Effects of levetiracetam on blood-brain barrier disturbances following hyperthermia-induced seizures in rats with cortical dysplasia

Page 1

Life Sciences 87 (2010) 609–619

Contents lists available at ScienceDirect

Life Sciences

j ourna l homepage: www.e lsev ie r.com/ locate / l i fesc ie

Effects of levetiracetam on blood-brain barrier disturbances followinghyperthermia-induced seizures in rats with cortical dysplasia

Bulent Ahishali a,⁎, Mehmet Kaya b, Nurcan Orhan c, Nadir Arican d, Oguzhan Ekizoglu e, Imdat Elmas d,Mutlu Kucuk c, Gonul Kemikler f, Rivaze Kalayci c, Candan Gurses g

a Department of Histology and Embryology, Istanbul Faculty of Medicine, Istanbul University, Istanbul, Turkeyb Department of Physiology, Istanbul Faculty of Medicine, Istanbul University, Istanbul, Turkeyc Institute of Experimental Medicine, Istanbul University, Istanbul, Turkeyd Department of Forensic Medicine, Istanbul Faculty of Medicine, Istanbul University, Istanbul, Turkeye Dr. Sadi Konuk Training and Research Hospital, Istanbul, Turkeyf Department of Radiotherapy, Istanbul Faculty of Medicine, Istanbul University, Istanbul, Turkeyg Department of Neurology, Istanbul Faculty of Medicine, Istanbul University, Istanbul, Turkey

⁎ Corresponding author. Department of Histology andof Medicine, Istanbul University, Capa 34093 Istanbul, Tu32360; fax: +90 212 414 22 84.

E-mail address: [email protected] (B. Ahishali).

0024-3205/$ – see front matter © 2010 Elsevier Inc. Aldoi:10.1016/j.lfs.2010.09.014

a b s t r a c t

a r t i c l e i n f o

Article history:

Received 15 April 2010Accepted 18 September 2010

Keywords:LevetiracetamBlood-brain barrierSeizures

Aims: The mechanisms underlying the changes in blood-brain barrier (BBB) integrity and the generation ofseizures in childhood associated with preexisting brain lesions like cortical dysplasia (CD) are poorlyunderstood. We investigated the effects of levetiracetam (LEV) on BBB integrity and the survival duringhyperthermic seizures in rats with CD.Main methods: Pregnant rats were exposed to 145 cGy of gamma-irradiation on embryonic day 17. Onpostnatal day 28, hyperthermia-induced seizures were evoked in offspring with CD. To show the functionaland morphological alterations in BBB integrity, quantitative analysis of sodium fluorescein (NaFlu)

extravasation, immunohistochemistry and electron microscopy were performed.Key findings: Seizure scores and mortality rates were decreased by LEV during hyperthermia-induced seizuresin rats with CD (Pb0.01). Increased NaFlu extravasation into brain by hyperthermia-induced seizures inanimals with CD was decreased by LEV (Pb0.01). While glial fibrillary acidic protein (GFAP) immunoreac-tivity slightly increased in brain sections of animals with CD during hyperthermia-induced seizures, LEV led toGFAP immunoreactivity comparable to that of controls. Decreased occludin immunoreactivity and expressionin CD plus hyperthermia-induced seizures was increased by LEV. Opening of tight junctions and abundance ofpinocytotic vesicles representing ultrastructural evidences of BBB impairment and severe perivascular edemawere observed in animals with CD exposed to hyperthermia-induced seizures and LEV treatment led to theattenuation of these findings.Significance: These results indicate that LEV may present a novel approach for the protection of the BBBbesides its antiepileptic impact on hyperthermic seizures in the setting of CD.

© 2010 Elsevier Inc. All rights reserved.

Introduction

Febrile seizures are themost frequently seen convulsions in infantsand young children. On the other hand, cortical dysplasia (CD), adevelopmental malformation of the neocortex, is recognized as one ofthe major causes of pediatric epilepsy. A number of studies indicatethat the underlying pathology in CD may lower the threshold toseizures (Toth et al., 1998; Chen et al., 1999; Porter et al., 2003; Dubeand Baram, 2006). Moreover, the presence of preexisting brain injurysuch as CD renders the immature more susceptible to seizures and

Embryology, Istanbul Facultyrkey. Tel.: +90 212 414 20 00

l rights reserved.

susceptibility is shown to be more pronounced in kindled animals(Germano et al., 1996; Kaya et al., 2008; Lin and Roper, 2006).

