Tolerance and cross-tolerance to theophylline-induced stimulation of locomotor activity in rats

Rats treated chronically with caffeine become tolerant to caffeine-induced stimulation of locomotor activity and cross-tolerant to theophylline. This study was performed to determine if the cross-tolerance between these two methylxanthine drugs is symmetrical. Symmetrical cross-tolerance produced by two different drugs implies a common underlying mechanism of action. Separate groups of rats were given scheduled access to drinking bottles containing either drug-free tap water or 1.0 mg/ml theophylline solution. Daily theophylline intake averaged 59 mg/kg. Dose-effect curves were determined in both control and theophylline-treated groups for 5 drugs: the methylxanthines theophylline and caffeine, a nonxanthine psychomotor stimulant, d-amphetamine, and the adenosine analogs R(-)-N6-2-(phenylisopropyl)adenosine and 5'-(N-ethyl)carboxamidoadenosine. All drugs were injected i.p. and locomotor activity was measured for 30 min beginning 35 min later. Rats that were maintained chronically on theophylline were completely tolerant to the locomotor activity stimulant effects of acutely administered theophylline and cross-tolerant to caffeine-induced stimulation of locomotor activity. In contrast, both control and treated groups were fully responsive to the stimulant effects of d-amphetamine. Low doses of the adenosine analogs produced stimulation of locomotor activity in both groups of rats. Higher doses produced a dose-dependent depression of locomotor activity in control rats; curves for the theophylline-treated rats were shifted to the right of the control curves. Thus, adenosine antagonist activity of theophylline remained evident at a time of complete tolerance to the stimulant effect of the drug on locomotor activity.     

Circadian rhythm in adenosine A1 receptor of mouse cerebral cortex

In order to investigate diurnal variation in adenosine A1 receptors binding parameters, Bmax and Kd values of specifically bound N6 - cyclohexyl-[3H]adenosine were determined in the cerebral cortex of mice that had been housed under controlled light-dark cycles for 4 weeks (light on from 7.00 to 19.00 h). Significant differences were found for Bmax values measured at 3-hr intervals across a 24-h period, with low Bmax values during the light period and high Bmax values during the dark period. The amplitude between 03.00 and 18.00 hr was 33%. No substantial rhythm was found in the Kd values. It is suggested that the changes in the density of A1 receptors could reflect a physiologically-relevant mechanism by which adenosine exerts its modulatory role in the central nervous system.     

Caffeine tolerance: behavioral, electrophysiological and neurochemical evidence

The development of tolerance to the stimulatory action of caffeine upon mesencephalic reticular neurons and upon spontaneous locomotor activity was evaluated in rats after two weeks of chronic exposure to low doses of caffeine (5-10 mg/kg/day via their drinking water). These doses are achievable through dietary intake of caffeine-containing beverages in man. Concomitant measurement of [3H]-CHA binding in the mesencephalic reticular formation was also carried out in order to explore the neurochemical basis of the development of tolerance. Caffeine, 2.5 mg/kg i.v., markedly increased the firing rate of reticular neurons in caffeine naive rats but failed to modify the neuronal activity in a group exposed chronically to low doses of caffeine. In addition, in spontaneous locomotor activity studies, our data show a distinct shift to the right of the caffeine dose-response curve in caffeine pretreated rats. These results clearly indicate that tolerance develops to the stimulatory action of caffeine upon the reticular formation at the single neuronal activity level as well as upon spontaneous locomotor activity. Furthermore, in chronically caffeine exposed rats, an increase in the number of binding sites for [3H]-CHA was observed in reticular formation membranes without any change in receptor affinity. We propose, therefore, that up-regulation of adenosine receptors may underlie the development of tolerance to the CNS effects of caffeine.     

The effect of hyperthyroidism on opiate receptor binding and pain sensitivity

This study was conducted to determine the effect of thyroid hormone on opiate receptor ligand-binding and pain sensitivity. Specific opiate receptor-binding was performed on brain homogenates of Swiss-Webster mice. There was a significant increase in 3H-naloxone-binding in thyroxine-fed subjects (hyperthyroid). Scatchard analysis revealed that the number of opiate receptors was increased in hyperthyroid mice (Bmax = 0.238 nM for hyperthyroid samples vs. 0.174 nM for controls). Binding affinity was unaffected (Kd = 1.54 nM for hyperthyroid and 1.58 nM for control samples). When mice were subjected to hotplate stimulation, the hyperthyroid mice were noted to be more sensitive as judged by pain aversion response latencies which were half that of control animals. After morphine administration, the hyperthyroid animals demonstrated a shorter duration of analgesia. These findings demonstrate that thyroxine increases opiate receptor number and native pain sensitivity but decreases the duration of analgesia from morphine.     

