Region-Specific Overexpression of P-glycoprotein at the Blood-Brain Barrier Affects Brain Uptake of Phenytoin in Epileptic Rats

Recent studies have suggested that overexpression of the multidrug transporter P-glycoprotein (P-gp) in the hippocampal region leads to decreased levels of antiepileptic drugs and contributes to pharmacoresistance that occurs in a subset of epileptic patients. Whether P-gp expression and function is affected in other brain regions and in organs that are involved in drug metabolism is less studied. Therefore, we investigated P-gp expression in different brain regions and liver of chronic epileptic rats, several months after electrically induced status epilepticus (SE), using Western blot analysis. P-gp function was determined by measuring phenytoin (PHT) levels in these brain regions using high-performance liquid chromatography, in the absence and presence of a P-gp-specific inhibitor, tariquidar (TQD). In addition, the pharmacokinetic profile of PHT was determined. PHT concentration was reduced by 20 to 30% in brain regions that had P-gp overexpression (temporal hippocampus and parahippocampal cortex) and not in brain regions in which P-gp expression was not changed after SE. Inhibition of P-gp by TQD significantly increased the PHT concentration, specifically in regions that showed P-gp overexpression. Despite increased P-gp expression in the liver of epileptic rats, pharmacokinetic analysis showed no significant change of PHT clearance in control versus epileptic rats. These findings show that overexpression of P-gp at the blood-brain barrier of specific limbic brain regions causes a decrease of local PHT levels in the rat brain.

Anticonvulsant drugs differentially suppress individual ictal signs: A pharmacokinetic/pharmacodynamic analysis in the cortical stimulation model in the rat

Antiepileptic drugs can suppress seizures completely, but they may also modify the appearance of drug-resistant seizures. In this study, the effects of three antiepileptic drugs on a seizure pattern were assessed by means of population pharmacokinetic/pharmacodynamic (PK/PD) modeling, yielding estimates of baseline response, EC50, and Hill slope. Lamotrigine did not affect eye closure, although it did suppress the other ictal signs in a concentration-dependent fashion. Midazolam suppressed forelimb clonus less potently than the other ictal signs; the same was observed for tiagabine with respect to eye closure. This study shows that ictal component analysis (ICA) in combination with PK/PD modeling may facilitate drug selection and dose optimization. The application of ICA is not restricted to a single seizure type or anticonvulsant drug and can be used to identify drug combinations that have a complementary action.

Pharmacokinetic-pharmacodynamic modelling of tiagabine CNS effects upon chronic treatment in rats: lack of change in concentration-EEG effect relationship

The pharmacodynamics of the gamma-aminobutyric acid (GABA) uptake inhibitor (R)-N-(4,4-di-(methylthien-2-yl)but-3-enyl) nipecotic acid (tiagabine) was quantified in rats following chronic (14 days) administration by an integrated pharmacokinetic-pharmacodynamic (PK/PD) modelling approach. The increase in beta activity (11.5-30 Hz) of the EEG as derived by fast Fourier transformation analysis was used as pharmacodynamic endpoint. Two groups of male Wistar rats were treated for 14 days with either tiagabine at a steady-state concentration of 198+/-10 ng ml(-1) or placebo. Chronic treatment with tiagabine resulted in an increase of the EEG effect parameter by 38+/-2 microV. In the PK/PD experiment the time course of the EEG effect was determined in conjunction with the decline of drug concentrations after an i.v. bolus administration of 10 mg kg(-1). The pharmacokinetics of tiagabine was most adequately described by a bi-exponential function. No influence of chronic treatment on the pharmacokinetics was observed. Hysteresis between plasma concentration and EEG effect was accounted for by incorporation of an 'effect-compartment' in the model. The observed relationship between tiagabine concentrations and EEG effect was non-linear and described on the basis of the Hill equation. Between the treatment groups no differences in the pharmacodynamic parameters were observed. The population means for the different pharmacodynamic parameters were: maximum EEG effect 82 microV, EC(50) 486 ng ml(-1), Hill factor 2.0 and k(e0) 0.060 min(-1). In the in vitro [(3)H]GABA uptake assay no changes in affinity or functionality for the GABA uptake transporter were observed, consistent with the absence of adaptation. It is concluded that chronic treatment with tiagabine in an effective dose range for 14 days does not produce functional adaptation to tiagabine-induced GABA-ergic inhibition in vivo.

Effect of amygdala kindling on the central nervous system effects of tiagabine: EEG effects versus brain GABA levels

1. The objective of this investigation was to determine the influence of amygdala kindling on the pharmacodynamics of tiagabine in vivo, using quantitative EEG parameters and extracellular GABA concentrations as pharmacodynamic endpoints. In integrated pharmacokinetic/pharmacodynamic (PK/PD) studies the time course of these effects was determined in conjunction with plasma concentrations following intravenous administration of 10 mg kg(-1). An 'effect compartment' model was used to derive individual concentration - effect relationships. 2. ++Tiagabine produced an increase in the amplitude of the 11.5 - 30 Hz frequency band of the EEG. The relationship between concentration and EEG effect was non-linear and described by the Hill equation. 3. In kindled rats the EC(50) was reduced to 291 ng ml(-1) from the original value of 521 ng ml(-1) in controls. The values of all other parameters were unchanged. In kindled rats the baseline extracellular GABA concentration was increased to 1.58 microM from 0.74 microM in controls. The relationships between tiagabine concentration and extracellular GABA concentration were again non-linear and described by the Hill equation. No differences were observed between kindled rats and controls. In the synaptoneurosmal preparation in vitro no changes in the functioning of the GABA transporter were observed. 4. It is concluded that unlike the situation with midazolam, there is no resistance to the EEG effect of tiagabine in the kindling model of experimental epilepsy. The observed shift in the concentration - EEG effect relationship to lower concentrations can presumably be explained by the increase in the baseline GABA levels.

