Postnatal caffeine treatment affects differently two pentylenetetrazol seizure models in rats

Seizure 18 (2009) 463–469

Postnatal caffeine treatment affects differently two pentylenetetrazol seizuremodels in rats

Jana Tchekalarova a,b,1,*, Hana Kubova b,1, Pavel Mares b,1

a Laboratory of Experimental Psychopharmacology, Acad. G. Bonchev Street, Block 23, Bulgarian Academy of Sciences, Sofia 1113, Bulgariab Institute of Physiology, Academy of Sciences of the Czech Republic, Prague, Czech Republic

A R T I C L E I N F O

Article history:

Received 10 November 2008

Received in revised form 13 February 2009

Accepted 3 April 2009

Keywords:

Caffeine

Pentylenetetrazol

Spike-and-wave rhythm

Development

Rats

A B S T R A C T

Effects of repeated postnatal administration of caffeine (10 and 20 mg/kg s.c. daily from P7 to P11) were

studied in two models of epileptic seizures characterized by spike-and-wave EEG rhythm in 18- and 25-

day-old rats. Rhythmic metrazol activity (RMA, model of human absences) was induced by low dose of

pentylenetetrazol (PTZ, 20 mg/kg or 40 mg/kg, i.p.) and minimal clonic seizures (model of human

myoclonic seizures) by two successive doses of PTZ (20 and 40 mg/kg i.p.). Early postnatal caffeine

treatment resulted in significant changes of RMA only in 18-day-old rats. Anticonvulsant effects were

observed in RMA episodes elicited by the 20-mg/kg dose of PTZ in both caffeine groups whereas latency

of RMA episodes induced by the 40-mg/kg dose was shortened and their duration was prolonged. No

changes were found in 25-day-old animals. Incidence, EEG and motor pattern of minimal clonic seizures

were not changed. Some animals in both control age groups exhibited transition to generalized tonic–

clonic seizures. This type of seizures never appeared in caffeine-treated 25-day-old animals. Mixed

effects of postnatal caffeine exposure were demonstrated; these predominantly anticonvulsant effects

are age- and model-specific.

� 2009 Published by Elsevier Ltd on behalf of British Epilepsy Association.

Contents lists available at ScienceDirect

Seizure

journal homepage: www.e lsev ier .com/ locate /yse iz

1. Introduction

Adenosine is commonly accepted as a neuromodulator in centralnervous system (CNS) with an inhibitory action on synaptic activityand neurotransmitter release.1 Low extracellular levels of adenosinein brain activate adenosine A1 and A2a receptors under physiologicalconditions.2 Both receptors are blocked by methylxanthines.2 Themethylxanthine caffeine is the most common CNS stimulant ineveryday life. The proconvulsant action of adenosine antagonistssuch as caffeine, aminophylline and other methylxanthines wasdescribed both in adult3 and in young rats.4,5 Moreover, it has beenshown that immature rats are more susceptible to methylxanthinesthan the adult animals.6 Nowadays, methylxanthines are among themost often prescribed drugs in neonatal medicine7, and in particular,in the treatment of apneic episodes in premature infants.8

Methylxanthines are given to preterm infants both acutely (loadingdose) or chronically.7 The long-term drug-induced blockade ofadenosine receptors at early stages of maturation might modifybrain development. The outcome of such blockade cannot bepredicted from clinical data therefore experimental studies are

* Corresponding author. Tel.: +359 2979 2172; fax: +359 2 719 109.

E-mail addresses: [email protected] (J. Tchekalarova),

[email protected] (P. Mares).1 Fax: +420 2 41062488.

1059-1311/$ – see front matter � 2009 Published by Elsevier Ltd on behalf of British

doi:10.1016/j.seizure.2009.04.002

necessary. In experimental animals, chronic exposure to caffeineduring peri- and postnatal period has been shown to result inadaptive long-lasting neurochemical and behavioral responses thatare usually opposite to acute drug effects under normal as well aspathological conditions.9–13 Only few reports considered theconsequences of repeated treatment with this methylxanthine onseizure susceptibility.11,14 Similar to the data from adult rodents,repeated caffeine treatment during postnatal period results in adecreased seizure susceptibility in different models of convulsiveseizures. Moreover, caffeine effects may delay the decrease in theseizure threshold that occurs for many agents since the late juvenileage.11,14 The observed attenuation in the seizure susceptibility afterrepeated administration of caffeine at low doses of 15–20 mg/kg for5 days both during adulthood and developmental period correlateswith an increase in adenosine A1 receptor density in specific brainstructures.10,15

Present experiments represent continuation of our recentfindings that repeated caffeine administration at postnatal days7–11 caused both transient and durable age-specific changes inbehaviour and seizure susceptibility.16–18 Furthermore, repeatedpostnatal caffeine administration attenuated sensitivity to con-vulsant drugs with different mechanisms of action (aminophylline,pentylenetetrazol (PTZ), picrotoxin, bicuculline) in an age-depen-dent manner.18 These data suggest that repeated postnatal caffeineadministration affects not only adenosinergic system but at leastalso (probably indirectly) GABAergic inhibitory system.

