-receptor ligands. In light of these and other data possible correlations between phospholipid methylation and gamma-aminobutyric acid (GABA) receptor function are discussed.

Sustained synergism by chronic caffeine of the motor control deficit produced by midazolam

To evaluate the effects of chronic caffeine on the impairment of discriminative fine motor control produced by midazolam, rats were trained to hold a force transducer steady to deliver food pellets. Chronic, daily doses of midazolam (3 mg/kg SC) led to a stable level of motor impairment. Chronic caffeine (20 mg/kg IP) alone usually produced a more moderate deficit or, for one animal, no deficit. Combined, chronic administration of these doses yielded a sustained synergism in motor performance impairment, which contrasted with the antagonism usually found between the benzodiazepines and methylxanthines when performance is evaluated by psychomotor tests not requiring fine motor control. The observed synergism was not explicable in terms of measured disposition of the drugs. The synergistic production of fine motor dyskinesia by the concurrent administration of caffeine and midazolam may be relevant to the triggering of anxiety attacks by caffeine observed in panic disorder patients.

Synergism by caffeine and by cocaine of the motor control deficit produced by midazolam

To evaluate the effects of caffeine and cocaine on the impairment of discriminative motor control produced by midazolam, rats were trained to hold a force transducer operated with a paw so that it remained between upper and lower limits of a force band for a continuous 1.5-s period to deliver each food pellet. Acute doses of 3 mg/kg midazolam SC impaired motor performance. Except for one animal, caffeine (10-40 mg/kg IP) had little or no effect on performance, while cocaine (3.75-22.5 mg/kg IP) produced dose-related impairment. When each dose of caffeine was combined with 3 mg/kg midazolam, a marked synergism in motor performance impairment occurred. Cocaine plus midazolam produced mainly an additive synergism. The conspicuous synergistic action of caffeine on the motor control deficit produced by midazolam contrasts with the typical antagonism found between the benzodiazepines and methylxanthines when performance is evaluated by psychomotor tests not requiring fine motor control.

Anxiolytic effect of caffeine and caffeine-clonazepam interaction: evaluation by NaCl solution intake

The administration of drugs with anxiolytic action to rehydrating rats augments the intake of 1.5% NaCl solution. In order to clarify the status of caffeine as an anxiolytic agent and its possible interaction with a benzodiazepine having high potency and efficacy in this regard, caffeine (0.78-100 mg/kg) alone and caffeine (0.78-50 mg/kg) plus clonazepam (0.05 or 0.50 mg/kg) injections (IP) were administered to rehydrating rats prior to 1-hr sessions during which they drank 1.5% NaCl solution. When given alone, caffeine, within a particular dose range, and clonazepam at both doses, augmented NaCl solution intake, but when administered in combination, caffeine antagonized the effects of clonazepam.

Caffeine produces contralateral rotation in rats with unilateral dopamine denervation: comparisons with apomorphine-induced responses

Like the dopamine agonist apomorphine, the methylxanthines caffeine, theophylline and theobromine produced dose-dependent contralateral rotation in rats with unilateral 6-hydroxydopamine denervation, a response considered to be dependent upon dopamine receptors rendered supersensitive. This response was also observed after the injection of the substances into the denervated striatum. Indeed, intrastriatal administration of caffeine into the dopamine denervated striatum produced, dose-dependently (1.0-50.0 micrograms/ul), contralateral rotation. However, while apomorphine produced ipsilateral rotation in rats with unilateral striatal kainic acid lesions, a response considered to be dependent upon normosensitive dopamine receptors, neither caffeine nor theophylline produced rotational responses. As for apomorphine, the rotational behaviour elicited by caffeine (15.0 mg/kg SC) and theophylline (25.0 mg/kg SC) was inhibited by the dopamine antagonists cis-(Z)flupentixol, haloperidol and sulpiride. Nevertheless, despite the fact that cis-(Z)flupentixol was the most potent inhibitor of the caffeine response, no more than 50% inhibition was produced with doses as high as 1.0-10.0 mg/kg SC of cis-(Z)flupentixol. Pretreatment with alpha methyl-p-tyrosine inhibited the rotational response produced by caffeine in 6-OHDA-lesioned animals, but did not significantly modify the apomorphine response. Furthermore, the benzodiazepine diazepam produced a dose-dependent inhibition of the caffeine rotation, but again, the apomorphine response, although qualitatively modified, was not significantly inhibited.

Adenosine analogues. The temperature-dependence of the anticonvulsant effect and inhibition of 3H-D-aspartate release

Following the intraperitoneal administration of the adenosine analogues, 2-chloro-adenosine (1-4 mg/kg) or 5'-N-ethylcarboxamidoadenosine (NECA; 0.01-0.5 mg/kg) to audiogenic DBA/2 mice, there is a potent protection against sound-induced seizures and a simultaneous large (2-5 degrees) reduction in body temperature. The anticonvulsant potency of the adenosine analogues is almost completely abolished by pretreatment with methylxanthines or warming the mice to prevent the adenosine-induced temperature decrease. Adenosine (0.01-1 mM), 2-chloro-adenosine (0.1-1 mM) and NECA (0.1 mM) also significantly inhibit potassium-evoked release of 3H-D-aspartate from rat hippocampal slices. This inhibition is not affected by theophylline (1 mM).