Pharmacokinetic-pharmacodynamic correlation of lamotrigine, flunarizine, loreclezole, CGP40116 and CGP39551 in the cortical stimulation model

The purpose of this study was to assess the concentration-anti-convulsant effect relationships of a number of anti-convulsant drugs in the direct cortical stimulation model, to obtain more insight in the properties and predictive value of this model. The time course of the effect of lamotrigine, loreclezole, flunarizine, CGP40116 and CGP39551 was determined after iv. administration in conjunction with their pharmacokinetics. Convulsive activity was induced by stimulation of the motor cortex with a ramp-shaped pulse train. This technique allows consecutive measurements of the treshold for localized (TLS) and for generalized (TGS) seizure activity. Increase in threshold was used as measure of the anti-convulsant effect. After administration of lamotrigine, pronounced elevation of the TGS, with little change in the TLS, was observed. Flunarizine caused a similar effect, but much less intense. Loreclezole strongly elevated the TGS and to a lesser extent the TLS, also. The concentration-anti-convulsant effect relationship of the three compounds could be fitted by an exponential model. The NMDA antagonists, CGP40116 and CGP39551, induced minor changes in the TLS and a slight increase in the TGS. The onset of this effect was marked by a delay relative to blood concentrations. The biophase equilibration kinetics was estimated and a linear model was applied to describe the concentration-effect relationship of both NMDA antagonists. The present results show that the cortical stimulation model is a suitable technique for integrated pharmacokinetic-pharmacodynamic modelling and for assessing anti-convulsant efficacy. The results show that the model is rather insensitive to calcium channel blockers and NMDA antagonists.

Stereoselective central nervous system effects of the R- and S-isomers of the GABA uptake blocker N-(4, 4-di-(3-methylthien-2-yl)but-3-enyl) nipecotic acid in the rat

1. The 'effect compartment' model was applied to characterize the pharmacodynamics of the R- and S-isomers of tiagabine in conscious rats in vivo using increase in the beta activity of the EEG as a pharmacodynamic endpoint. 2. No pharmacokinetic differences in plasma were observed between R- and S-tiagabine. The values for clearance and volume of distribution at steady-state were 103+/-10 versus 90+/-6 ml min(-1) kg(-1) and 1.8+/-0.2 versus 1.6+/-0.2 l kg(-1) for the R- and S-isomer, respectively. In contrast, plasma protein binding showed a statistically significant difference with values of the free fraction of 5.7+/-0.5 and 11.4+/-0.6%. In addition the rate constant for transport to the effect compartment was also different with values of 0.027 versus 0.067 min(-1). 3. For both isomers the relationship between concentration and EEG effect was non-linear and successfully characterized on basis of the Hill equation. A statistically significant difference in the value of EC(50) of 328+/-11 versus 604+/-18 ng ml(-1) was observed for R- and S-tiagabine respectively. The values of the other pharmacodynamic parameters were identical. 4. It is concluded that the differences in in vivo pharmacodynamics of R- and S-tiagabine can be explained by stereoselective differences in both the affinity to the GABA-uptake transporter and the degree of non-specific protein binding in plasma and at the effect site.

The antiarrhythmic and anticonvulsant effects of dietary N-3 fatty acids

It has been shown in animals and probably in humans, that n-3 polyunsaturated fatty acids (PUFAs) are antiarrhythmic. We report recent studies on the antiarrhythmic actions of PUFAs. The PUFAs stabilize the electrical activity of isolated cardiac myocytes by modulating sarcolemmal ion channels, so that a stronger electrical stimulus is required to elicit an action potential and the refractory period is markedly prolonged. Inhibition of voltage-dependent sodium currents, which initiate action potentials in excitable tissues, and of the L-type calcium currents, which initiate release of sarcoplasmic calcium stores that increase cytosolic free calcium concentrations and activate the contractile proteins in myocytes, appear at present to be the probable major antiarrhythmic mechanism of the PUFAs.

Experimental studies on antiarrhythmic and antiseizure effects of polyunsaturated fatty acids in excitable tissues

It has been shown that in animals, and probably in humans, n-3 polyunsaturated fatty acids (PUFAs) are antiarrhythmic. We discuss our recent studies on the antiarrhythmic actions of PUFAs. PUFAs stabilize the electrical activity of isolated cardiac myocytes by requiring a stronger electrical stimulus to elicit an action potential and by markedly prolonging the refractory period. These electrophysiologic effects are the result of specific modulation of ion currents, particularly of the voltage-dependent sodium current and of the L-type calcium currents across sarcolemmal phospholipid membranes. This appears to be the probable major antiarrhythmic mechanism of PUFAs. However, they also similarly affect neuronal ion channels with potentially important functional effects on the nervous system.

Application of a combined "effect compartment/indirect response model" to the central nervous system effects of tiagabine in the rat

Pharmacological inhibition of GABA uptake transporters provides a mechanism for increasing GABAergic transmission, which may be useful in the treatment of various neurological disorders. The purpose of our investigations was to develop an integrated pharmacokinetic-pharmacodynamic (PK/PD) model for the characterization of the pharmacological effect of tiagabine, R-N-(4,4-di-(3-methylthien-2-yl)but-3-enyl)nipecotic acid, in individual rats in vivo. The tiagabine-induced increase in the amplitude of the EEG 11.5-30 Hz frequency band (beta), was used as pharmacodynamic endpoint. Chronically instrumented male Wistar rats were randomly allocated to four groups which received an infusion of 3, 10, or 30 mg kg-1 of tiagabine or vehicle over 10 min. The EEG was continuously recorded in conjunction with frequent arterial blood sampling. The pharmacokinetics of tiagabine could be described by a biexponential equation. The pharmacokinetics of tiagabine were not dose dependent, and the pooled values for clearance, volume of distribution at steady state and terminal half-life were (mean +/- SE, n 23) 96 +/- 9 ml min-1 kg-1, 1.5 +/- 0.1 L kg-1 and 20 +/- 0.2 min. A time delay was observed between the occurrence of maximum plasma drug concentrations and maximal response. A physiological PK/PD model has been used to account for this time delay, in which a biophase was postulated to account for tiagabine available to the GABA uptake carriers in the synaptic cleft and the increase in EEG effect was considered an indirect response due to inhibition of GABA uptake carriers. The population values for the pharmacodynamic parameters characterizing the delay in pharmacological response relative to plasma concentrations were keo = 0.030 min-1 and kout = 81 min-1, respectively. Because of the large difference in these values the PK/PD model was simplified to the effect compartment model. Population estimates (mean +/- SE) were E0 = 155 +/- 6 microV, Emax = 100 +/- 5 microV, EC50 = 287 +/- 7 ng ml-1, Hill factor = 1.8 +/- 0.2 and keo = 0.030 +/- 0.002 min-1. The results of this analysis show that for tiagabine the combined "effect compartment-indirect response" model can be simplified to the classical "effect compartment" model.

