- Top of page
Summary: Purpose: Adenosine is a neuromodulator that has been proposed to act as an anticonvulsant mainly via inhibitory A1 receptors, but recent data show that genetic deletion of facilitatory A2A receptors might also attenuate convulsions. Since both A1 and A2A receptors are prone to down- and upregulation in different stressful situations, we investigated if convulsive behavior leads to a long-term change in A1 and A2A receptor density in the rat cerebral cortex.
Methods: Stage 4-5 convulsions (Racine's scale) were induced in adult Wistar rats either through amygdala stimulation (kindling) or by intraperitoneal injection of kainate (10 mg/ml). Rats were killed after 4 weeks to evaluate adenosine A1 and A2A receptor density in the cerebral cortex using both Western blot and membrane binding assays.
Results: The binding density of the A1 antagonist, 3H-DPCPX, decreased by 40. ± 4.4% and by 20.7 ± 0.5% after kindling or kainate injection. Likewise, A1 receptor immunoreactivity in cortical membranes from kindled or kainate-injected rats decreased by 19.1 ± 3.3% and 12.7 ± 5.7%, respectively. In contrast, the binding density of the A2A receptor antagonist 3H-SCH 58261 increased by 293 ± 34% and by 159 ± 32% in cortical membranes from kindled or kainate-injected rats, and A2A receptor immunoreactivity also increased by 151 ± 12% and 79.6 ± 7.0%.
Conclusions: This indicates that after convulsive behavior there is a long-term decrease of A1 receptors accompanied by an increased density of A2A receptors, suggesting that A2A antagonists rather than A1 agonists may be more promising anticonvulsive drugs.
Adenosine is a ubiquitous neuromodulator that mainly inhibits synaptic transmission and neuronal excitability through activation of the predominant adenosine A1 receptors (1). The ability of adenosine to selectively depress glutamatergic excitatory pathways makes A1 receptors interesting anticonvulsant targets [reviewed in (2)]. Accordingly, exogenous administration of A1 receptor agonists attenuates seizures, and the use of A1 receptor antagonists has proconvulsant effects [reviewed in (2)]. It has even been proposed that the loss of the A1 receptor-mediated control of glutamatergic function could contribute to the implementation of epileptic status (3). However, conflicting results have been reported in relation to the impact of convulsive behavior on cortical A1 receptors [reviewed in (4)]. In other brain regions, chronic stressful situations, such as amygdala kindling (5) or Alzheimer's disease (6), decreased the density of A1 receptors.
Adenosine can also activate another less abundant adenosine receptor subtype, i.e., A2A receptors, with effects generally opposite to these mediated by A1 receptors (7). These facilitatory A2A receptors are most abundant in the basal ganglia and have a density 20 times lower in the cerebral cortex (8). However, despite their low abundance, and by mechanisms still to be resolved, pharmacological blockade or genetic inactivation of these A2A receptors confer robust neuroprotection in the limbic and neocortex in different noxious brain situations [reviewed in (9)]. In particular, genetic inactivation of A2A receptors decreases ethanol withdrawal-induced seizures (10), suggesting a role for A2A receptors also in the control of convulsive behavior. Interestingly, studies in nonbrain preparations have documented that prolonged stressful situations, such as hypoxia (11) or exposure to cytokines (12), enhance the expression and density of these A2A receptors, but this has not yet been documented in brain tissue.
Therefore, there is ground to consider both A1 and A2A receptors as possible targets for the development of anticonvulsants. However, since noxious situations lead to changes in the density of both A1 and A2A receptors, it appears necessary to determine if the occurrence of seizures might also lead to changes in A1 and A2A receptor densities. Thus, this study was designed to investigate the long-term changes of the densities of A1 and A2A receptors in the cerebral cortex using two different strategies to induce episodic convulsive behavior: amygdala kindling (13) and intraperitoneal injection of kainate (14).
- Top of page
The neuromodulator adenosine has been associated with epileptic phenomena [reviewed in (2)] since there is a massive amount of adenosine released during seizures (16) and manipulation of the activity of either A1[reviewed in (2)] or A2A receptors (10) affects convulsive behavior. We now found that convulsive behavior triggers a long-term decrease of A1 receptor density and an increase of A2A receptor density in the cerebral cortex. This conclusion was reached using two different and well-established paradigms to trigger convulsive behavior, amygdala kindling (13) and kainate injection (14), and two different experimental approaches to measure the density of adenosine A1 and A2A receptors, i.e., radioligand binding with selective antagonists and using selective antibodies against each adenosine receptor subtype. Thus, the observed remarkable qualitative similarity between the findings obtained when using two different ways to trigger convulsions and two different methods to evaluate changes in adenosine receptor density emphasizes the robustness of this major conclusion, i.e., that convulsive behavior causes a long-term increase in A2A receptor density and a decrease of the density of A1 receptors.