Loss of myocardial protection from ischemic preconditioning following chronic exposure to R(-)-N6-(2-phenylisopropyl)adenosine is related to defect at the adenosine A1 receptor

Exogenously administered adenosine agonist will protect myocardium against infarction during ischemia. However, long-term exposure to adenosine agonists is associated with loss of this protection. To determine why this protection is lost, isolated, perfused rabbit hearts were studied after administration of R(-)-N⁶-(2-phenylisopropyl)adenosine (PIA), 0.25 mg/h IP, for 3-4 days to intact animals. All hearts experienced 30 min of regional ischemia and 120 min of reperfusion. Control groups 1 and 2 were untreated. In group 1 this ischemia/reperfusion was the only intervention, whereas group 2 hearts were preconditioned with a cycle of 5 min global ischemia/10 min reperfusion preceding the 30 min regional ischemia. Groups 3-5 had been chronically exposed to PIA. Group 3 hearts had 1 preconditioning ischemia/reperfusion cycle before the prolonged ischemia. Group 4 received a 5 min infusion of 0.1 mol/L phenylephrine in lieu of global ischemia, whereas group 5 was instead treated with 1 mol/L carbachol. Infarct size averaged 32% of the risk zone in group 1, whereas ischemic preconditioning limited infarction to 8.2 in group 2. Prolonged exposure of group 3 hearts to PIA resulted in the inability of preconditioning with 5 min global ischemia to protect (28.7 ± 4.4% infarction). However, protection was restored by either phenylephrine, an agonist of ₁-adrenergic receptors which couple to G_q and stimulate PKC, or carbachol, an agonist of M₂-muscarinic receptors which couple instead to G_i as do adenosine A₁ receptors (5.2 ± 1.7% and 9.2 ± 2.1% infarction, resp.). Therefore, cross tolerance to ischemic preconditioning develops after chronic PIA infusion. Since both the G_i and the PKC components of the preconditioning pathway were shown to be intact, tolerance must have been related to downregulation or desensitization of the A₁ adenosine receptor.     

Possible Involvement of Pertussis Toxin-Sensitive G Proteins and D2 Dopamine Receptors in the A1 Adenosine Receptor-Adenylate Cyclase System in Rat Cerebral Cortex

To identify the involvement of dopamine receptors in the transmembrane signaling of the adenosine receptor-G protein-adenylate cyclase system in the CNS, we examined the effects of pertussis toxin (islet-activating protein, IAP) and apomorphine on A1 adenosine agonist (-)N6-R-[3H]phenylisopropyladenosine ([3H]PIA) and antagonist [3H]xanthine amine congener ([3H]XAC) binding activity and adenylate cyclase activity in cerebral cortex membranes of the rat brain. Specific binding to a single class of sites for [3H]XAC with a dissociation constant (KD) of 6.0 +/- 1.3 nM was observed. The number of maximal binding sites (Bmax) was 1.21 +/- 0.13 pmol/mg protein. Studies of the inhibition of [3H]XAC binding by PIA revealed the presence of two classes of PIA binding states, a high-affinity state (KD = 2.30 +/- 1.16 nM) and a low-affinity state (KD = 1.220 +/- 230 nM). Guanosine 5'-(3-O-thio)triphosphate or IAP treatment reduced the number of the high-affinity state binding sites without altering the KD for PIA. Apomorphine (100 microM) increased the KD value 10-fold and decreased Bmax by approximately 20% for [3H]PIA. The effect of apomorphine on the KD value increase was irreversible and due to a conversion from high-affinity to low-affinity states for PIA. The effect was dose dependent and was mediated via D2 dopamine receptors, since the D2 antagonist sulpiride blocked the phenomenon. The inhibitory effect of PIA on adenylate cyclase activity was abolished by apomorphine treatment. There was no effect of apomorphine on displacement of [3H]quinuclidinyl benzilate (muscarinic ligand) binding by carbachol. These data suggest that A1 adenosine receptor binding and function are selectively modified by D2 dopaminergic agents.     