Functional and electrophysiologic effects of polyunsaturated fatty acids on exictable tissues: heart and brain

It has been shown in animals and probably in humans, that n-3 polyunsaturated fatty acids (PUFAs) are antiarrhythmic. The free PUFAs stabilize the electrical activity of isolated cardiac myocytes by inhibiting sarcolemmal ion channels, so that a stronger electrical stimulus is required to elicit an action potential and the relative refractory period is markedly prolonged. This appears at present to be the probable major antiarrhythmic mechanism of the PUFAs. They similarly inhibit the Na+ and Ca2+ currents in rat hippocampal neurons which results in an increase in the electrical threshold for generalized seizures using the cortical stimulation model in rats.

Rate of change of blood concentrations is a major determinant of the pharmacodynamics of midazolam in rats

The objective of this investigation was to characterize quantitatively the influence of the rate of increase in blood concentrations on the pharmacodynamics of midazolam in rats. The pharmacodynamics of midazolam were quantified by an integrated pharmacokinetic-pharmacodynamic modelling approach. Using a computer controlled infusion technique, a linear increase in blood concentrations up to 80 ng ml(-1) was obtained over different time intervals of 16 h, resulting in rates of rise of the blood concentrations of respectively, 1.25, 1.00, 0.87, 0.46, 0.34 and 0.20 ng ml(-1) min(-1). In one group of rats the midazolam concentration was immediately brought to 80 ng ml(-1) and maintained at that level for 4 h. Immediately after the pretreatment an intravenous bolus dose was given to determine the time course of the EEG effect in conjunction with the decline of midazolam concentrations. The increase in beta activity (11.5-30 Hz) of the EEG was used as pharmacodynamic endpoint. For each individual animal the relationship between blood concentration and the EEG effect could be described by the sigmoidal Emax model. After placebo, the values of the pharmacodynamic parameter estimates were Emax = 82+/-5 microV, EC50,u = 6.4+/-0.8 ng ml(-1) and Hill factor = 1.4+/-0.1. A bell-shaped relationship between the rate of change of midazolam concentration and the value of EC50,u was observed with a maximum of 21+/-5.0 ng ml(-1) at a rate of change of 0.46 ng ml(-1) min(-1); lower values of EC50,u were observed at both higher and lower rates. The findings of this study show that the rate of change in plasma concentrations is an important determinant of the pharmacodynamics of midazolam in rats.

Effects of repeated seizure induction on seizure activity, post-ictal and interictal behavior

Individual variability and numerous interactions between pharmacokinetics, pharmacodynamics, and homeostatic factors complicate the study of the anticonvulsant effect in animal models of seizure activity. In theory, both individual variability and the contribution of these factors to the anticonvulsant effect can be determined by following the time course of the pharmacological response and the corresponding plasma concentrations in individual animals. Currently, there are several formal pharmacokinetic-pharmacodynamic models available for the analysis of such data, which yield accurate estimates of drug intrinsic activity and potency. However, most models of seizure activity are not suited for such an approach, either because they can be applied only once, or because the expression of seizures is not constant over time. In addition, the induction of seizures constitutes repeated jeopardy to the animals, which may profoundly change behavior and interfere in the anticonvulsant response as well as in different physiological processes. In this paper, we compare ictal, post-ictal, and interictal behavior in three different models of seizure activity in rats, namely, the electroconvulsive shock, amygdala kindling and the cortical stimulation model (CSM). The methods were compared in the same way as they are currently in use for the assessment of antiepileptic drug effect. Our results show that repeated seizure activity induced by cortical stimulation does not exacerbate ictal activity (eye closure, jerk, gasp, forelimb clonus, and hind-limb tonus) nor post-ictal behavior (chewing and freezing), while producing less serious changes in interictal behavior (walk, lean, upright rearing, exploratory, grooming, and rest) than kindling or electroconvulsive shock. We conclude that seizures induced by cortical stimulation are reproducible and qualitatively similar to kindling seizures. Our results also suggest that the electroconvulsive shock model is not suited for pharmacokinetic-pharmacodynamic studies and that the assessment of interictal behavior may contribute to the evaluation of overall antiepileptic drug effect in seizure disorders.

Modelling of the pharmacodynamic interaction between phenytoin and sodium valproate

Treatment of epilepsy with a combination of antiepileptic drugs remains the therapeutic choice when monotherapy fails. In this study, we apply pharmacokinetic-pharmacodynamic modelling to characterize the interaction between phenytoin (PHT) and sodium valproate (VPA). Male Wistar rats received a 40 mg kg(-1) intravenous dose of PHT over 5 min either alone or in combination with an infusion of VPA resulting in a steady-state concentration of 115.5+/-4.9 microg ml(-1). A control group received only the infusion of VPA. The increase in the threshold for generalized seizure activity (ATGS) was used as measure of the anticonvulsant effect. PHT pharmacokinetics was described by a pharmacokinetic model with Michaelis-Menten elimination. The concentration-time course and plasma protein binding of PHT were not altered by VPA. The pharmacokinetic parameters Vmax and Km were, respectively, 294+/-63 microg min(-1) and 7.8+/-2.4 microg ml(-1) in the absence of VPA and 562+/-40 microg min(-1) and 15.6+/-0.9 microg ml(-1) upon administration in combination with VPA. A delay of the onset of the effect relative to plasma concentrations of PHT was observed. The assessment of PHT concentrations at the effect site was based on the effect-compartment model, yielding mean ke0 values of 0.128 and 0.107 min(-1) in the presence and absence of VPA, respectively. A nonlinear relationship between effect-site concentration and the increase in the TGS was observed. The concentration that causes an increase of 50% in the baseline TGS (EC50%TGS) was used to compare drug potency. A shift of EC50%TGS from 13.27+3.55 to 4.32+/-0.52 microg ml(-1) was observed upon combination with VPA (P<0.01). It is concluded that there is a synergistic pharmacodynamic interaction between PHT and VPA in vivo.