Epilepsy Association.

J. Tchekalarova et al. / Seizure 18 (2009) 463–469464

The aim of the present study was to further specify effects ofrepeated postnatal caffeine treatment in immature ratson seizures inlater life. We hypothesized that effects will depend on the type ofelicited seizures. Two models of human seizures with differentpathophysiological mechanisms were used – a model of generalizedabsence seizures with an important role of inhibitory mechanismsand a model of myoclonic seizures with marked convulsions (forreview19). Such two models can be elicited by different doses ofpentylenetetrazol (PTZ); both are characterized by EEG spike-and-wave (SW) rhythm but markedly differ in their behavioral correlates.An acute model of absence seizures can be induced by low doses ofPTZ. This model is accepted as a valid model of this type of primarygeneralizedhumanabsenceseizures.20It ischaracterized only byEEGspike-and-wave rhythm (rhythmic metrazol activity, RMA) but alsoby the same motor pattern (behavioral arrest and minimal motorphenomena21) and pharmacological sensitivity22 as other experi-mental models of absences such as feline generalized penicillinepilepsy,23 gamma-hydroxybutyrate-induced model24 and geneti-cally determined absence seizure GAERS and WAG/Rij rats.25,26

Higher doses of PTZ elicit in rodents clonic seizures usuallyrestricted to head and forelimb muscles without a loss of rightingreflexes (minimal metrazol seizures of the older literature); toniccomponent if present consists mainly of torsion of the trunk.27

These seizures are also characterized by a spike-wave rhythm inthe EEG what is in agreement with the clinical data that thisrhythm accompanies not only absences but also other types ofseizures in epileptic patients.28 Motor pattern and pharmacolo-gical sensitivity suggest that these seizures could be considered asan experimental model of myoclonic seizures.29,30 Early postnatalcaffeine exposure caused transient dose-dependent pro- or antic-onvulsant action in another model of myoclonic seizures elicitedby electrical stimulation of sensorimotor cortex – epilepticafterdischarges (ADs).17,31 Comparison with this model maydemonstrate a role of an epileptogenic agent. The 40-mg/kg doseof PTZ never inducing minimal clonic seizures in naıve rats failed toshow possible proconvulsant effect of repeated caffeine exposuretherefore two successive doses of PTZ (20 and 40 mg/kg i.p.) wereadministered. Only two age groups (18- and 25-day-old rats) werestudied because SW episodes and minimal clonic seizures cannotbe induced reliably with PTZ before postnatal day 18.27,32

2. Methods

2.1. Animals

The experiments were carried out on immature Wistar rats(breeding of the Institute of Physiology, Academy of Sciences,Prague, Czech Republic). Litters consisted from 10 pups; postnatalday 0 was a day of birth. The rats were housed together with theirmothers in a temperature-controlled environment (22 � 1 8C andhumidity 50–60%) with a 12/12 h light/dark cycle (lights on at 6 a.m.).Food and water were provided ad libitum (with the exception of thetest period). Rat pups were taken from their mothers just beforetesting. All experiments were approved by the Animal Care and UseCommittee of the Institute of Physiology to be in agreement withAnimal Protection Law of the Czech Republic (fully compatible withEuropean Community Council directives 86/609/EEC).

2.2. Caffeine treatment

Rat pups in each litter were randomly assigned to one controland two experimental groups. Each group consisted of at least 10animals and contained pups from four or five litters. Caffeineadministration started at P7; animals assigned to the experimentalgroups were injected subcutaneously with either 10 or 20-mg/kgcaffeine (Sigma, St. Louis, MO, #C 0750) in a volume of 1-ml/kg

body weight. Control rats received saline (1 ml/kg). Injections wererepeated daily for 5 days.

