Low dose of bupropion significantly enhances the anticonvulsant activity of felbamate, lamotrigine and topiramate in mice.

Experimental evidence indicates that bupropion hydrochloride, an antidepressant and a first-line smoking cessation aid, exerts dose-dependently anticonvulsant and convulsant effects. In this study, chronic bupropion pretreatment intraperitoneally (i.p.) for 14 days in a dose of 5 mg/kg reduced the ED(50) (i.e. the dose protecting 50% of mice against electroconvulsions) of lamotrigine, topiramate, and felbamate from 4.58, 60.95, and 48.79 (antiepileptic+vehicle) to 3.01, 41.68, and 37.28 mg/kg (antiepileptic+bupropion), respectively, against maximal electroshock-induced seizures in mice. Bupropion significantly increased the plasma and brain concentrations of lamotrigine. Plasma concentration of topiramate was elevated, however, the brain concentration of the drug was not affected. Neither plasma nor brain concentrations of felbamate were elevated by bupropion administration. Bupropion did not exacerbate motor coordination impairment caused by the antiepileptic drugs in the rotarod test. Chronic administration of bupropion significantly potentiates the protective activity of lamotrigine, topiramate, and felbamate against maximal electroshock-induced seizures. A pharmacokinetic interaction is responsible for the effect of bupropion co-administered with lamotrigine.

Effects of androsterone on convulsions in various seizure models in mice.

It is believed that a deficiency of androgens, including free testosterone, may promote the development of convulsions. The present study revealed differences in the action of androsterone (AND), a major excreted metabolite of testosterone and a neurosteroid, in three commonly used seizure models in mice. AND administered intraperitoneally exhibited dose-dependent protection against tonic-clonic convulsions caused by maximal electroshock (MES) with ED(50) (effective dose(50)) of 227 mg/kg. The compound also inhibited the convulsive action of pentylenetetrazole (PTZ), increasing its CD(50) (convulsive dose(50)) for clonic convulsions from 77.2 (PTZ + saline) to 93.9 (p < 0.05) for PTZ + AND 40 mg/kg and 113.9 mg/kg (p < 0.001) for PTZ + AND 60 mg/kg. In mice pretreated with 60 mg/kg AND, the CD(50) for PTZ-induced tonic convulsions increased from 102 to 127.6 mg/kg (p < 0.01). Surprisingly, doses of 50 and 100 mg/kg AND lowered the CD(50) for kainate (KA)-induced convulsions from 40.8 to 28.7 (p < 0.05) and 25.4 mg/kg (p < 0.001), respectively. In summary, for two of the mouse seizure models, our findings confirmed previous studies that demonstrated protective activity of AND. However, the potentiation of KA-induced convulsions by AND was somewhat unexpected and suggested that AND may also possess proconvulsant activity.

beta-adrenergic enhancement of brain kynurenic acid production mediated via cAMP-related protein kinase A signaling.

The central levels of endogenous tryptophan metabolite kynurenic acid (KYNA), an antagonist of N-methyl-d-aspartate (NMDA) and alpha7-nicotinic receptors, affect glutamatergic and dopaminergic neurotransmission. Here, we demonstrate that selective agonists of beta(1)-receptors (xamoterol and denopamine), beta(2)-receptors (formoterol and albuterol), alpha- and beta-receptors (epinephrine), 8pCPT-cAMP and 8-Br-cAMP (analogues of cAMP) increase the production of KYNA in rat brain cortical slices and in mixed glial cultures. Neither betaxolol, beta(1)-adrenergic antagonist, nor timolol, a non-selective beta(1,2)-adrenergic antagonist has influenced synthesis of KYNA in both paradigms. In contrast, KT5720, a selective inhibitor of protein kinase A (PKA), strongly reduced KYNA formation in cortical slices (2-10 microM) and in glial cultures (100 nM). beta-adrenergic antagonists and KT5720 prevented the beta-adrenoceptor agonists-induced increases of KYNA synthesis. In vivo, beta-adrenergic agonist clenbuterol (0.1-1.0 mg/kg) increased the cortical endogenous level of KYNA; the effect was blocked with propranolol (10 mg/kg). beta-adrenoceptors agonists, cAMP analogues and KT5720 did not affect directly the activity of KAT I or KAT II measured in partially purified cortical homogenate. In contrast, the exposure of intact cultured glial cells to pCPT-cAMP, 8-Br-cAMP and formoterol has lead to an enhanced action of KATs. These findings demonstrate that beta-adrenoceptor-mediated enhancement of KYNA production is a cAMP- and PKA-dependent event. PKA activity appears to be an essential signal affecting KYNA formation. Described here novel mechanism regulating KYNA availability may be of a potential importance, considering that various stimuli, among them clinically used drugs, activate cAMP/PKA pathway, and thus could counteract the central deficits of KYNA.