Pharmacodynamic interaction between phenytoin and sodium valproate changes seizure thresholds and pattern

1. In this study we used cortical stimulation to assess the effects of phenytoin (PHT), sodium valproate (VPA), and their interaction on total motor seizure and on the constituent elements of the seizure. 2. PHT (40 mg kg(-1)) was administered as an intravenous bolus infusion to animals receiving either a continuous infusion of VPA or saline. VPA plasma concentration was maintained at levels that produced no detectable anticonvulsant effect. 3. Analysis of ictal components (eyes closure, jerk, gasp, forelimb, clonus, and hindlimb tonus) and their durations revealed both qualitative and quantitative differences in drug effects. 4. The anticonvulsant effect is represented by the increase in the duration of the stimulation required to reach a given seizure threshold. PHT significantly increased the duration of the stimulation and of the motor seizure. This increase was greatly enhanced by VPA. In addition, ictal component analysis revealed that the combination of PHT and VPA causes the reduction of a specific seizure component (JERK). 5. Neither the free fraction of PHT nor the biophase equilibration kinetics changes in the presence of VPA. It is concluded that the synergism may be due to a pharmacodynamic rather than a pharmacokinetic interaction.

Adaptive changes in the pharmacodynamics of midazolam in different experimental models of epilepsy: kindling, cortical stimulation and genetic absence epilepsy

1. The objective of this investigation was to determine quantitatively whether experimental epilepsy is associated with a change in the pharmacodynamics of benzodiazepines in vivo. For that purpose the pharmacodynamics of midazolam were quantified by an integrated pharmacokinetic-pharmacodynamic approach in three different models of experimental epilepsy: amygdala kindling, cortical stimulation and genetic absence epilepsy. 2. The time course of the EEG effect was determined in conjunction with the decline of drug concentrations after intravenous administration of 10 mg kg(-1) midazolam. The pharmacokinetics of midazolam were most adequately described by a bi-exponential equation. No influence of epilepsy on the pharmacokinetics of midazolam was observed. 3. The increase in beta activity (11.5-30 Hz) of the EEG as derived by Fast Fourier Transformation analysis was used as pharmacodynamic endpoint. For each individual rat the increase in beta activity was directly related to the concentration in blood on the basis of the sigmoidal Emax pharmacodynamic model. In all three models a significant reduction in the maximal effect was observed, in amygdala kindling 28%, in the cortical stimulation model 49% and in genetic absence epilepsy 37%. No differences in the other pharmacodynamic parameters, E0 EC50,u and Hill factor, were observed. 4. It is inferred that in three different models of epilepsy there is a similar change in GABAergic functioning which is associated with a significant reduction in the intrinsic activity of midazolam in vivo. These models provide therefore a useful basis for further studies on the mechanism of epilepsy-induced changes in pharmacodynamics of anti-epileptic drugs.

Pharmacokinetic-pharmacodynamic modeling of the anticonvulsant and electroencephalogram effects of phenytoin in rats

In this study a pharmacokinetic-pharmacodynamic model is proposed for drugs with nonlinear elimination kinetics. We applied such an integrated approach to characterize the pharmacokinetic-pharmacodynamic relationship of phenytoin. In parallel, the anticonvulsant effect and the electroencephalogram (EEG) effect were used to determine the pharmacodynamics. Male Wistar-derived rats received a single intravenous dose of 40 mg . kg-1 phenytoin. The increase in the threshold for generalized seizure activity (TGS) was used as the anticonvulsant effect and the increase in the total number of waves in the 11.5 to 30 Hz frequency band was taken as the EEG effect measure. Phenytoin pharmacokinetics was described by a saturation kinetics model with Michaelis-Menten elimination. Vmax and Km values were, respectively, 386 +/- 31 microg . min-1 and 15.4 +/- 2.2 microg . ml-1 for the anticonvulsant effect in the cortical stimulation model and 272 +/- 31 microg . min-1 and 5.9 +/- 0.7 microg . ml-1 for the EEG effect. In both groups, a delay to the onset of the effect was observed relative to plasma concentrations. The relationship between phenytoin plasma concentrations and effect site was estimated by an equilibration kinetics routine, yielding mean ke0 values of 0.108 and 0.077 min-1 for the anticonvulsant and EEG effects, respectively. The EEG changes in the total number of waves could be fitted by the sigmoid Emax model, but Emax values could not be estimated for the nonlinear relationship between concentration and the increase in TGS. An exponential equation (E = E0 + Bn . Cn) derived from the sigmoid Emax model was applied to describe the concentration-anticonvulsant effect relationship, under the assumption that Emax values cannot be reached within acceptable electric stimulation levels. This approach yielded a coefficient (B) of 2.0 +/- 0.4 microA . ml . microg-1 and an exponent (n) of 2.7 +/- 0.9. The derived EC50 value of 12.5 +/- 1. 3 microg . ml-1 for the EEG effect coincides with the "therapeutic range" in humans.