2.3. Surgery

Four cortical recording electrodes were implanted epidurallyover sensorimotor and occipital areas symmetrically over bothhemispheres to 18- or 25-day-old rats. Surgical preparation wasperformed under ether anesthesia. The coordinates were recalcu-lated from the adult brain (AP = 0; L = 2 mm and AP = 6, L = 4 mm,respectively) based on bregma-lambda distance. A referenceelectrode was placed into the nasal bone and the ground electrodeinto the occipital bone. All electrodes were fixed to the skull byfast-curing dental acrylic. The surgery lasted less than 10 min. Ratswere then placed into plexiglas cages over an electrical heating pad(34 8C, i.e. the temperature in the nest). The animals were allowedto recover for at least 1 h. After the recovery period, righting andplacing reflexes were examined and the animals were offered 5%sucrose solution not only as a nutrient but also to check thesuckling reflex. Only then the experiment started.

2.4. Recording

EEG was registered in a monitoring system (KaminskijBiomedical Research Systems, Prague); the rate of digitalizationwas 200 Hz. All rats were allowed to habituate to the environmentfor 15 min. Control EEG was registered for 10 min, then PTZ wasinjected and registration continued for at least 30 min, allbehavioral phenomena were coded directly into EEG recording.

2.5. Experiment 1: rhythmic metrazol activity (RMA) induced

with the 20-mg/kg dose of PTZ

Pentetrazol (PTZ; free base) (Sigma, St. Louis, Mo., USA) wasfreshly dissolved in 0.9 % NaCl solution and administered in a doseof 20 mg/kg (in a volume of 1 ml/kg) intraperitoeneally. Latency tothe appearance of the first RMA episode and of the first generalizedRMA (GRMA) (i.e. recorded in all four cortical areas) was measured.All RMA episodes (their number and duration) between 10–15 minand 20–25 min after PTZ injection were counted and both total andmean duration of episodes were calculated.

2.6. Experiment 2: rhythmic metrazol activity induced with

the 40-mg/kg dose of PTZ

Experimental design was the same as in Experiment 1, only thedose of PTZ was 40 mg/kg i.p. Latency to the appearance of the firstand the first generalized RMA episode, number and duration of theseepisodes were counted in the same two 5-min periods as in the firstexperimentand total andmeandurationofepisodeswerecalculated.

2.7. Experiment 3: seizures induced with successive doses of

PTZ (20 + 40 mg/kg)

Experimental design was again the same as in the first andsecond experiments, the first 20-mg/kg dose of PTZ was followedafter 20 min by the second injection (40 mg/kg). Continuous EEGregistration was stopped 30 min after the second dose of PTZ.Presence of RMA episodes after the first dose served as an indicatorof PTZ action. Attention was focused on incidence of seizures, theirmotor pattern and duration.

2.8. Statistical analysis

Means and S.E.M. were calculated for all RMA episodes data.The data were subjected to analysis by two-way ANOVA, with

Fig. 1. EEG recording of two spike-wave episodes from a 25-day-old rat 11 minutes

after the PTZ administration (40 mg/kg i.p.). Individual leads from top to bottom: LF

– left frontal, LO – left occipital, RF – right frontal, RO – right occipital always in

reference connection. Amplitude callibration 0.5 mV, time mark 5 s.

J. Tchekalarova et al. / Seizure 18 (2009) 463–469 465

factors age (P18 and P25) and drug (saline, caffeine 10 mg/kg,caffeine 20 mg/kg). For all analyses, significant main effects offactors were followed by post hoc analysis using Tukey’s test. Theincidence of seizures in Experiment 3 was evaluated by Fisherexact test (SigmaStat1 SPSS). The level of statistical significancewas set at 5%.

3. Results

3.1. Experiment 1: rhythmic metrazol activity induced by

the 20-mg/kg dose of PTZ

The 20-mg/kg dose of PTZ induced episodes of spike-and-waveactivity with a frequency of 4–5 Hz (RMA, Fig. 1) accompanied bybehavioral arrest in both age groups. This spike-wave rhythmpredominated in frontal over occipital regions. Two-way ANOVAshowed only a main effect of Age factor on the latency for theappearance of the first RMA [F1,55 = 18.70, p < 0.001], but neitherthe overall drug nor the age � drug interaction was significant. Thedifference between the latencies to the first SW and the firstgeneralized SW episode was not significant (Fig. 2A). Twenty-five-day-old rats exhibited shorter latencies for appearance of RMAepisodes than 18-day-old ones (Tukey’s test: *p < 0.01).

Analysis of variance demonstrated a main effect of drug on thenumber of RMA episodes in the first (10–15 min) interval after PTZinjection, [F2,56 = 11.70, p < 0.027]. Subsequent pairwise compar-isons confirmed that RMA episodes in 18-day-old rats were lessfrequent in the group treated with the higher dose of 20 mg/kgcaffeine at P7–P11 than in control animals (p < 0.03) (Fig. 2B). Inthe 20–25 min period after PTZ injection, similar difference wasfound but the level of significance was reached only in rats treatedwith lower dose of caffeine (p < 0.005) (Fig. 2B).