Although many forms of epilepsy are successfully managed withantiepileptic drug therapy, epilepsy due to CD is markedly resistant topharmacological treatment which is characteristic of these develop-mental lesions (Smyth et al., 2002; Wong, 2008). In recent humanand animal studies focusing on the involvement of the efflux pumpsin pharmacoresistance, overexpression of blood-brain barrier (BBB)efflux transporters in the brain capillary endothelial cells of theepileptic focus is reported (Sisodiya et al., 2003; Schmidt and Löscher,2009; Kuteykin-Teplyakov et al., 2009). While a large number ofstudies have mainly focused on the susceptibility to seizure induction,accumulating data concerning seizure-induced cell damage andantiepileptic treatment in cortical malformations (Smyth et al.,2002; Aronica et al., 2003; Oh et al., 2008) are being obtained from

Page 2

610 B. Ahishali et al. / Life Sciences 87 (2010) 609–619

recent studies concentrating on the BBB integrity in malformed brain(Marchi et al., 2006, 2007; Kaya et al., 2008; Gurses et al., 2009).Levetiracetam (LEV) is shown to be a highly effective antiepilepticdrug against partial and generalized seizures in several animal models(Löscher et al., 1998; Klitgaard et al., 1998; Cilio et al., 2009). AlthoughLEV has no effects on acutely generated seizures in animals (Löscheret al., 1998; Brandt et al., 2007), it markedly suppresses audiogenicseizures and kindling development (Löscher et al., 1998; Klitgaardand Pitkänen, 2003). In a recent study, it is shown that LEV may haveinhibitory effects in hippocampus and attenuates other neuronalpathways to secondary generalization in Noda epileptic rats (Ishimaruet al., 2010). Meanwhile intravenous administration of LEV in achemoconvulsant model of status epilepticus resulted in the atten-uation of behavioral manifestations of seizure discharge and in thereduction of neuronal injury (Zheng et al., 2010).

Studies investigating the impact of febrile seizures on BBB integrityhave shown disruption of BBB in humans and animals (van Eijsden et al.,2004; Suenaga et al., 2008). Because alterations in BBB integrity mayprecede neuronal damage in many neuropathologic conditions such asepilepsy, BBB impairment may also be involved in the initiation,progression andmaintenance of seizures (Seiffert et al., 2004; Ransohoff,2009). Two recent studies reported that antiepileptic drugs, LEV andtopiramate, could exert protective effects on BBB components againstfebrile seizures and seizures induced by pentylenetetrazole in kindledanimalswithCD(Łotowska et al., 2008;Gurses et al., 2009).However, theimpact of febrile seizures on BBB integrity in the setting of CD and theinfluence of LEV on the altered BBB components are unclear. The presentstudy is therefore, designed to investigate whether LEV alters seizureseverity and/or exerts protective effects on functional and structuralproperties of BBB during hyperthermia-induced seizures in rats with CD.

Materials and methods

Animal protocol and irradiation procedure

The experimental procedures were conducted in timed-pregnantSprague–Dawley rats (n=20) and their litters of both sexes. The day onwhich insemination was detected was determined as the embryonic day(E) 0. In order to induce CD, the animals were treated according to aprotocol that was previously described (Roper et al., 1995). On E17, thepregnant rats were exposed to gamma-irradiation. For this purpose, theanimals were anesthetized with sodium pentothal (35 mg/kg, i.p), andwere placed in groups of ten, each in prone position on awooden board. Apolystyrene phantom was put under the board to achieve an acceptablebackscatter. A nominal single dose of 145 cGy to mid-plane of theabdominal area was delivered by 6×27 cm posterior field using Co-60tele-therapyunit (Alcyon II,General Electric, France).After irradiation, ratswere taken back to the animal facility and cared for routinely until birth.After parturition, litters were allowed to live a 12-hour light/dark cyclewith free access to food andwater until the ageof 28 days. The litterswererandomly divided into the following experimental groups; CD, CD plushyperthermia, CD plus LEV, and CD plus hyperthermia plus LEV, while in-utero untreated litters were randomly assigned to control, hyperthermiaand hyperthermia plus LEV groups. The experimental protocols used inthis study were approved by the Ethics Committee for AnimalExperimentation of the Institute of Experimental Medicine, IstanbulUniversity. All effortsweremade tominimize animal suffering and reducethe number of animals used. Separate sets of experimental groupsconsisting of 8 rats were used for each experimental procedure exceptwestern blotting which was performed in experimental groups of 4 rats.