Anticonvulsant effect of polyunsaturated fatty acids in rats, using the cortical stimulation model

Recent studies have shown that long-chain polyunsaturated fatty acids can prevent cardiac arrhythmias, attributed to the reduction in excitability of cardiomyocytes, owing mainly to a shift in hyperpolarizing direction of the inactivation curves of both Na+ and Ca2+ currents and to a slowed recovery from inactivation. Qualitatively similar effects of polyunsaturated fatty acids on inactivation parameters have been observed in freshly isolated hippocampal neurons. Since the same effects are presumed to underlie the action of some established anticonvulsant drugs, polyunsaturated fatty acids might have an anticonvulsant action as well. We have investigated this for eicosapentaenoic acid, docosahexaenoic acid, linoleic acid and oleic acid, employing cortical stimulation in rats, a seizure model allowing the determination of the full anticonvulsant effect-time profile in freely moving, individual animals. I.v. infusion of 40 micromol of eicosapentaenoic acid or docosahexaenoic acid over a period of 30 min, modestly increased the threshold for localized seizure activity after 6 h by 73 +/- 13 microA (mean +/- S.E.M.; n = 7) and 77 +/- 17 microA (n = 7), respectively, and the threshold for generalized seizure activity by 125 +/- 20 and 130 +/- 19 microA, respectively (P < 0.001). The thresholds remained elevated for 6 h after infusion, but returned to baseline the next day. Free plasma concentrations in rats treated with eicosapentaenoic acid or docosahexaenoic acid, averaged 5.7 +/- 1.6 microM (n = 4) for eicosapentaenoic acid and 12.9 +/- 1.8 microM (n = 5) for docosahexaenoic acid at the end of infusion, but declined to undetectable levels within 3 h. Linoleic acid and oleic acid were less effective. Possible mechanisms for the modest anticonvulsant effect but of long duration with the polyunsaturated fatty acids are discussed.

Seizure patterns in kindling and cortical stimulation models of experimental epilepsy

A large number of animal models has been proposed for the evaluation of the anticonvulsant effect of antiepileptic drugs. Various seizure patterns are produced and differences are frequently observed in anticonvulsant effect estimates obtained for the same drug in different models. The incidence of seizures and the threshold for the induction are usually the only measures used for the determination of the anticonvulsant effect. However, behavioural components expressed during seizures induced by different means are likely to differ considerably. The aim of this study was to provide a detailed behavioural description of ictal and post-ictal components in two models of electrically induced seizure activity: kindling and cortical stimulation model (CSM). Seizure activity was induced in two groups of 6 Wistar-derived rats. Ictal and post-ictal behaviours were recorded on video tape and quantified using a computer supported frame-by-frame encoding of the behavioural components. We encoded the duration and rate of occurrence of the following behavioural items: whisker movements, eye closure, myoclonic jerk, facial gasping, forelimb clonus, forelimb tonus, hindlimb tonus, immobility and chewing. It appears that both models are, in many respects, qualitatively similar. However, the models differ quantitatively. Behavioural expression of seizure activity differs in the following respects: (1) the total duration of the seizure induced by cortical stimulation is shorter than by kindling; (2) seizure activity in the CSM occurs mainly during stimulation, while in amygdala kindling, it occurs thereafter; and (3) seizures evoked in the CSM comprise relatively less violent behavioural items than in the amygdala kindling. The evaluation of the ictal and post-ictal behavioural components suggests that behavioural analysis could assist in the detection of differences in the mechanisms of action of antiepileptic drugs. In addition, observational measures can also be used to assess animal distress inflicted by different experimental procedures.

Pharmacokinetic/pharmacodynamic relationship of benzodiazepines in the direct cortical stimulation model of anticonvulsant effect

The in vivo concentration-anticonvulsant effect relationships of six benzodiazepines, midazolam, clonazepam, oxazepam, flunitrazepam, diazepam and clobazam were quantified in individual rats and correlated with the affinity to the GABAA-benzodiazepine receptor complex. Furthermore the interaction between midazolam and the benzodiazepine antagonist flumazenil was characterized. All benzodiazepines exhibited a nonlinear relationship between concentration and anticonvulsant effect without ceiling of the effect at higher concentration. The potency of most benzodiazepines was similar with values of the EC250, between 0.067 +/- 0.01 mg. l-1 for midazolam and 0.21 +/- 0.03 mg. l-1 for diazepam. The EC250,u of clobazam was 2.8 +/- 0.9 mg. l-1. These values were considerably larger than the Ki for binding at the GABAA-benzodiazepine receptor complex. No correlation was observed between EC250,u and Ki. Antagonism of the anticonvulsant effect of midazolam by flumazenil was associated with a remarkable change in the concentration-anticonvulsant effect relationship. Analysis of these data on basis of a composite model provided evidence for two separate effects of which only one is antagonized by flumazenil. The anticonvulsant effect at low midazolam concentration was characterized on basis of the sigmoid E maximal effect pharmacodynamic model. The value of the EC50,u was 0.0086 +/- 0.0013 mg. l-1 which is similar to the Ki for binding at the GABAA-benzodiazepine receptor complex. The second more pronounced anticonvulsant effect occurred at higher concentration and was described by an exponential function. The findings of this study indicate that the effect of benzodiazepines against seizures induced by cortical stimulation in vivo cannot be fully accounted for by an interaction at the GABAA-benzodiazepine receptor complex.

Polyunsaturated fatty acids modulate sodium and calcium currents in CA1 neurons

Recent evidence indicates that long-chain polyunsaturated fatty acids (PUFAs) can prevent cardiac arrhythmias by a reduction of cardiomyocyte excitability. This was shown to be due to a modulation of the voltage-dependent inactivation of both sodium (INa) and calcium (ICa) currents. To establish whether PUFAs also regulate neuronal excitability, the effects of PUFAs on INa and ICa were assessed in CA1 neurons freshly isolated from the rat hippocampus. Extracellular application of PUFAs produced a concentration-dependent shift of the voltage dependence of inactivation of both INa and ICa to more hyperpolarized potentials. Consequently, they accelerated the inactivation and retarded the recovery from inactivation. The EC50 for the shift of the INa steady-state inactivation curve was 2.1 +/- 0.4 microM for docosahexaenoic acid (DHA) and 4 +/- 0.4 microM for eicosapentaenoic acid (EPA). The EC50 for the shift on the ICa inactivation curve was 2.1 +/- 0.4 for DHA and > 15 microM for EPA. Additionally, DHA and EPA suppressed both INa and ICa amplitude at concentrations > 10 microM. PUFAs did not affect the voltage dependence of activation. The monounsaturated oleic acid and the saturated palmitic acid were virtually ineffective. The combined effects of the PUFAs on INa and ICa may reduce neuronal excitability and may exert anticonvulsive effects in vivo.