Clenbuterol enhances the production of kynurenic acid in brain cortical slices and glial cultures.

The effect of a beta(2)-adrenergic agonist, clenbuterol on the production of a glutamate receptor antagonist, kynurenic acid was studied in vitro. Clenbuterol enhanced the production of kynurenic acid in brain cortical slices (0.1-1.0 mM) and in glial cultures (1-50 muM). Timolol, a non-selective beta-adrenergic antagonist prevented this effect. The presented data indicate a novel mechanism of action of beta(2)-adrenoceptor agonists and suggest that an increased formation of the endogenous glutamate receptor antagonist, kynurenic acid could partially contribute to their neuroprotective activity.

Memantine increases brain production of kynurenic acid via protein kinase A-dependent mechanism.

We describe a novel aspect of action of memantine ex vivo, in the brain cortical slices and in vitro, in mixed glial cultures. The drug potently increased the production of kynurenic acid, an endogenous tryptophan metabolite blocking N-methyl-D-aspartate (NMDA) and nicotinic alpha7 receptors. In cortical slices memantine, an open-channel NMDA blocker (100-150 microM), but not the competitive NMDA receptor antagonist, LY235959 increased the production of kynurenic acid. Neither SCH23390, D1 receptor antagonist (50 microM) nor raclopride, D2 receptor antagonist (10 microM) changed the memantine-induced effects. Propranolol (100 microM) has partially reduced its action. Selective cAMP-dependent protein kinase (PKA) inhibitor, KT5720 (1 microM), but not selective protein kinase C (PKC) inhibitor, NPC15437 (30 microM) totally reversed the action of memantine. In mixed glial cultures, 2-24 h incubation with memantine (2-50 microM) enhanced the production of kynurenic acid. Memantine (up to 0.5 mM) has not affected the activity of kynurenic acid biosynthetic enzymes. The obtained data suggest that memantine enhances the production of kynurenic acid in PKA-mediated way. This effect may partially contribute to the therapeutic actions of memantine and be of a potential clinical importance.

Enhanced glutamatergic transmission reduces the anticonvulsant potential of lamotrigine but not of felbamate against tonic-clonic seizures.

The efficacy of lamotrigine and felbamate against maximal electroshock (MES)-induced seizures was assessed under conditions mimicking the pharmacoresistance associated with an increased excitatory neurotransmission. N-methyl-D-aspartate (NMDA), but not kainate applied at subconvulsive dose, reduced the activity of lamotrigine against MES-induced seizures increasing its ED50 value from 4.3 (3.2-5.6) to 6.1 (5.2-7.2) mg/kg (p < 0.001). This effect was reversed by co-application of an NMDAreceptor antagonist D-(E)-2-amino-4-methyl-5-phosphono-3-pentenoic acid (CGP 40116) at 0.1 mg/kg [4.5 (3.7-5.6) vs. 6.1 (5.2-7.2) mg/kg; p < 0.001]. The anticonvulsive action of felbamate was altered by neither NMDAnor kainate. In conclusion, the data presented here indicate that felbamate, but not lamotrigine, effectively prevents generalized tonic-clonic seizures, also when NMDA-mediated neurotransmission is enhanced. The impaired antiepileptic potential of lamotrigine might be restored in such scenario by the co-administration of a very low dose of NMDA receptor antagonist.

Different effects of nitric oxide synthase inhibitors on convulsions induced by nicotine in mice.