Administration of LEV and induction of hyperthermia

LEV (UCB Pharma, Belgium), dissolved in saline, was intravenouslyadministered in a dose of 100 mg/kg through a catheter inserted intoright femoral vein to 28-day-old rats under a mild dietyl ether

anaesthesia. Twentymin later when the animals were fully awakened,severe hyperthermia with an average core (rectal) temperature of40 °C (39.5–40.5 °C) was induced by placing the rats in a translucentfiberglass chamber (74 cm×42 cm×30 cm). A heated air flow systemwas adjusted tomaintain the temperature of the chamber at 55–60 °C.Baseline core temperature wasmeasured using a probe connected to atemperature indicator. To induce hyperthermia, the core temperatureof animals was kept at 39.5–40.5 °C. Once seizures commenced,hyperthermia was continued, and core temperature was recordedevery 2 min. The convulsive behavior was observed for 30 min andclassified into the following stages as described by Racine (Racine,1972): 0—no behavioral changes; 1—facial movements, ear andwhisker twitching; 2—myoclonic convulsions without rearing; 3—myoclonic convulsions with rearing; 4—clonic convulsion with loss ofposture; 5—generalized clonic–tonic seizures.

Measurement of arterial blood pressure and BBB permeability

The rats were anaesthetized with diethyl ether and polyethylenecatheters (PE-10) were inserted into the right femoral artery and vein.The former catheter was connected to a pressure transducer whichwasinterfaced to a data acquisition system (iWorx/ETH-256) to continu-ouslymonitor themean arterial blood pressure by a personal computer,and the latter catheterwasused to infuse sodiumfluorescein (NaFlu; 2%,5 ml/kg) tracer. BBB permeability was assessed by measuring the brainlevel of extravasated NaFlu. In animals which were exposed tohyperthermia, NaFluwas administeredwhen core temperature reached39.5 °C and allowed to circulate for 35 min. Then, rats were transcar-dially perfused with 100 ml of saline for about 15 min to removeintravascular NaFlu. After decapitation, brains were removed anddissected into four regions; left cerebral cortex, right cerebral cortex,diencephalon and cerebellum. Each brain region was weighed forquantitative measurement of NaFlu extravasation. Brain samples werehomogenized in 2.5 ml phosphate-buffered saline and mixed with avortex for 2 min after the addition of 2.5 ml 60% trichloroacetic acid toprecipitate protein. Samples were later cooled for 30 min andcentrifuged at 14000 g for 10 min. The concentration of tracer in thesupernatant was measured at excitation wavelength of 440 nm andemission wavelength of 525 nm using a spectrophotofluorometer(Microplate Reader; DTX880 Multimode Detector, Beckman Coulter).NaFlu was expressed as ng/mg of brain tissue against a standard curve.

Immunohistochemistry

To demonstrate tight junction (TJ) protein, occludin, and glialfibrillary acidic protein (GFAP) immunoreactivity in the brain sec-tions, rats were anaesthetized with diethyl ether. A bolus transcardialperfusion with 20 ml saline was administered at a pressure of110 mmHg for 15 s, followed by 100 ml fixative (4% paraformaldehydein phosphate buffer; pH: 7.4) for 10 min. After the perfusion, brainswere removed, immersed in the same fixative, kept for 24 h at 4 °C, andthen embedded inparaffin. 5-μmthick sectionswere deparaffinized andincubated in 10 mM citrate buffer solution (pH: 6.0) at 1 atm pressurefor 2 min for GFAP or incubated with protease (1 mg/ml; Sigma,USA) for 10 min for occludin to achieve antigen retrieval. Endogenousperoxidase activity was quenched using 0.3% hydrogen peroxide for20 min. A nonspecific blocking reagent (Ultra-V-Block, Lab Vision,Westinghouse, CA) was used to prevent nonspecific binding. Monoclo-nal mouse anti-GFAP (Neomarker, Fremont, CA; 1/100, 60 min) andpolyclonal rabbit anti-occludin (Zymed, CA; 1/50, 2 h) were used asprimary antibodies. Secondary antibodies were biotinylated goat anti-mouse (Lab Vision, Westinghouse, CA) for GFAP and biotinylated goatanti-polyvalent for occludin. After washing, streptavidin–peroxidasecomplex was applied and aminoethyl carbazole chromogen was used.The sectionswere counterstainedwithMayer's hematoxylin to enhancenuclear staining. Images were obtained from the hippocampus by