Can human polyclonal immunoglobulin raise the threshold for convulsions in rats?

In the present preliminary study we explored the possibility of an anticonvulsant effect of normal human immunoglobulin in an animal epilepsy model based on direct cortical stimulation in freely moving rats. After human immunoglobulin administration a significant and prolonged elevation of the threshold for convulsions was measured in 12% (6/49) of the total group of outbred Wistar rats. In the subgroup of more than seven months old Wistar rats this was 67% (6/9). When a threshold increasing effect of immunoglobulin occurred, it was detectable within 0.5-1 hour after administration, reached its maximum after approximately two hours and continued for at least 40 hours.

Nerve fiber size-related block of action currents by phenytoin in mammalian nerve

We investigated nerve fiber size-related actions of phenytoin (PHT) by applying the anticonvulsant on 2-mm-long stretches of desheathed whole nerves, excised from rat sural nerve. Compound action potentials (APs) were elicited by voltage pulses of increasing amplitude and recorded as monophasic action currents of the A alpha beta-type along the surface of the nerve. The area under the action current Q at supramaximal stimulation was reduced by 11 and 30% in solutions containing 10 and 100 microM PHT, respectively, similar to the reduction in peak action current. However, a greater reduction in Q induced by PHT was observed with smaller stimuli at both concentrations. This stimulus-dependent reduction was believed to originate from selective inhibition of the thicker nerve fibers. Using a mathematical model, we separated Q into contributions Q alpha of the alpha-fibers and Q beta of the beta-fibers. In solutions containing 10 microM PHT, Q alpha was reduced by 15% maximally, whereas Q beta was not affected. Both fiber types were reduced < or = 30% in the presence of 100 microM PHT, whereas the relations between Q alpha and Q beta, respectively, and stimulus voltage shifted along the voltage axis for 0.3 V, suggesting that the larger fibers in the A alpha beta-groups were more inhibited by PHT than the smaller ones. Abolition of the early phases of the compound action currents by PHT also indicated loss mainly of faster conducting nerve fibers. We conclude that primarily the larger fibers in the A alpha beta populations were inhibited by the anticonvulsant, strongly suggesting a differential mode of action by PHT on myelinated nerve fibers.

Characterization of the pharmacodynamics of several antiepileptic drugs in a direct cortical stimulation model of anticonvulsant effect in the rat

In this investigation a newly developed direct cortical stimulation technique was evaluated for measurement of anticonvulsant efficacy in rats. The kinetics of drug action for carbamazepine, phenytoin, valproate, phenobarbital, ethosuximide and oxazepam were studied in conjunction with their pharmacokinetics. Motor cortex stimulation with a ramp-shaped pulse train allowed successive determination of a threshold for localized seizure activity (TLS) and for generalized seizure activity (TGS). For each drug the time course of effect was followed in individual animals. Differential effects on the pharmacodynamic parameters were seen. Phenytoin and carbamazepine clearly elevated the TGS. However, phenytoin did not affect TLS and carbamazepine only marginally. Valproate increased both TLS and TGS to the same extent. Phenobarbital and oxazepam elevated both thresholds, but the effect on TGS was more pronounced. Ethosuximide had little effect on both thresholds. Comparison with other animal models suggested that elevation of TLS reflects an effect on seizure initiation, whereas elevation of TGS above TLS reflects an effect on seizure propagation. All drugs exhibited a nonlinear relationship between plasma concentration and anticonvulsant efficacy, without ceiling of anticonvulsant intensity at the highest concentrations. The effective concentration range of most compounds coincided with the "therapeutic" range in humans. The direct cortical stimulation technique is useful for preclinical monitoring of anticonvulsant efficacy with most antiepileptic drugs because it allows detection of both qualitative and quantitative differences. In addition the model is particularly useful for time course studies.

Hypoxic-ischemic encephalopathy sustained in early postnatal life may result in permanent epileptic activity and an altered cortical convulsive threshold in rat

The aim of this study was to investigate whether the rat cerebral cortex, damaged by hypoxia-ischemia in early postnatal life, would show an increased seizure susceptibility and/or spontaneous epileptic discharges in adulthood. To that end 12-13-day-old Wistar rat pups were unilaterally exposed to hypoxic-ischemic conditions. After a recovery period of about 2.5 months, recording and stimulation electrodes were permanently implanted over the left and right fronto-parietal neocortex. Long-term recording of baseline electrocortical activity showed that only those animals that had incurred severe brain damage, as was reflected by the presence of a cortical infarction, ran a high risk of developing permanent epileptic activity. With the aid of the stimulation electrodes the initial threshold for localized seizure activity was found to be the same for the experimental and non-treated groups. However, when the kindling-like decline of this threshold was assessed by repeated testing over a 2-week period, the infarcted animals tended to a more rapid decline but a higher stabilization level than the non-infarcted and control animals.

Valproate and sodium currents in cultured hippocampal neurons

Cultured rat hippocampal neurons with short processes were investigated using the whole cell voltage clamp under conditions appropriate for isolating Na+ currents. After incubation of the neuron culture for a period of 15-30 min in 1 mM sodium valproate, several parameters of the Na+ current were changed. The peak Na+ conductance gp, measured using hyperpolarizing prepulses, was reduced by valproate in a voltage-dependent manner. In the membrane voltage range from -30 to +20 mV, this reduction showed a linear dependence on voltage, increasing from about zero to approximately 30% of gp, the maximum peak Na+ conductance of the neuron. At the holding voltage of -70 mV, the inactivation parameter h infinity decreased from 0.88 in the control to 0.64 in the valproate solution. This reduction originated mainly from a 10 mV shift in the sigmoid relation between h infinity and membrane voltage along the voltage axis to hyperpolarizing potentials. The decay of the maximum peak Na+ current (inactivation) could be fitted by a biexponential function. Time constants of the fast and slow component at -20 mV decreased in valproate by about 50%. Valproate also retarded the recovery from inactivation, as determined at the holding voltage. The sigmoid recovery from inactivation could reasonably be described by an exponential function with time constant tau r and delay time delta t. Both tau r and delta t increased more than 200% in valproate. Our results indicate that valproate affected the Na+ current in hippocampal neurons in a way that contributed to a considerable depression of Na+ reactivation.(ABSTRACT TRUNCATED AT 250 WORDS)