Shorter mean and total duration of RMA episodes weredemonstrated in the first interval in the 10-mg/kg caffeine groupin comparison with control 18-day-old rats (p < 0.05) (Fig. 2C,D).Data for the 20-mg/kg group are missing because of a nearlycomplete absence of the episodes. The interaction Age x Drug andthe main effect of age factor were evident for the mean and thetotal duration of episodes [F2,48 = 20.56, p < 0.0019 andF2,48 = 9.10, p < 0.02, respectively] 20-25 min after PTZ injection.Post hoc comparisons showed that the duration (both mean andtotal) of episodes was significantly decreased in the two caffeine-treated groups at P18 (Fig. 2 C, D).

No significant differences were found between control andcaffeine-treated groups at the age of 25 days.

3.2. Experiment 2: rhythmic metrazol activity induced by

the 40-mg/kg dose of PTZ

PTZ given in a dose of 40 mg/kg elicited a marked RMA in allcontrol rats in both age groups. This activity was again formed bySW rhythm with a frequency of about 4–5 Hz. There was nomarked difference between 18- and 25-day-old rats in the EEGpattern. Electroclinical correlation was evident with locomotorarrest and/or jerks of vibrissae or other facial muscles accompany-ing RMA episodes.

As in Experiment 1, 25-day-old control rats showed shorterlatencies for appearance of RMA episodes than 18-day-oldcontrols. The difference between the latencies to the first RMAand the first generalized RMA episode was negligible in either agegroup (Fig. 3A). Two-way ANOVA revealed a main effect of drug onthe latency of the first episode of RMA, [F2,51 = 15.85, p < 0.0035],and a significant age � drug interaction on both the latency of thefirst episode of RMA and of the first episode of generalized RMA[F2,51 = 18.64, p < 0.015; F2,51 = 13.03; p < 0.01, respectively].Tukey’s test confirmed that 18-day-old rats in either caffeine groupshowed shorter latencies than controls for both the first and thefirst generalized episodes of RMA (*p < 0.05) (Fig. 3A).

Although ANOVA demonstrated a main effect of drug factor onthe number of RMA episodes only in the second period after PTZinjection, [F2,51 = 11.27, p < 0.04], post-hoc test revealed thatcaffeine-treated groups did not significantly differ from controls(Fig. 3B). The 20-mg/kg group of 18-day-old rats showed asignificant decrease in the incidence of RMA episodes (20% ofanimals vs. 80% of control rats).

Age factor had an overall effect on the number and theduration (mean and total) of RMA episodes in the first (10–15 min)period after PTZ injection: number [F1,51 = 19.22, p < 0.001];mean duration [F1,47 = 19.43, p < 0.0002]; total duration[F1,47 = 11.48, p < 0.005], respectively. Furthermore, main effectof drug factor on the mean duration, [F1,47 = 19.43, p < 0.0002],and age � drug interaction on both mean and total duration ofRMA episodes [F2,47 = 12.29, p < 0.01; F2,47 = 19.95, p < 0.0015,respectively] were evident. Subsequent pairwise comparisonsconfirmed that both mean and total duration of RMA episodes wassignificantly prolonged in the first period (10–15 min) after PTZinjection in both caffeine groups at P18 (Fig. 3C and D). Mean butnot total duration of RMA episodes was significantly increased inthe second period after PTZ injection in 18-day-old rats treatedwith the lower dose of caffeine. Because of insufficient number of18-day-old animals with RMA episodes 20–25 min after PTZ in thehigher caffeine dose group, data are missing. There were nosignificant changes of these parameters in 25-day-old rats (Fig. 3Cand D).

3.3. Experiment 3: motor seizures after two subsequent doses of PTZ

There were no significant changes in the incidence and latencyof RMA episodes after the first, 20-mg/kg dose of PTZ whencompared with data from Experiment 1.

EEG pattern of minimal seizures after the second dose of PTZ(40 mg/kg) was represented by spike-wave activity with afrequency of about 3 Hz. Motor pattern of seizures was formedby clonic seizures of forelimb and head muscles, tonic componentwhen present was represented by a dorsiflexion of the head ortorsion of head and body. Caffeine-treated rats did not exhibit anychanges in the EEG and/or motor pattern of seizures. Incidence ofminimal clonic seizures were not significantly changed. They wereobserved in 57%, 88% and 63% of rats in the control, 10- and 20-mg/kg caffeine 18-day-old groups, respectively. Twenty-five-day oldrats exhibited these seizures in 29, 50 and 70% of animals in thecontrol, 10- and 20-mg/kg caffeine groups, respectively (Table 1).