Aminophylline and CGS 8216 reverse the protective action of diazepam against electroconvulsions in mice

Aminophylline (50 and 100 mg/kg) and CGS 8216 (20 and 40 mg/kg) decreased the anticonvulsant potency of diazepam (5 and 10 mg/kg) against electroshock-induced seizures. It should be emphasized that aminophylline moderately affected the protective action of the benzodiazepine at a dose of 5 mg/kg, whereas it was equipotent with CGS 8216 with regard to diazepam at a dose of 10 mg/kg. Consequently, participation of a purinergic component in the anticonvulsant action of diazepam is suggested. On the other hand, the use of aminophylline in epileptic patients suffering from asthma seems unjustified.

Diazepam potentiates GABA-, but not adenosine-mediated, inhibition of neurons of the nigral pars reticulata

Experiments were conducted to assess the relative roles of gamma-aminobutyric acid (GABA) and adenosine in mediating the inhibition of neuronal activity by diazepam injected intravenously. Recent studies have shown that benzodiazepines inhibit, in a dose-dependent manner, the firing of neurons in the substantia nigra pars reticulata. In support of a predominantly GABAergic mechanism for this inhibitory action, a small dose of diazepam (50 micrograms/kg, i.v.), which itself had little effect on cell firing, significantly potentiated the inhibitory responses of neurons of the pars reticulata to muscimol, a potent GABA agonist given intravenously, and significantly and selectively potentiated the inhibition of reticulata neurons by GABA applied iontophoretically. In contrast to their extreme sensitivity to GABAergic inhibition, neurons of the pars reticulata were comparatively insensitive to systemically and iontophoretically administered adenosine-related drugs. However, in those instances when inhibitions could be achieved with iontophoretically applied adenosine-5'-monophosphate, the inhibitory responses were not significantly modified by a 50 micrograms/kg (i.v.) dose of diazepam. These findings, considered in light of differences in GABA and adenosine receptor densities within the substantia nigra, suggest that the benzodiazepine-induced inhibition of neurons of the nigral pars reticulata most likely involves potentiation of GABA but not adenosine-mediated influences.

Benzodiazepine inhibition of nucleoside transport in human erythrocytes

The interaction of several benzodiazepines (BDZs) with the nucleoside transport system of fresh erythrocytes from humans was investigated. The affinities of BDZs for the nucleoside transport system were estimated by measuring BDZ inhibition of (a) the site-specific binding of nitrobenzylthioinosine, a potent and specific inhibitor of nucleoside transport, and (b) the uridine transport processes, zero-trans influx, zero-trans efflux, and equilibrium exchange influx. The BDZs inhibited both the inward and outward transport processes, and, for individual agents, inhibition constants (Ki) were similar for the inhibition of each transport process and for the inhibition of the site-specific binding of nitrobenzylthioinosine. The order of potencies of the BDZs in their interactions with the nucleoside transport mechanism (Ro 5-4864 greater than diazepam greater than clonazepam greater than lorazepam greater than flurazepam) is distinct from the potencies of these compounds at BDZ recognition sites. The affinities of the BDZs for the nucleoside transport system, which are about 1000-fold lower than for BDZ recognition sites, suggest that significant inhibition is unlikely to occur with the plasma concentrations (less than 1 microM) that result from usual anxiolytic doses of these agents.

Benzodiazepine inhibition of adenosine uptake is not prevented by benzodiazepine antagonists

Uptake of [3H]adenosine into rat cerebral cortex synaptosomes was studied. Hexobendine (10(-5) M) and the benzodiazepine agonists diazepam (10(-5) M) and flurazepam (10(-4) M) significantly inhibited this uptake, but only if the compounds were pre-incubated for 10 min in the case of the benzodiazepines. The benzodiazepine antagonists Ro15-1788 (10(-5) M) and CGS 8216 (10(-5) M) failed to reverse the action of benzodiazepine agonists or hexobendine on [3H]adenosine uptake. The results add weight to the view that inhibition of adenosine uptake processes by benzodiazepines do not contribute to their behavioural effects.

Theophylline antagonizes diazepam-induced psychomotor impairment

Eight healthy men received an oral dose of 0.25 mg/kg diazepam followed 40 min later by an intravenous infusion of 100 ml physiological sodium chloride solution, with or without 4.4 mg/kg theophylline. Psychomotor function was assessed after each blood sampling up to 5 h post-infusion. Thirty min after diazepam psychomotor performance measured by Card Sorting test and Digit Symbol Substitution test was impaired and subjects felt sleepy and could think less clearly (two factors of the Clyde Mood Scale). Theophylline antagonized the diazepam-induced impairment statistically significantly for up to 5 h and subjects felt less tense and less apprehensive (State Anxiety Inventory). Since pharmacokinetic parameters of diazepam seemed not to be different after theophylline, interaction at receptor level can be assumed.