Pharmacodynamics of the anticonvulsant effect of oxazepam in aging BN/BiRij rats

1. The purpose of this investigation was to examine the influence of increasing age on the pharmacokinetics and the time course of the anticonvulsant response of oxazepam in BN/BiRij rats as an animal model of aging. 2. Oxazepam was administered intravenously in a dose of 12 mg kg-1 body weight and the anticonvulsant effect intensity was measured as elevation above baseline of a threshold for induction of localized seizure activity (TLS). Direct cortical stimulation with ramp shaped electrical pulse trains of increasing intensity was used to determine this threshold. 3. The pharmacological effect vs. time profile showed in young rats an anticonvulsant component followed by proconvulsant component which is suggestive for the occurrence of acute tolerance and/or withdrawal syndrome. With increasing age the proconvulsant component disappeared, resulting in a monophasic effect profile (anticonvulsant effect only) at the age of 35 months with significantly higher anticonvulsant effect intensity immediately following drug administration. No age-related changes in the pharmacokinetic parameters of oxazepam were observed. 4. In five animals of each age group, benzodiazepine receptor binding characteristics were determined in vitro with [3H]-flunitrazepam as a ligand. Both receptor density and affinity did not show age-related changes. Available literature data on post-receptor events do not indicate conclusive age-related changes. 5. It is concluded, that the observed change in the pharmacodynamics of anticonvulsant effect of oxazepam can be explained by the disappearance of the tolerance/withdrawal phenomenon. This is compatible with a decreased efficiency of homeostatic control mechanisms in the elderly.

Properties of the convulsive threshold determined by direct cortical stimulation in rats

The threshold for convulsions in rats can be determined by applying ramp-shaped pulse trains directly to the cerebral cortex in rats, which provides a convenient model for investigating anticonvulsant drug effects. This study was undertaken to extend a previous study on the properties of this model. Analysis of the cortical EEG, recorded from two motor areas and one somatosensory area, showed that the start of clonic forepaw movement, marking the convulsive threshold, is preceded by the appearance of sharp negative spikes at the electrodes in the two motor areas. There was a strong linear relation between the clinically determined threshold and the EEG derived threshold (r = 0.93, slope 0.99, SD 0.04), confirming the validity of the clonic movement threshold as an objective and accurate measure. Examination of the seizure patterns seen with various degrees of suprathreshold stimulation led to the distinction between a threshold for localized and for generalized seizure activity (TLS and TGS respectively). Carbamazepine selectively and strongly increased the TGS, whereas it only slightly affected the TLS, indicating that cortical stimulation can be used to select drugs that specifically prevent seizure spread, for which carbamazepine is a prototype. It was found that the TLS was not affected by testing at intervals as short as 1 min, provided that no self-sustained seizures were induced. However, if the TGS was passed, the TLS was increased substantially for at least 10 min, while complete recovery could take several hours. The intensity of stimulation, rather than seizure duration, appeared to be the determinant for the TLS increase. There was no seasonal influence or effect of stimulation electrode depth. There may be a minor effect of experience in using the test. It was concluded that the observed variability was mainly an intrinsic property of the individual animal.

Pharmacokinetic-pharmacodynamic modelling of the anticonvulsant effect of oxazepam in individual rats

1. The purpose of this investigation was to examine in vivo drug-concentration anticonvulsant effect relationships of oxazepam in individual rats following administration of a single dose. 2. Whole blood concentration vs time profiles of oxazepam were determined following administration of doses of 4, 8 and 12 mg kg-1. The pharmacokinetics could be described by an open 2-compartment pharmacokinetic model. Following 12 mg kg-1 the values (mean +/- s.e., n = 11) of clearance and volume of distribution were 28 +/- 2 ml min-1 kg-1 and 2.6 +/- 0.31 kg-1, respectively, and were not significantly different from the values obtained at the other doses. 3. The anticonvulsant effect was quantitated by a new technique which allows repetitive determination of the convulsive threshold by direct cortical stimulation within one rat. Significant dose-dependent elevations of the seizure threshold were observed. 4. By pharmacokinetic-pharmacodynamic modelling, a log-linear relationship was found between concentration and anticonvulsant effect. Following 12 mg kg-1 the values (mean +/- s.e., n = 11) of the pharmacodynamic parameters slope and minimal effective concentration (Cmin) were 243 +/- 27 microA and 0.11 +/- 0.02 mg l-1, respectively and not significantly different from the values obtained at the other doses. 5. In a repeatability study the pharmacodynamic parameters were determined twice on two different occasions with an interval of two weeks in the same group of 11 rats. The inter-animal variability in the pharmacodynamic parameter slope was 46%, whereas the intra-animal variability was 24 +/- 18%. The value of the minimal effective concentration was in each animal and on each occasion close to zero within the relatively narrow range of 0.01-0.30mgI. 6. The results of this study showed that it is possible to determine in vivo concentration-anticonvulsant effect relationships of oxazepam under non-steady-state conditions in individual rats. The anti-convulsant effect of oxazepam appeared to be a rapidly reversible direct effect and acute tolerance did not develop within the time frame of the experiments.