Fig. 2. Effects of chronic treatment with caffeine (10 and 20 mg/kg, s.c.) during P7–P11 on spike-and-wave episodes (RMA) elicited by PTZ injection (20 mg/kg i.p.) in 18- and

25-day-old rats. (A) Latencies of 1st RMA and 1st GRMA (generalized RMA), (B) number of RMA episodes, (C) mean duration of RMA episodes and (D) total duration of RMA

episodes between the 10th and 15th min, and between the 20th and 25th min, respectively. Abscissae: control (white column), group with caffeine 10 mg/kg (vertical line

column), group with caffeine 20 mg/kg (hatched column). Ordinates: A – latency in seconds, B – number of episodes and C and D – duration in seconds. Asterisks mark

significant differences between caffeine groups and controls (ANOVA with subsequent pairwise comparisons using Tukey’s test).

J. Tchekalarova et al. / Seizure 18 (2009) 463–469466

Unexpectedly, generalized tonic–clonic seizures appeared,especially in 43% of control rats in both age groups. The incidenceof these seizures tended to increase to 75 in the 10-mg/kg caffeinegroup of 18-day-old animals (Table 1). Again, this tendency wasnot significant. Only control 25-day-old rats exhibited generalized

Table 1Effects of chronic treatment with caffeine (10 and 20 mg/kg, s.c.) during P7–P11 on

the incidence of clonic and generalized tonic–clonic (GTCS) seizures after

20 + 40 mg/kg PTZ at P18 and P25. Asterisks denote significant difference between

control and caffeine groups (Fisher exact test).

Age P18 P25

Incidence

Clonic GTCS Clonic GTCS

Controls 4/7 3/7 2/7 3/7

Caffeine 10 7/8 6/8 5/10 0/10

Caffeine 20 5/8 4/8 7/10 0/10

tonic-clonic seizures (three out of seven animals, i.e. 43%), whereasthis type of seizures was never observed in either caffeine group.Due to a low incidence of these seizures in control animals,complete suppression did not reach the level of significance.

4. Discussion

Our present results demonstrated marked difference in theeffect of repeated postnatal caffeine treatment on the two modelsof epileptic seizures characterized by spike-and-wave rhythm inthe EEG (rhythmic metrazol activity and minimal clonic seizures).Nonconvulsive seizures, a model of human absence seizures,induced by low doses of PTZ were markedly influenced by early lifecaffeine exposure. In contrast, convulsive model of humanmyoclonic seizures was not significantly affected as the incidence,motor and EEG patterns are concerned. This is the firstdemonstration of the influence of postnatal caffeine treatmenton a model of absence seizures.

Fig. 3. Effects of chronic treatment with caffeine (10 and 20 mg/kg, s.c.) during P7–P11 on RMA induced by the 40-mg/kg dose of PTZ in 18- and 25-day-old rats. All details as in

Fig. 2.

J. Tchekalarova et al. / Seizure 18 (2009) 463–469 467

Analysis of SW rhythm demonstrated thalamocortical mechan-ism of its generation33 with a marked role of thalamic reticularnucleus34 and GABAergic inhibition.35,36 Under control conditions,SW EEG rhythm cannot be induced before the third postnatal weekin rats20,32,37 therefore 18- and 25-day-old animals were studied.Our present results confirmed that the only difference in EEGpatterns between the two studied models is in the frequency ofEEG rhythmic activity.38 On the other hand, behaviorally theseseizures are completely different: minute jerks of vibrissaeaccompany RMA episodes induced by low doses of PTZ, whereasminimal clonic seizures present marked convulsions usuallyrestricted to head and forelimb muscles. These motor patternsindicate different spread of epileptic activity into the motor system– to the nuclei of cranial nerves in a model of absences and to thegenerator of minimal seizures localized in basal forebrain.39,40

These two models differ substantially in pharmacological sensi-tivity. They are differently affected by ethosuximide and to GABA-Breceptor agonist baclofen, especially in 18-day-old rats.38,41 Thethird type of epileptic activity recorded in our study, generalizedtonic–clonic seizures generated in the brainstem and accompanied

by a different EEG pattern (rhythmic spikes transformed intospike-and-wave activity around the transition from the tonic toclonic phase), was not expected under our experimental condi-tions. They appeared in spite of the fractionated administration ofPTZ used to avoid elicitation of this type of seizures. The intervalbetween the two PTZ injections was probably too short to preventaccumulation of the convulsant drug to a suprathreshold level inthe brain.