Inhibition of Purkinje cell firing by systemic administration of phenylisopropyl adenosine: effect of central noradrenaline depletion by DSP4

The effect of the metabolically stable adenosine analog (-)-N6(R-phenyl-isopropyl)-adenosine (PIA) on the rate of spontaneous Purkinje cell firing was studied in anesthetized rats. In control animals, systemically administered PIA elicited only small and inconsistent changes in firing rate. However, in animals previously treated with DSP4 (50 mg/kg i.p.), which selectively lesions central noradrenergic afferents, or with the adrenergic antagonist sotalol (15 mg/kg), PIA elicited consistent decreases in firing rate. These effects were antagonized by the systemic administration of the adenosine receptor antagonist aminophylline (50-150 mumol/kg). Local administration of adenosine by pressure ejection caused a dose-dependent depression of Purkinje cell firing that was likewise inhibited by the methylxanthine. In DSP4 treated rats the depression of synaptic transmission by adenosine in rat hippocampus in vitro was unaltered, and theophylline did not cause any marked rise in Purkinje cell firing, suggesting that DSP4 does not sensitize neurons to the depressant effects of adenosine derivatives. PIA also caused a dose-dependent decrease in arterial blood pressure and a decrease in heart rate that was of equal magnitude in control and DSP4 treated rats. The results show that the central effects of systemically administered adenosine analogs are altered by procedures that disrupt the normal depressant effect of tonic noradrenergic input.     

Solubilized Rat Brain Adenosine Receptors Have Two High-Affinity Binding Sites for l, 3-Dipropyl-8-Cyclopentylxanthine

The specific binding of L-N6-[3H]phenylisopropyladenosine (L-[3H]PIA) to solubilized receptors from rat brain membranes was studied. The interaction of these receptors with relatively low concentrations of L-[3H]PIA (0.5-12.0 nM) in the presence of Mg2+ showed the existence of two binding sites for this agonist, with respective dissociation constant (KD) values of 0.24 and 3.56 nM and respective receptor number (Bmax) values of 0.28 +/- 0.03 and 0.66 +/- 0.05 pmol/mg of protein. In the presence of GTP, the binding of L-[3H]PIA also showed two sites with KD values of 24.7 and 811.5 nM and Bmax values of 0.27 +/- 0.09 and 0.93 +/- 0.28 pmol/mg of protein for the first and the second binding site, respectively. Inhibition of specific L-[3H]PIA binding by 1,3-dipropyl-8-cyclopentylxanthine (DPCPX) (0.1-300 nM) performed with the same preparations revealed two DPCPX binding sites with Ki values of 0.29 and 13.5 nM, respectively. [3H]DPCPX saturation binding experiments also showed two binding sites with respective KD values of 0.81 and 10.7 nM and respective Bmax values of 0.19 +/- 0.02 and 0.74 +/- 0.06 pmol/mg of protein. The results suggest that solubilized membranes from rat brain possess two adenosine receptor subtypes: one of high affinity with characteristics of the A1 subtype and another with lower affinity with characteristics of the A3 subtype of adenosine receptor.     

Adenosine; a physiologic modulator of superoxide anion generation by human neutrophils. Adenosine acts via an A2 receptor on human neutrophils

Adenosine specifically inhibits superoxide anion generation by N-formyl-methionyl-leucyl-phenylalanine-stimulated neutrophils without affecting either degranulation or "aggregation." We present data that also supports the hypothesis that adenosine engages a specific cell surface receptor to mediate inhibition of stimulated neutrophils. Theophylline (10 and 100 mu M), a competitive antagonist at adenosine receptors, reversed the effects of adenosine (0.1 mu M) on superoxide anion generation by stimulated neutrophils. The adenosine analogue 5'N-ethylcarboxamidoadenosine (NECA) was a more potent inhibitor of superoxide anion generation than either N6-phenylisopropyladenosine (PIA) or adenosine, an order of potency consistent with that previously demonstrated for adenosine A2 receptors. 2-Chloroadenosine inhibited superoxide anion generation at concentrations similar to NECA. [3H]-NECA and [3H]-2-chloroadenosine bound to a single receptor on intact neutrophils. The characteristics of the receptors for [3H]-NECA and [3H]-2-chloroadenosine were similar (Kd = 0.22 and 0.23 mu M, respectively; number of binding sites = 9.31 and 11.1 X 10(3) sites/cell, respectively). NECA, 2-chloroadenosine, adenosine, and PIA inhibited binding of [3H]-NECA with a rank order similar to that for inhibition of superoxide anion generation (NECA = 2-chloroadenosine greater than adenosine greater than PIA). There was 50% inhibition of superoxide anion generation by NECA at approximately 20% receptor occupancy. Adenosine, derived from damaged tissues, may serve as a specific, endogenous modulator of superoxide anion generation by activated neutrophils through interaction at this newly described receptor on human neutrophils.     