Determination of the threshold for convulsions by direct cortical stimulation

In this study we investigated whether determination of the convulsion threshold by electrical stimulation of the cortex could be used as a simple test for measuring anticonvulsant drug activity in unrestrained, unanaesthetized rats. Pulse trains delivered to electrodes implanted in the frontoparietal cortex elicited convulsions, similar to those seen in the classical electroshock tests. The threshold could be determined rapidly with pulse trains which increased in strength in a ramp-shaped fashion (bipolar pulses of 2 msec, 50 pulses/sec, increment 1.3 microA/pulse). The threshold was defined as the current needed to elicit forelimb clonus. Upon repeated stimulation the threshold declined from a value of about 600 microA to about 350 microA in 20 sessions. Thereafter, continued testing did not result in considerable changes in threshold. After stabilization, the convulsion threshold could be determined repeatedly with intervals as short as 5 min. Following i.p. injection of 5 mg/kg of diazepam an elevation of the threshold of 30% was observed 0.5 h after injection. After 5 daily injections, evidence for the development of complete tolerance was obtained. After i.v. injection of 8 mg/kg oxazepam, the threshold increase reached a peak level of 75% after 20 min. The changes in threshold followed arterial blood concentration of oxazepam, which was maximally 4.8 micrograms/ml immediately after injection. The threshold returned to baseline in approximately 6 h. The results of the present study show that with our procedure anticonvulsant drug activity can be accurately, rapidly and repeatedly determined in individual animals, both in acute and chronic experiments.

Enhancement of recurrent inhibition by angular bundle kindling is retained in hippocampal slices

Rats were kindled in the right angular bundle. EEG and monosynaptically evoked responses were monitored in the ipsilateral fascia dentata. Although every animal was kindled, Long Term Potentiation (LTP) of monosynaptic responses was observed only in part of the kindling sessions, suggesting that LTP is not required for kindling. Paired pulse inhibition of granule cell discharge was progressively enhanced by kindling. Transverse hippocampal slices (400-700 micron) of fully kindled rats were prepared 1 hour after the last seizure. Field potentials evoked by stimulation appeared completely normal. Spontaneous epileptiform discharges were not observed in control solution, in low [Ca++]0, in high [K+]0 or after high frequency stimulation. Paired pulse inhibition was enhanced in fascia dentata, but not in area CA1. Enhancement of inhibition may be caused by increased activity of inhibitory synapses or by a nonsynaptic hyperpolarizing current. The relation between the absence of spontaneous activity and enhanced inhibition in the fascia dentata is unclear.

Valproate reduces excitability by blockage of sodium and potassium conductance

Effects of the antiepileptic drug valproate on sodium and potassium currents in the nodal membrane of peripheral nerve fibers of Xenopus laevis were determined by voltage- and current-clamp experiments. Under voltage-clamp conditions, a reduction of both sodium and potassium conductance (in a ratio of 2:1) was observed. Typically, 2.4 mM (400 mg/L) valproate reduced the sodium current 54% and the potassium current 26%, at a membrane potential of 5 mV. Valproate did not affect the leakage conductance. The reduction of potassium conductance was voltage dependent, being more pronounced at more positive membrane potentials. For the sodium system, a voltage dependency of the blockage could not be established. Under current-clamp conditions, valproate caused a reduction of excitability of nerve membrane: amplitude of the action potential and maximum rate of rise were decreased, whereas threshold potential was increased. The ability to follow high-frequency stimulation was impaired.

Spontaneous epileptiform discharges in hippocampal slices induced by 4-aminopyridine

4-Aminopyridine (4-AP) induced 2 types of spontaneous field potentials (SFPs) in the hippocampal slice. Type I resembled spontaneous activity induced by other convulsants. They occurred at a rate of approximately 1 Hz, started in the CA2/CA3 region and spread at a velocity of 0.3 m/s to area CA1. Transsection experiments and laminar profiles indicated that they spread synaptically along the Schaffer collateral pathway. Synaptic blockade by low Ca2+/high Mg2+ or kynurenic acid reversibly abolished type I SFPs. Increasing [Ca2+]o lowered the rate and slightly increased the amplitude. Possibly, increased spontaneous transmitter release, and not disinhibition, is responsible for the generation of type I SFPs. Type II occurred at a rate of about 0.15 Hz and travelled in the same direction, but a factor 10 slower. They could not be blocked by separation of the CA1 and CA3 region; coupling remained until stratum moleculare was severed. Type II could not be suppressed by blockade of synaptic transmission. The laminar profile is similar in shape to that of type I but not identical. Increasing [Ca2+]o had the same but stronger effect as on type I. Type II SFPs depressed evoked population spikes up to a second and delayed the next type I SFP. The mechanisms involved remain largely speculative; further analysis is needed to help understand the epileptogenic action of 4-AP.

Excitability increase of neurons in olfactory cortex slices of the guinea pig after penicillin administration

The mechanism underlying the effects of penicillin on slices of the olfactory cortex of the guinea pig was examined. In a previous report it was shown that penicillin increases the amplitude of the presynaptic action potential, the population EPSP and, more strongly, the population responses of the postsynaptic cells. Moreover, the postsynaptic population responses increased in number and suggested strong repetitive firing. These results were confirmed in the present study. Analysis of stimulus-response relationships suggested that the enhancement of the postsynaptic response was due to an increase in excitability of the postsynaptic neurons by penicillin. The amplitude changes of the presynaptic action potential and the EPSP were probably largely, if not completely, due to an increase in resistance of the bathing fluid. It was found that the changes in population responses paralleled to a large extent changes in cell discharge. In addition, penicillin was found to induce spontaneous firing of the postsynaptic cells. The changes in cell discharge were consistent with an increase in excitability of the postsynaptic cells.

Actions and interactions of dipropylacetate and penicillin on evoked potentials of excised prepiriform cortex of guinea pig

Slices from guinea pig brain containing the lateral olfactory tract (LOT) and the prepiriform cortex were studied in vitro. Field potentials, evoked by stimulation of the LOT, were recorded extracellularly. This field potential comprises a compound action potential, a surface negative wave (identified as EPSP), and superimposed positive peaks ("population spikes" or PSs) reflecting postsynaptic activity. In a previous article the penicillin-induced increase of EPSP and of both amplitude and number of PSs was described. Now we are reporting the slight depression of EPSP and PSs and the prevention of the appearance of penicillin-induced PSs by an antiepileptic drug sodium dipropylacetate (Depakine). The effect was dose-dependent. Models explaining the effects of penicillin and dipropylacetate are discussed.