Page 3

611B. Ahishali et al. / Life Sciences 87 (2010) 609–619

means of a digital camera (Nikon, Coolpix 4500) attached to a lightmicroscope. Hematoxylin and eosin staining was performed on parallelparaffin brain sections using standard techniques for histologicalconfirmation of CD development in in-utero irradiated rats.

Western blotting

Brain tissues from the CA1 region of hippocampus from rats werehomogenized in lysis buffer [50 mM Hepes, 150 mM NaCl, 0.5%Triton-X 100, 1 mM DTT and one protease inhibitor tablet (Roche)].Samples were centrifuged at 13,000 rpm at 4 °C for 10 min, and thesupernatant was extracted and stored at –20 °C. Tissue lysates werenormalized according to Lowry-method. Denatured 20 μg cell lysateswere separated by 12.5% SDS-PAGE at 90 V for 2 h and transferred to0.45 μm polyvinylidene difluoride membranes (Millipore) by semidrytransfer (Bio-Rad, CA, USA). Membranes were blocked in tris-bufferedsaline (TBS; 50 mM Tris–HCl, pH 8.0, 150 mMNaCl, 5% BSA fraction V)for 30 min at room temperature (RT). Incubation with polyclonalrabbit anti-occludin antibody (Zymed, CA) at a concentration of2.5 μg/ml in TBS was performed at RT for 2 h. After three washes withTBS, membranes were incubated for 45 min with horseradishperoxidase-conjugated secondary antibody diluted 1.0 μg/ml in TBSand subsequently washed with TBST [50 mM Tris–HCl, pH 8.0,150 mM NaCl, 0.1% (v/v) Tween 20]. Chemiluminescence detectionwas performed by exposing Biomax MR films (Kodak) to Lumi-Lightwestern blotting substrate-treated membranes (Pierce, USA). Themolecular masses of protein fragments were determined using aprotein ladder (Fermentas, Germany). Occludin band intensity wasquantified using Image-Pro Plus software (Ver.6.0). The intensitysignal of each band in each group was determined and the meanintensity was then recorded for statistical comparisons.

Electron microscopy

Animals were anesthetized with diethyl ether and perfusion-fixedthrough the heart with 2% paraformaldehyde and 2.5% glutaraldehydein phosphate buffer following a brief bolus administration of saline.The brains were removed and 1 mm3 tissue samples obtained fromthe CA1 region of hippocampus were kept in the same fixativeovernight at 4 °C. The samples were post-fixed in 1% osmiumtetroxide for 1 h, and then embedded in Epon. CA1 region ofhippocampus was selected as the area of interest on 1 μm semi-thinsections. Then, ultrathin sections (60 nm) were stained with uranylacetate and lead citrate and examined under a transmission electronmicroscope (JEOL, 1011, Japan).

Statistical analysis

Data were presented as mean±S.E.M. Group differences weredetermined by one way analysis of variance followed by Tukey andWilcoxon tests using a computer program (SPSS 11.0 ver.). In all cases,differences between the means were considered significant if Pb0.05.

Results

Immediately after the onset of hyperthermia-induced seizures, theblood pressure increased from 115±7.5 to 154±12mmHg (Pb0.01) inintact rats, and from 115.5±2 to 156±13mmHg in rats with CD(Pb0.01), while LEV treatment did not decrease the increased bloodpressure in hyperthermia-induced seizures in intact animals (151±8mmHg). An increase in blood pressure from 106±4.5 to 148±6.2 mmHg was also observed in LEV treated animals with CD uponexposure to hyperthermic seizures (Pb0.01). Baseline core (rectal)temperature values ranged between 35.8 and 36.6 °C in all rats. Followingthe maintenance of the rats in the chamber for 10–15 min, coretemperature values increased to 39.5–40.5 °C. Racine scales for hyper-

thermic seizures in intact animals and animalswithCDwere 3.47±0.20and 4.48±0.20, respectively. These scales were decreased by LEV to2.38±0.27 and 3.30±0.27, respectively (Pb0.01). The durations ofhyperthermic seizureswere 11±1.4 min in intact animals, 15±2.1 minin animals with CD and 8±1.5 min in LEV treated animals with CD.Following hyperthermic seizures,mortality rates in intact rats, rats withCD and LEV treated rats with CD were 28%, 76% and 16%, respectively.