Effects of CCK antagonists on CCK-induced suppression of locomotor activity in mice.

Sulfated cholecystokinin octapeptide (CCK-8) was administered either intraperitoneally or into the cerebral ventricle of fully conscious mice, and locomotor activity was quantified. CCK-8 administered by either route suppressed locomotor activity. Subcutaneously administered selective CCK-A receptor antagonist, L-364,718 (1 mg/kg), reversed the inhibitory effect of centrally as well as peripherally administered CCK-8, but the selective CCK-B receptor antagonist, L-365,260 (1 mg/kg), did not. These results demonstrate that centrally as well as peripherally administered CCK-8 suppresses locomotor activity in mice through an interaction with CCK-A, but not CCK-B, receptors.     

Caffeine and a selective adenosine A<Subscript>2A</Subscript> receptor antagonist induce sensitization and cross-sensitization behavior associated with increased striatal dopamine in mice

Background  

Caffeine, a nonselective adenosine A₁ and A_2A receptor antagonist, is the most widely used psychoactive substance in the world. Evidence demonstrates that caffeine and selective adenosine A_2A antagonists interact with the neuronal systems involved in drug reinforcement, locomotor sensitization, and therapeutic effect in Parkinson's disease (PD). Evidence also indicates that low doses of caffeine and a selective adenosine A_2A antagonist SCH58261 elicit locomotor stimulation whereas high doses of these drugs exert locomotor inhibition. Since these behavioral and therapeutic effects are mediated by the mesolimbic and nigrostriatal dopaminergic pathways which project to the striatum, we hypothesize that low doses of caffeine and SCH58261 may modulate the functions of dopaminergic neurons in the striatum.     

N6-(Phenylisopropyl)adenosine prevents glucagon both blocking insulin''s activation of the plasma-membrane cyclic AMP phosphodiesterase and uncoupling hormonal stimulation of adenylate cyclase activity in hepatocytes.

Glucagon (10nM) prevented insulin (10nM) from activating the plasma-membrane cyclic AMP phosphodiesterase. This effect of glucagon was abolished by either PIA [N6-(phenylisopropyl)adenosine] (100nM) or adenosine (10 microM). Neither PIA nor adenosine exerted any effect on the plasma-membrane cyclic AMP phosphodiesterase activity either alone or in combination with glucagon. Furthermore, PIA and adenosine did not potentiate the action of insulin in activating this enzyme. 2-Deoxy-adenosine (10 microM) was ineffective in mimicking the action of adenosine. The effect of PIA in preventing the blockade by glucagon of insulin's action was inhibited by low concentrations of theophylline. Half-maximal effects of PIA were elicited at around 6nM-PIA. It is suggested that adenosine is exerting its effects on this system through an R-type receptor. This receptor does not appear to be directly coupled to adenylate cyclase, however, as PIA did not affect either the activity of adenylate cyclase or intracellular cyclic AMP concentrations. Insulin's activation of the plasma-membrane cyclic AMP phosphodiesterase, in the presence of both glucagon and PIA, was augmented by increasing intracellular cyclic AMP concentrations with either dibutyryl cyclic AMP or the cyclic AMP phosphodiesterase inhibitor Ro-20-1724. PIA also inhibited the ability of glucagon to uncouple (desensitize) adenylate cyclase activity in intact hepatocytes. This occurred at a half-maximal concentration of around 3 microM-PIA. However, if insulin (10 nM) was also present in the incubation medium, PIA exerted its action at a much lower concentration, with a half-maximal effect occurring at around 4 nM.     

Effect of Chronic Manganese Treatment on Adenosine Tissue Levels and Adenosine A2a Receptor Binding in Diverse Regions of Mouse Brain

In the present study the effects of chronic manganese (Mn) treatment on adenosine A_2a receptor binding in mouse brain have been assessed. Male albino mice were divided in two groups: In the Mn-treated group, the animals were injected intraperitoneally (i.p.) with MnCl₂ (5 mg/kg/day) five days per week during 9 weeks; in the control group, they were injected likewise with a saline solution. A significant decrease of the Kd without alteration of Bmax in the cerebellum and, an increase of the Kd and Bmax in hippocampus of mice treated with Mn were found. Also, an increase of Kd in frontal cortex was observed. The binding parameters in caudate nucleus, olfactory bulb and hypothalamus were not altered by Mn. A significant decrease in the adenosine concentration in caudate nucleus, olfactory bulb and hypothalamus, without significant changes in hippocampus, frontal cortex and cerebellum was also detected. These findings suggest that chronic administration of Mn could affect adenosine receptor function and turnover, depending on the brain region analyzed.     