Acute intraperitoneal (i.p.) administration of N(G)-nitro-L-arginine (NNA, 10, 20 and 40 mg/kg), a non-selective nitric oxide synthase (NOS) inhibitor, significantly and dose-dependently decreased the incidence of convulsions induced by i.p. nicotine (NIC) in mice, whereas 7-nitroindazole (7NI, 50 and 100 mg/kg i.p.), a selective neuronal NOS inhibitor, had a proconvulsant effect. Aminoguanidine (100 mg/kg ip), a specific inducible NOS inhibitor, remained without an effect on convulsive behavior. L-arginine, a nitric oxide (NO) precursor, which independently has no effect on convulsions, markedly reversed the anticonvulsant effect of NNA; yet only partially reversed the proconvulsant effect of 7NI when injected at 500 mg/kg i.p.. Convulsions evoked by intracerebroventricular injection of NIC were significantly suppressed by ip NNA(40 mg/kg i.p.) and enhanced by i.p. 7NI (100 mg/kg i.p.); however, these effects of NNA and 7NI were less potent than those seen when NIC was administered i.p.. The present study revealed essential differences in the action of NOS inhibitors in NIC-induced convulsions. It appears that only NO produced by constitutive NOS is involved in the mechanism of NIC-induced convulsions. The proconvulsant effect of 7NI may result from the mechanisms unrelated to NOS inhibition.

Hyperglycemia enhances the inhibitory effect of mitochondrial toxins and D,L-homocysteine on the brain production of kynurenic acid.

We have evaluated the effect of diabetes-mimicking conditions on the inhibition of kynurenic acid (KYNA) production exerted by mitochondrial toxins: 3-nitropropionic acid (3-NPA) and aminooxyacetic acid (AOAA), by endogenous agonists of glutamate receptors: L-glutamate and L-cysteine sulfinate, and by a risk factor of atherosclerosis, D,L-homocysteine. Hyperglycemia (30 mM; 2 h) itself did not influence KYNA synthesis in brain cortical slices. However, it significantly enhanced the inhibitory effects of 3-NPA, AOAA and D,L-homocysteine, but not of L-glutamate and L-cysteine sulfinate, on KYNA production. Their IC(50) values were lowered from 5.8 (4.5-7.4) to 3.7 (3.1-4.5) mM (p < 0.01), from 11.6 (8.6-15.5) to 7.1 (4.9-10.3) microM (p < 0.05), and from 4.5 (3.5-5.8) to 2.4 (1.8-3.2) mM (p < 0.01), respectively. The obtained data suggest that during hyperglycemia, the mitochondrial impairment and high levels of D,L-homocysteine evoke stronger inhibition of KYNAsynthesis what may further exacerbate brain dysfunction and play a role in central complications of diabetes.

Kynurenic acid in human saliva--does it influence oral microflora?

Kynurenic acid (KYNA) is an endogenous antagonist of alpha7 nicotinic receptors and all ionotropic glutamate receptors. Its neuroprotective activity has been suggested. In this study, the presence of KYNAin human saliva and its potential bactericidal role was investigated. KYNAwas found in all samples of human saliva with mean concentration of 3.4 nM. The concentration of KYNA in saliva obtained from patients with odontogenic abscesses was 3.5 times higher than in healthy subjects. We have shown that the human gingival fibroblasts produce KYNAand an inflammatory stimulant, lipopolysaccharide, enhanced its synthesis in vitro. The bactericidal effect of KYNA was also presented. We hypothesize that KYNA may contribute to the control of oral microflora.

Enhancement of brain kynurenic acid production by anticonvulsants--novel mechanism of antiepileptic activity?

In this study, we describe the effect of antiepileptic drugs on the production of kynurenic acid in rat cortical slices, and on the activity of kynurenic acid biosynthetic enzymes, kynurenine aminotransferases (KATs I and II) in the brain tissue. Phenobarbital, felbamate, phenytoin and lamotrigine (all at 0.5-3.0 mM) enhanced kynurenic acid production in vitro, and stimulated the activity of KAT I. In contrast, vigabatrin, gabapentin and tiagabine inhibited kynurenic acid synthesis in cortical slices with IC(50) of 3.9 (2.8-7.9), 3.7 (2.5-5.4) and 7.5 (3.5-14.3) mM, respectively. Vigabatrin, gabapentin and tiagabine reduced also the activity of KAT I with IC(50) of 1.6 (1.1-2.4), 0.1 (0.01-0.15), 0.9 (0.7-1.2) mM, and the activity of KAT II with IC(50) values of 6.0 (4.8-7.5), 0.2 (0.1-0.3) and 2.0 (1.5-2.6) mM, respectively. In conclusion, the enhancement of kynurenic acid formation displayed by carbamazepine, phenytoin, phenobarbital, felbamate and lamotrigine seems to be a novel mechanism, synergistic with other actions of these drugs, and potentially valuable in terms of better control of epilepsy.