There is strong evidence that the effects of caffeine at the lowconcentrations are due to the antagonism of endogenousadenosine actions.10 Moreover, it could be speculated that eithermultiple systems are directly influenced or adenosinergic systemcould modulate activity of other systems. Although most studiessupport the hypothesis that antagonism of adenosine receptors isresponsible for convulsant action of methylxanthines, Chakra-barti42 suggested that other mechanisms might also be involved.Methylxanthines are considered to be weak ligands for benzodia-zepine site in vitro, in vivo studies suggest that they affect channelfunction of the GABAA receptor–channel complex.9 In spite of thefact that the affinities of caffeine and theophylline for diazepam-

J. Tchekalarova et al. / Seizure 18 (2009) 463–469468

binding sites associated with the GABAA receptor channel are verylow and might play a role in the convulsant action of high doses ofcaffeine and theophylline,43 it cannot be excluded that postnatalcaffeine treatment may cause early adaptive changes affecting alsoGABAA receptor supramolecular complex. Our previous data onchanges in susceptibility to convulsants acting on this inhibitorysystem18 speak in favour of this possibility.

Early life caffeine treatment resulted in changes in the model ofnonconvulsive (absence) seizures in 18-day-old rats but it did notinfluence this model in 25-day-old rats, i.e. this effect is onlytransient. This effect changed qualitatively with the intensity ofepileptogenic stimulus (dose of PTZ). Anticonvulsant effect wasseen if low dose of PTZ was administered whereas the action of thehigher dose of PTZ (40 mg/kg) was potentiated. Reduced effects ofthe 20-mg/kg dose of PTZ (lower number of RMA episodes andtheir shorter duration) are in agreement with the action of thehigher caffeine dose on cortical epileptic afterdischarges.17

Surprisingly, we found a proconvulsant effect of postnatal caffeineexposure in RMA induced by 40 mg/kg PTZ. Because caffeine andother methylxanthines antagonize both adenosine A1 and A2

receptor subtypes with nearly equal affinities (with Ki values of 29and 48 mM, respectively) it is reasonable to assume that thesereceptor subtypes mediate equally caffeine effects. Adenosine A2A

receptors are present in the forebrain during the postnatal period15

and caffeine could modify the seizure threshold by its action onthis receptor subtype. The results from adult rat hippocampusshowed an interaction between co-expressed and co-localized A2A

and A1 receptors, providing a mixture of excitatory and inhibitorymodulation of neuronal excitability.44 Therefore, the observedincrease in seizure susceptibility in caffeine-treated groups at P18showed in RMA induced by 40 mg/kg PTZ might reflect disturbedbalance between inhibitory A1 and excitatory A2A receptors inbrain areas associated with these seizure phenomena. Thecontroversy between effects on RMA episodes induced by twodifferent doses of PTZ might be due to a change of the dose-effectrelationship. Our results suggest that this curve might be shifted tothe right but its slope may be steeper than in control animals butfurther analysis is needed.

A failure of significant effects of early life caffeine treatment inour second model – clonic seizures accompanied also by spike-wave rhythm – indicates participation of different mechanisms ordifferent brain structures in the effects on models of different typesof seizures. In contrast to lack of effect on minimal clonic seizureselicited by PTZ, our previous experiments demonstrated that the10-mg/kg dose of caffeine administered from P7 to P11 resulted ina proconvulsant effect in cortical epileptic afterdischarges inducedby cortical electrical stimulation.17 Both minimal clonic (metrazol)seizures and cortical epileptic afterdischarges exhibit identicalmotor pattern and are considered as an analogue of humanmyoclonic seizures.29,30 It indicates that not only motor pattern ofconvulsions but also the way how these seizures are triggeredplays a role in the consequences of early life caffeine treatment.

Transition to generalized tonic–clonic seizures in the twocontrol groups and their failure in both caffeine groups of 25-day-old rats was unexpected. Absence of these seizures in caffeine-treated 25-day-old animals might be due to earlier maturation ofadenosine receptors in the brainstem than in more rostral parts ofthe brain. Therefore, they may be influenced by caffeine exposureduring early postnatal development in a different way than not yetfully mature adenosine receptors in the forebrain. It remains to bestudied if this is a permanent change, i.e. if this effect will appearalso in adult animals. Data for cortical epileptic afterdischarges17

speak in favour of a transient effect.To conclude, our study demonstrated mixed anti- and

proconvulsant effects of early postnatal caffeine exposure in amodel of human absence seizures. These effects were age- and

dose-dependent. An outlined anticonvulsant effect on generalizedtonic–clonic seizures is in accordance with previous reportsdemonstrating anticonvulsant consequences of early life caffeineexposure in majority of seizure models.11,14,18

References

1. Jarvis M, Williams M. Adenosine in the central nervous system. In: Williams M,editor. Adenosine and Adenosine Receptors. New Jersey: Humana; 1990. p. 423–60.