Specific photoaffinity labelling of inhibitory adenosine receptors

N6(L-phenylisopropyl)adenosine (L-PIA) and N6(3-iodo-4-azido benzyl)-adenosine (IAzBA) inhibit the adenylate cyclase activity in synaptic membranes of chick cerebellum via Ri adenosine receptors. [3H]L-PIA and [125I]AzBA bind to these membranes with Kd values of approximately 1 nM and Bmax values of approximately 1000 fmol/mg protein. Photolysis of [125I]AzBA bound to synaptic membranes results in the specific incorporation of radioactivity into a protein with Mr = 36,000. This photoincorporation is blocked by simultaneous exposure to L-PIA, theophylline, an adenosine receptor antagonist, or Gpp(NH)p, but not by cytosine, suggesting that the 36,000 dalton protein is the Ri adenosine receptor or a subunit of the receptor that contains the adenosine binding site.     

Modulation of A2A adenosine receptor(s) by K+(ATP) channels in bovine brain striatal membranes

The modulation of adenosine receptor with K+(ATP) channel blocker, glibenclamide, was investigated using the radiolabeled A2A-receptor selective agonist [3H]CGS 21680. Radioligand binding studies in bovine brain striatal membranes (BBM) indicated that unlabeled CGS 21680 displaced the bound [3H]CGS 21680 in a concentration-dependent manner with a maximum displacement being approximately 65% at 10(-4) M. In the presence of 10(-5) M glibenclamide, unlabeled CGS 21680 increased the displacement of bound [3H]CGS 21860 by approximately 28% at 10(-4) M. [3H]CGS 21680 bound to BBM in a saturable manner to a single binding site (Kd = 10.6+/-1.71 nM; Bmax = 221.4+/-6.43 fmol/mg of protein). In contrast, [3H]CGS 21680 showed saturable binding to two sites in the presence of 10(-5) M glibenclamide; (Kd = 1.3+/-0.22 nM; Bmax = 74.3+/-2.14 fmol/mg protein; and Kd = 8.9+/-0.64 nM; Bmax = 243.2+/-5.71 fmol/mg protein), indicating modulation of adenosine A2A receptors by glibenclamide. These studies suggest that the K+(ATP) channel blocker, glibenclamide, modulated the adenosine A2A receptor in such a manner that [3H]CGS 21680 alone recognizes a single affinity adenosine receptor, but that the interactions between K+(ATP) channels and adenosine receptors.     

Antagonism by theophylline of respiratory inhibition induced by adenosine

The effects on respiration of an analogue of adenosine, L-2-N6-(phenylisopropyl)adenosine (PIA), and of the methylxanthine, theophylline, were determined in 19 vagotomized glomectomized cats whose end-tidal PCO2 was kept constant by means of a servo-controlled ventilator. Integrated phrenic nerve activity was used to represent respiratory output. Our results show that PIA, whether given systemically or into the third cerebral ventricle, depressed respiration. Systemically administered theophylline stimulated respiration. Theophylline given intravenously, or into the third ventricle not only reversed the depressive effects of previously administered PIA but caused further increases of respiration above the control level. Prior systemic administration of theophylline blocked both respiratory and hypotensive effects of subsequently administered PIA. Effects of either agent on medullary extracellular fluid pH did not explain the results. We conclude that the adenosine analogue PIA, acts to inhibit neurons in the brain that are involved in the control of respiration and that its effects are blocked by theophylline. We suggest that adenosine acts as a tonic modulator of respiration and that theophylline stimulates breathing by competitive antagonism of adenosine at neuronal receptor sites.     

Adenosine analogs inhibit fighting in isolated male mice

The potent adenosine analogs N-ethylcarboxamide adenosine (NECA) and phenylisopropyladenosine (PIA) inhibit fighting and associated agonistic behaviors in isolated male mice. These effects are reversed by methylxanthines; moderate doses of NECA which inhibit fighting have minimal effects on spontaneous locomotor activity. At very low doses, both NECA and PIA increase fighting in parallel with previously reported increases of motor activity. Brain levels of [3H]-NECA and [3H]-PIA achieved at behaviorally effective doses suggest an involvement of adenosine receptors. The biochemical mechanism of adenosine receptor action with respect to fighting is unknown, but may include neuromodulatory effects on the release of other, more classical neurotransmitters.     