Influence of cimetidine on the anticonvulsant activity of conventional antiepileptic drugs against pentetrazole-induced seizures in mice.

The aim of this study was to evaluate the effects of acute (1 day) and chronic (7 days) administrations of cimetidine, an H2 histamine receptor antagonist, on the protective activity of conventional antiepileptic drugs (AEDs) against pentetrazole (PTZ)-induced seizures in mice. Cimetidine (up to 100 mg/kg), given alone either acutely or chronically, did not alter significantly PTZ-induced seizures in mice. However, the drug (at 20 mg/kg, administered acutely) potentiated the anticonvulsant activity of ethosuximide (ETX) by reducing its ED50 from 134 to 103 mg/kg (p < 0.05). This effect was associated with a 74% elevation of plasma ETX level (p < 0.01). In contrast, chronic (7 days) administration of cimetidine (20 mg/kg) did not affect the anticonvulsant activity of ETX in the PTZ test and its plasma levels. On the other hand, cimetidine (20 mg/kg), given either acutely or chronically, when co-administered with valproate, clonazepam, and phenobarbital had no significant impact on the anticonvulsant properties of these AEDs against PTZ-induced seizures and their plasma levels in mice. The results indicate that there may be no risk in prescribing cimetidine for other than epilepsy reasons in patients treated with valproate, clonazepam or phenobarbital.

Dexamethasone reduces locomotor stimulation induced by dopamine agonists in mice.

The interaction between glucocorticosteroids and the dopaminergic system has attracted considerable attention in recent years since this link could be involved in certain psychopathological conditions including depression. Radioligand binding studies have shown the presence of glucocorticoid receptors in neurons of the limbic system, a structure involved in mood control and subtle regulation of hypothalamic-pituitary-adrenal (HPA) axis. Structures of the limbic system are also rich in dopaminergic innervation. It has been hypothesized that glucocorticoids may be important in causing and perpetuating depression. The aim of the present study was to investigate the effect of dexamethasone (DEX) on hyperactivity induced by dopamine agonists (amphetamine, amantadine, quinpirole and bromocriptine) in mice. Male Albino Swiss mice received DEX at a single dose (2, 4 or 8 mg/kg) or for 14 days at the doses of 0.5, 2 or 4 mg/kg/day. After a single or the last injection (in the chronic experiment) of DEX, dopamine agonists were given in the following regimen: D-amphetamine (0.4 mg/kg) and quinpirole (3 mg/kg) - 30 min, amantadine (50 mg/kg) - 60 min and bromocriptine (10 mg/kg) 180 min before the measurement of locomotor activity. The obtained results show that DEX may decrease the locomotor activity and reduce the hyperactivity induced by dopamine agonists in mice. These observations may suggest that DEX weakens the activity of dopamine agonists in the mesolimbic system.

Effect of histamine receptor antagonists on aminophylline-induced seizures and lethality in mice.

The aim of this study was to evaluate the effects of H(1) (antazoline and astemizole) or H(2) (cimetidine and famotidine) histamine receptor antagonists on the clonic phase, tonic seizures and morality of mice challenged with aminophylline to induce convulsions in mice. Moreover, the total plasma and brain concentrations of theophylline were evaluated. Astemizole (1 mg/kg) did not affect the threshold for aminophylline-induced seizures, but when administered at a dose of 2 mg/kg, it significantly reduced the CD(50) value of aminophylline from 249 mg/kg to 211 mg/kg (p < 0.01). The remaining histamine receptor antagonists studied i.e., antazoline (up to 1 mg/kg), cimetidine (up to 40 mg/kg) and famotidine (up to 10 mg/kg) had no impact on seizure susceptibility in aminophylline-induced convulsions. Furthermore, astemizole (2 mg/kg) decreased latency to the clonic phase of aminophylline-induced convulsions from 51.1 +/- 4.5 to 32.1 +/- 4.3 min (p < 0.01). It is noteworthy that astemizole, a novel H(1) receptor antagonist, did not alter the brain and plasma levels of theophylline, so the existence of pharmacokinetic interactions was excluded. Our results indicate that some interactions between methylxanthines and histamine receptor antagonists may be clinically important since these drugs are usually combined during the treatment of status asthmaticus.