2. Fredholm BB. Adenosine, adenosine receptors and the action of caffeine.Pharmacol Toxicol 1995;76:93–101.

3. Dragunow M. Adenosine receptor antagonism accounts for the seizure-prolonging effects of aminophylline. Pharmacol Biochem Behav 1990;36:751–5.

4. Trommer BL, Pasternak JF, Suyeoka GM. Proconvulsant effects of aminophyl-line during amygdala kindling in developing rats. Dev Brain Res 1989;46:169–72.

5. Bernaskova K, Mares P. Proconvulsant effect of aminophylline on corticalepileptic afterdischarges varies during ontogeny. Epilepsy Res 2000;39:183–90.

6. Mares P, Kubova H, Czuczwar SJ. Aminophylline exhibits convulsant action inrats during ontogenesis. Brain Develop 1994;16:296–300.

7. Millar D, Schmidt B. Controversies surrounding xanthine therapy. Semin Neo-natol 2004;9:239–44.

8. Comer AM, Perry CM, Figgitt DP. Caffeine citrate: a review of its use in apnoea ofprematurity. Paediat Drugs 2001;3:61–79.

9. Lopez F, Miller LG, Greenblatt DJ, Kaplan GB, Shader RI. Interaction of caffeinewith the GABAA receptor complex: alterations in receptor function but notligand binding. Eur J Pharmacol 1989;172:453–9.

10. Nehlig A, Daval JL, Debry G. Caffeine and the central nervous system: mechan-isms of action, biochemical metabolic and psychostimulant effects. Brain ResRev 1992;17:139–70.

11. Guillet R, Dunham L. Neonatal caffeine exposure and seizure susceptibility inadult rats. Epilepsia 1995;36:743–9.

12. Holloway WR. Caffeine: effects of acute and chronic exposure on the behavior ofneonatal rats. Neurobehav Toxicol Teratol 1982;4:21–32.

13. Hughes RN, Beveridge IJ. Sex and age-dependent effects of prenatal exposurecaffeine on open field behavior, emergence latency and adrenal weights in rats.Life Sci 1990;47:2075–88.

14. Guillet R. Neonatal caffeine exposure alters seizure susceptibility in rats in anage-related manner. Dev Brain Res 1995;89:124–8.

15. Johansson B, Georgiev V, Kuosmanen T, Fredholm BB. Long-term treatmentwith some methylxanthines decreases the susceptibility to bicuculline- andpentylenetetrazol-induced seizures in mice. Relationship to c-fos expressionand receptor binding. Eur J Neurosci 1996;8:2447–58.

16. Tchekalarova J, Kubova H, Mares P. Postnatal caffeine exposure: effects onmotor skills and locomotor activity during ontogenesis. Behav Brain Res2005;160:99–106.

17. Tchekalarova J, Kubova H, Mares P. Biphasic effect of chronic postnatal caffeinetreatment on cortical epileptic afterdischarges during ontogeny in rats. BrainRes 2006;1082:43–9.

18. Tchekalarova J, Kubova H, Mares P. Effects of postnatal caffeine exposure onseizure susceptibility in developing rats. Brain Res 2007;1150:32–9.

19. Cavazos JE, Sanchez R. Pathophysiology of seizures and epilepsy. In: Rho JM,Sankar R, Cavazos JE, editors. Epilepsy. New York: Marcel Dekker; 2004. p. 5–20.

20. Snead 3rd OC. Pharmacological models of generalized absence seizures inrodents. J Neural Transm 1992;35(Suppl.):7–19.

21. Schickerova R, Mares P, Trojan S. Rhythmic metrazol activity in rats as a modelof human absences. Activ Nerv Super 1989;31:16–20.

22. Brabcova R, Kubova H, Velısek L, Mares P. Effects of a benzodiazepine, bretazenil(Ro 16 16-60028) on rhythmic metrazol EEG activity. Comparison with stan-dard anticonvulsants. Epilepsia 1993;34:1135–40.