Reduced muscarinic receptor plasticity in frontal cortex of aged rats after chronic administration of cholinergic drugs

Age-related differences in muscarinic receptor plasticity were observed in young, adult and senescent Fischer 344 rats (3, 9 and 27 months old, respectively) following the chronic, intracerebroventricular (ivt) administration of a cholinergic agonist, oxotremorine, or antagonist, methylatropine. After three weeks treatment of young rats with ivt oxotremorine, the maximum number (Bmax) of 3H-QNB binding sites in frontal cortex, determined by saturation experiments, was reduced by 27%, with no apparent change in the affinity (Kd) of 3H-QNB for the muscarinic receptor. Conversely, chronic ivt methylatropine administered to 3 month old animals caused a 29% increase in Bmax with no significant change in Kd. Adult animals showed a somewhat lesser degree of muscarinic receptor plasticity (16% down-regulation after oxotremorine, 22% up-regulation after methylatropine). However, 3H-QNB binding parameters in frontal cortex from senescent rats were not significantly altered following identical treatments with oxotremorine or methylatropine. Thus, muscarinic receptor adaptation to chronic, cholinergic drug administration was impaired in aged animals. This reduced receptor plasticity with aging could have important implications for the long-term drug treatment of elderly patients and for the therapeutic efficacy of cholinergic drugs in age-related neurological disorders, such as Alzheimer's disease.     

Pyrazolo[3,4-d]pyrimidines as adenosine antagonists

A large number of nitrogen heterocycles structurally related to caffeine and theophylline have been tested for activity as adenosine antagonists. Preliminary screening, utilizing displacement of [3H]N6-phenylisopropyladenosine (PIA) binding to rat brain membranes, identified several pyrazolo[3,4-d]pyrimidines with potential antagonist activity. These were then tested for their ability to antagonize adenosine-stimulated adenylate cyclase of guinea-pig slices and to block adenosine receptors which mediate presynaptic inhibition of transmitter release from cholinergic nerves in guinea-pig ileum. Of several compounds found to have antagonist activity, one of these, 4,6-bis-alpha- carbamoylethylthio -1-phenylpyrazolo[3,4-d]pyrimidine ( DJB -KK) was approximately an order of magnitude more potent than theophylline in both tests. GTP greatly reduces the potency of purine agonists, but not antagonists, as inhibitors of [3H] PIA binding; the potency of the pyrazolo[3,4-d]pyrimidine compounds was not altered by GTP. The compounds have no significant activity against [3H]adenosine uptake or on the binding of ligands to muscarinic cholinergic, beta-adrenergic, GABA or L-glutamate receptors.     

Potent convulsant actions of the adenosine receptor antagonist, xanthine amine congener (XAC)

The convulsant properties of xanthine amine congener (XAC, 8-(4-(2-aminoethyl)-aminocarboxylmethyloxy)phenyl-1,3-dipropylxant hine) are compared to those of caffeine. Male Swiss albino mice were infused with convulsants through a lateral tail vein. Convulsion thresholds (i.e. the amount of convulsants required to elicit convulsions) of 39.8 +/- 2.0 mg/kg (n = 10) and 109.8 +/- 2.3 mg/kg (n = 10) were calculated for XAC and caffeine respectively. Pretreatment of animals with the adenosine receptor agonists 2-chloroadenosine, N6-cyclohexyladenosine or 5'-N-ethylcarboxamido-adenosine (1 mg/kg, i.p., 20 minutes prior to infusion) significantly decreased the seizure threshold of both XAC and caffeine. The adenosine uptake blockers, 6-nitrobenzylthioinosine or dipyridamole (0.25 mg/kg, i.p., 20 minutes prior to infusion) did not significantly affect the seizure threshold to either XAC or caffeine. The benzodiazepine agonist diazepam (5 mg/kg, i.p., 20 minutes prior to infusion) significantly increased the seizure threshold to both XAC (p less than 0.05) and caffeine (p less than 0.01), whereas the benzodiazepine antagonist Ro 15-1788 (10 mg/kg, i.p., 20 minutes prior to infusion) significantly increased the seizure threshold to caffeine (p less than 0.01), but not XAC. The results suggest that actions at benzodiazepine receptors may be a tenable hypothesis to explain the convulsant actions of caffeine, but not those of XAC.