Effects of some convulsant agents on the protective activity of topiramate against maximal electroshock-induced seizures in mice.

The anticonvulsant activity of topiramate combined with some convulsant agents (bicuculline--BIC, N-methyl-D-aspartate--NMDA, and kainic acid--KA), given at subconvulsive doses, was evaluated in the maximal electroshock (MES)-test in mice. BIC (1.5 mg/kg), KA (10 mg/kg) and NMDA (50 mg/kg) significantly decreased the anticonvulsant activity of topiramate raising its ED(50) from 76.2 mg/kg to 135, 102, and 107 mg/kg, respectively. BIC (0.75 mg/kg) and KA (5 mg/kg) did not alter the protective activity of topiramate in the MES-test. Moreover, topiramate injected alone (up to 135 mg/kg) did not affect motor performance and long-term memory of animals tested in the chimney and passive avoidance tests, respectively. In contrast, combinations of topiramate with BIC (1.5 mg/kg), NMDA (50 mg/kg) or KA (10 mg/kg) considerably disturbed long-term memory in mice. Additionally, co-administration of topiramate with KA (10 mg/kg) or BIC (1.5 mg/kg) significantly impaired motor performance, whereas topiramate co-administered with NMDA (50 mg/kg) had no impact on motor coordination in mice. None of the studied convulsants affected the free plasma concentration of topiramate assayed with immunofluorescence method. The results of this study seem to indicate the expression of the anticonvulsant activity of topiramate is dependent on all ionotropic glutamate and GABAA receptor-mediated events.

Pharmacokinetics and metabolism of nicotine.

Nicotine (NIC), the major constituent of tobacco, is responsible for the compulsive use of tobacco. Advances in understanding of the pharmacokinetics and metabolism of NIC have been made rapidly over the past decade. The application of highly sensitive gas chromatography/mass spectrometry led to the identification and quantitation of new NIC metabolites as well as characterization of new pathways of NIC biotransformation. This review summarizes findings from human and animal studies concerning NIC kinetics and biotransformation as well as describes the factors that influence these processes. Recently, large individual, racial and species differences in the metabolism of NIC have been well documented. The differences in the metabolism of NIC may be a result of genetic, environmental, and developmental host influences. We review the scientific evidence from studies that supports a role for genetic mechanisms responsible for variability in the profile and the rate of the NIC metabolism. Actually, the majority of the genetic studies focus on the characterization of the CYP2A6 gene polymorphism, and on determining the relationship between the phenotype of NIC metabolism and the genotype of the CYP2A6 gene. There is good evidence that genetic polymorphisms associated with NIC metabolism are an important factor responsible for susceptibility to NIC dependence. It is anticipated that genetic findings can lead to the identification of individuals at a greater risk for tobacco addiction and will be used for more effective treatment and prevention strategies to reduce smoking.

Bupropion-induced convulsions: preclinical evaluation of antiepileptic drugs.

Bupropion, a unique, non-nicotine smoking cessation aid and an effective antidepressant, is well known to produce seizures following overdosing in humans. However, the experimental background for the usefulness of antiepileptic drugs in the protection against bupropion-induced convulsions has not been established yet. Therefore, we tested if the antiepileptic drugs were able to protect mice against clonic convulsions induced by intraperitoneally (i.p.) administered bupropion in the CD97 dose (139.5 mg/kg). Among 13 tested drugs, clonazepam showed the greatest potency (dose-dependent full protection; ED50 = 0.06 mg/kg, i.p.). No signs of locomotor impairment were observed in the rotarod test after anticonvulsive doses of clonazepam, resulting in a broad therapeutic window and favorable protective index (PI) (33.3). Gabapentin produced dose-dependent protection against convulsions at nontoxic doses (up to 1000 mg/kg), having PI>29. Diazepam in a very high dose showed full protection but its PI (1.7) was much less favorable than that of clonazepam. The PI values for ethosuximide, phenobarbital and valproate were slightly higher than unity and lower than 2, and for topiramate and felbamate were lower than unity. Phenytoin, carbamazepine, and lamotrigine as well as tiagabine failed to block the convulsant effects of bupropion even at doses that caused severe motor impairment. Our results encourage clinical testing of clonazepam against seizures developing after bupropion overdose.