23. Gloor P. Generalized epilepsy with spike-and-wave discharge: a reinterpreta-tion of its electrographic and clinical manifestations. Epilepsia 1979;20:571–88.

24. Snead 3rd OC. Gamma-hydroxybutyrate in the monkey. II. Effect of chronic oralanticonvulsant drugs. Neurology 1978;28:643–8.

25. Marescaux C, Vergnes M, Depaulis A. Antiepileptic drug evaluation in a newmodel: spontaneous petit mal epilepsy in the rat. Fed Proc 1984;43:280–1.

26. Coenen AM, Van Luijtelaar EL. The WAG/Rij rat model for absence epilepsy: ageand sex factors. Epilepsy Res 1987;1:297–301.

27. Velısek L, Kubova H, Pohl M, Stankova L, Mares P, Schickerova R. Pentylenete-trazol-induced seizures in rats: an ontogenetic study. Naunyn SchmiedebergsArch Pharmacol 1992;346:588–91.

28. Engel Jr J. Seizures and epilepsy. Philadelphia: FA Davis Co.; 1989.29. Loscher W, Schmidt D. Which animal models should be used in the search for

new antiepileptic drugs? A proposal based on experimental and clinical con-siderations. Epilepsy Res 1988;2:145–81.

30. Mares P, Zouhar A. Do we possess adequate models of childhood epilepsies?Physiol Bohemoslov 1988;37:1–9.

31. Mares P, Kubova H. Electrical stimulation-induced models of seizures. In:Pitkanen A, Schwartzkroin PA, Moshe SL, editors. Models of seizures andepilepsy. Amsterdam: Elsevier; 2006. p. 153–9.

32. Schickerova R, Mares P, Trojan S. Correlation between electrocorticographic andmotor phenomena induced by metrazol during ontogenesis in rats. Exp Neurol1984;84:153–64.

J. Tchekalarova et al. / Seizure 18 (2009) 463–469 469

33. Gloor P, Avoli M, Kostopoulos G. Thalamocortical relationship in generalizedepilepsy with bilaterally synchronous spike-and-wave discharge. In: Avoli M,Gloor P, Kostopoulos G, Naquet R, editors. Generalized epilepsy. Boston: Bir-khauser; 1990. p. 190–212.

34. Avanzini G, de Curtis M, Marescaux C, Panzica F, Spreafico R, Vergnes M. Roleof the thalamic reticular nucleus in the generation of rhythmic thalamo-cortical activities subserving spike and waves. J Neural Transm 1992;35(Suppl):85–95.

35. Mulle C, Madariaga A, Deschenes M. Morphology and electrophysiologicalproperties of reticularis thalami neurons in cat: in vivo study of thalamicpacemaker. J Neurosci 1986;6:2134–45.

36. Spreafico R, de Curtis M, Frassoni C, Avanzini G. Electrophysiological charac-teristics of morphologically identified reticular thalamic neurons from ratslices. Neuroscience 1988;27:629–38.

37. Mares P, Maresova D, Trojan S, Fischer J. Ontogenetic development of rhythmicthalamo-cortical phenomena in the rat. Brain Res Bull 1982;8:765–9.

38. Mares P. Ontogeny of ethosuximide action against two seizure models in rats isdifferent. Life Sci 1998;63:51–7.

39. Browning RA, Nelson DK. Modification of electroshock and pentylenetetrazolseizure patterns in rats after precollicular transections. Exp Neurol 1986;93:546–56.

40. Mares P. Generalized seizures in rodents – only two patterns? In: Hirsch E,Andermann F, Chauvel P, Engel J, Lopes da Silva F, Luders H, editors. Generalizedseizures: from clinical phenomenology to underlying systems and networks.Montrouge: John Libbey Eurotext; 2006. p. 70–2.

41. Mares P, Slamberova R. Opposite effects of a GABAB antagonist in two models ofepileptic seizures in developing rats. Brain Res Bull 2006;71:160–6.

42. Chakrabarti A, Saini HK, Garg SK. Dose-finding study with nimodipine: aselective central nervous system calcium channel blocker on aminophylline-induced seizure models in rats. Brain Res Bull 1998;45:495–9.

43. Daly JW. Mechanism of action of caffeine. In: Garattini S, editor. Coffee andhealth. New York: Raven Press; 1993. p. 97–150.

44. Cunha RA, Milusheva E, Vizi ES, Ribeiro JA, Sebastiao AM. Excitatory andinhibitory effects of A1 and A2A adenosine receptor activation on the elec-trically evoked [3H]acetylcholine release from different areas of the rat hippo-campus. J Neurochem 1994;63:207–14.