Several studies have shown that A1 receptors undergo desensitization on prolonged activation (17–19), which is expected due to the massive and prolonged increase of extracellular adenosine occurring upon ictal activity (16,20,21). However, previous studies failed to reach an agreement on the effects of seizure activity on the density of A1 receptors in the brain [reviewed in (4)]. In fact, most studies evaluating acute or short-term effects (i.e., after 24 h) of convulsions most commonly found an increase in the density of A1 receptors in different brain regions, either upon induction of seizures with pentylenetetrazole (22,23), bicuculline (24,25), or 3-mercaptopropionate (26). In contrast, no acute or short-term changes in the density of brain A1 receptors were found after electroconvulsive stimulation (27,28) or upon intraperitoneal injection of kainate (29). Changes in the density of A1 receptors occurring 2–15 days after the observation of convulsive behavior also seem to depend on the strategy used to trigger convulsions. Thus, convulsions triggered by pentylenetetrazole (22) or repeated electroconvulsive stimulation (27,28) enhanced the density of A1 receptors in most brain regions, whereas no measurable modification of A1 receptor density was reported upon amygdala kindling (30)] or kainate injection (29,31). Interestingly, most studies seem to find a long-term decrease in the density of adenosine A1 receptors, at least in the hippocampus (5,29,31). Thus, there is a general trend indicating that convulsive behavior leads to a long-term decrease in the density of A1 receptors, as we now report has occurred in the cerebral cortex. This conclusion is in general agreement with the development of tolerance in relation to the anticonvulsive effects of A1 receptor agonists (3,32).
However, the currently observed decrease of A1 receptor density does not necessarily exclude A1 receptors as potential targets for the development of new anticonvulsant drugs. In fact, while the efficiency of most anticonvulsants decreases markedly with increasing severity of seizures, A1 receptor activation is able to suppress seizures in an animal model of pharmacoresistant epilepsy (33). This may be understood if one keeps in mind that seizures cause a long-term greater modification of the extracellular levels of adenosine able to activate A1 receptors than a desensitization of responses mediated by A1 receptors [(5), reviewed in (4)]. This suggests that strategies aimed at increasing the extracellular levels of adenosine may be more effective than using A1 receptor agonists (34,35), which also have the disadvantages of poor brain penetration and causing potent peripheral side effects (36).
In contrast to the well-defined potential of A1 receptor activation to control seizures, the role of A2A receptors is less well established. In fact, some reports with purportedly selective A2A receptor agonists consistently observed anticonvulsant effects (32,37–39). However, despite the established presence of low amounts of A2A receptors in cortical brain regions, it has recently been concluded that A2A receptor agonists (like CGS 21680) mostly bind to A1 rather than to A2A receptors in cortical regions (8). This suggests that the activation of A1 receptors might underlie the anticonvulsive effects of these purportedly selective A2A receptor agonists (40). However, there are reasons to believe that A2A receptors might play a role in the control of seizures and/or epileptogenesis. In fact, the group of Vaugeois showed that pharmacological blockade or genetic deletion of A2A receptors decrease ethanol withdrawal-induced seizures in mice (10). Likewise, it has systematically been shown that blockade of A2A receptors is neuroprotective in different noxious situations that involve limbic or cortical degeneration [reviewed in (9)], as is the case of temporal lobe epilepsy (41). This preliminary evidence, together with the presently observed robust increase in the density of adenosine A2A receptors in the cerebral cortex of rats that had undergone a convulsive period, make A2A receptor antagonists an attractive novel class of anticonvulsive drugs.
In conclusion, the presently observed long-term decrease in the density of A1 receptors and the parallel increase of the density of A2A receptors in the cerebral cortex provide a preliminary rationale for the development of novel anticonvulsive strategies targeting the adenosine neuromodulatory systems, which might be based on the combined use of strategies to burst the extracellular levels of adenosine to activate inhibitory A1 receptors together with antagonists of facilitatory A2A receptors.