Adenosine A1 receptors and the anticonvulsant potential of drugs effective in the model of 3-nitropropionic acid-induced seizures in mice.

The role of adenosine A1 receptors in the activity of drugs and substances protecting against seizures evoked by mitochondrial toxin, 3-nitropropionic acid (3-NPA) was studied in mice. Non-selective A1/A2 adenosine receptor antagonist, aminophylline and selective A1 adenosine receptor antagonist, 8-cyclopentyl-1,3-dipropylxanthine (DPCPX) diminished the anticonvulsive effects of diazepam, phenobarbital, valproate and gabapentin. In contrast, A1/A2 adenosine receptor antagonist, 8-(p-sulfophenyl)theophylline (8pSPT) not penetrating via blood-brain barrier was ineffective. Aminophylline and DPCPX but not 8pSPT also reversed the protective action of A1/A2 adenosine receptor agonist, 2-chloroadenosine (2-CADO) and selective A1 adenosine receptor agonist, R-N6-phenylisopropyloadenosine (R-PIA), against 3-NPA-evoked convulsions. Obtained results suggest that the central adenosine A1 receptor stimulation may play a role in the anticonvulsive potential of diazepam, phenobarbital, valproate and gabapentin in a novel model of 3-NPA-evoked seizures. Moreover, concomitant application of aminophylline with these drugs may reduce their clinical antiepileptic efficacy, especially among patients suffering from seizures related to the disturbances of mitochondrial respiratory chain.

Effect of chronic treatment with dexamethasone on brain dopamine receptors in mice.

Glucocorticoids are expressed in the central nervous system. Radioligand binding studies have shown their presence in the neurons of the limbic system, a structure involved in mood control and subtle regulation of hypothalamic-pituitary-adrenal (HPA) axis. Structures of the limbic system are also rich in dopaminergic innervation. It has been hypothesized that glucocorticoids may be important in causing and perpetuating depression. Our previous study has demonstrated that dexamethasone decreases the locomotor activity of mice and counteracts the hyperactivity induced by agonists of dopamine receptors. The aim of the present study was to find the possible mechanism responsible for these behavioral effects of dexamethasone. So we sought to examine the influence of chronic dexamethasone treatment on selective radioligand binding to dopamine D(1) ([(3)H]SCH 23390) and D(2) ([(3)H]spiperone) receptors in the brain of mice. The male Albino Swiss mice received dexamethasone (4, 8 or 16 mg/kg/day) for 14 days. The striatum and limbic system structures were isolated and the binding procedure was performed 3.5 or 48 h after the last injection. It was shown that 3.5 h after the last dose of dexamethasone (4 mg/kg/day), specific D(2) receptor binding was statistically significantly increased (by 64%) in the limbic system. On the contrary, the tendency to the reduction of specific D(2) receptor binding was observed in the striatum. Dexamethasone treatment did not influence the specific binding to D(1) receptors in any structure of the brain.

Convulsant and anticonvulsant effects of bupropion in mice.

This study demonstrated that bupropion hydrochloride, an effective antidepressant and a commonly used smoking cessation aid, dose-dependently caused clonic convulsions in mice, with the CD50 (convulsive dose50, i.e., the dose producing convulsions in 50% of mice) at 119.7 mg kg(-1). An evaluation for anticonvulsant effects showed that bupropion in the doses of 15-30 mg kg(-1) protected against convulsions induced by maximal electroshock with the ED50 (effective dose50, i.e., the dose protected 50% of mice against convulsions) being 19.4 mg kg(-1). Bupropion had no effect on pentylenetetrazole- and kainic acid-induced convulsions. It is possible that the anticonvulsant activity of bupropion may be exploited for use in the treatment of epilepsy but it requires further investigations.

Carbamazepine enhances brain production of kynurenic acid in vitro.

Disturbed formation of kynurenic acid, an endogenous antagonist of glutamate ionotropic receptors, might contribute to the pathogenesis of seizures. Here, the effect of anticonvulsant drug, carbamazepine on the production of kynurenic acid was studied. Carbamazepine (0.5-3 mM) enhanced kynurenic acid synthesis in rat cortical slices and also increased the activity of kynurenine aminotransferase (KAT) I at 0.1-3.0 mM concentration. Thus, anticonvulsant drugs, such as carbamazepine, might act partially via stimulation of kynurenic acid production.