Hyperthermia or CD did not cause an alteration in BBB perme-ability to NaFlu. The increased NaFlu content in selected brain regionsin rats with CD during hyperthermia-induced seizures was observedto fall to levels comparable to that of controls by LEV treatment (Fig. 1;Pb0.01). NaFlu dye content in the brain regions of animals with CDwas observed to fall to levels below control values by LEV; however,the difference was not statistically significant.

Hematoxylin and eosin staining of brain sections from animalswith CD (Fig. 2B) revealed the presence of architectural abnormal-ities with lack of columnar organization and dyslamination in thecerebral cortex and neuronal dispersion in hippocampus comparedto the intact neuronal architecture of control animals (Fig. 2A).Immunoreactivity for GFAP in all regions and layers of hippocampalformation was observed to be slightly increased during hyperther-mia-induced seizures in rats with CD, while LEV treatment in theseanimals led to GFAP immunoreactivity comparable to that ofcontrols (Fig. 3A, F and G). GFAP immunohistochemistry in CD,hyperthermia, CD plus LEV and LEV plus hyperthermia revealed animmunostaining pattern similar to that of controls (Fig. 3A–E).Immunohistochemistry and western blotting analysis of occludin,an expected 65-kDa integral membrane protein, are shown in Figs. 4and 5, respectively. A decrease in the immunoreactivity of occludinwas noted during hyperthermia-induced seizures in rats with CDcompared to controls (Fig. 4A and F). In these animals, theimmunoreactivity for occludin was increased by LEV treatment(Fig. 4G). The pattern of occludin immunostaining in CD, hyper-thermia, CD plus LEV, and LEV plus hyperthermia was similar to thatof control (Fig. 4 A–E). Western blotting and quantification of theintensity of protein bands in experimental groups revealed that theslightly decreased expression of occludin in CD plus hyperthermiawas significantly increased by LEV treatment to values higher thancontrols (Fig. 5A and B; Pb0.05).

Electron microscopic examination revealed pericapillary edemawith intensive swelling of astrocytic end-feet frequently leading tocompression of capillaries and narrowing of their lumens follow-ing hyperthermia-induced seizures in rats with CD (Fig. 6F1–3).TJs between endothelial cells in the CA1 region of hippocampuswere occasionally observed to be open in rats with CD exposed tohyperthermia-induced seizures (Fig. 6F3). Frequent pinocytoticvesicles were encountered in the cytoplasm of endothelial cellsfollowing hyperthermia-induced seizures in rats with CD, and thesevesicles were mostly uniform in size and formed endocytotic pits inboth luminal and abluminal faces (Fig. 6F2). Occasionally, conglom-erates of vesicles of uniform size with approximately 40 nm diametersenclosed within a membrane were also observed (Fig. 6F4). Upon theadministration of LEV, TJs exhibited the intact closed ultrastructure,pericapillary edema was dramatically improved and the increasedfrequency of pinocytotic vesicles returned to levels comparable to thatof controls in these animals (Fig. 6G). No remarkable changes in theultrastructure of BBB components were observed in animals of CD, CDplus LEV and LEV plus hyperthermia groups, while the endothelium ofBBB was relatively normal with only a minimal perivascular astrocyticedema following hyperthermia-induced seizures in intact animals(Fig. 6A–E).

Discussion

In this study, we investigated the effects of LEV on the cellular andmolecular mechanisms involved in the BBB disruption induced by

Page 4

Fig. 1. NaFlu dye content in the brain regions of animals. Data are shown as mean±S.E.M. *Pb0.01 versus other groups.

612 B. Ahishali et al. / Life Sciences 87 (2010) 609–619

hyperthermic seizures in rats with CD. Our data demonstrated for thefirst time that the treatment of established hyperthermia-inducedseizures in rats in the setting of CDwith LEV exerted a protective effecton the seizure-induced BBB disruption. LEV also reduced the seizureseverity and mortality rate during hyperthermia-induced seizures inanimals with CD.

Hyperthermia-induced seizures in rats have been used as a modelof febrile seizures by precipitating hyperthermic convulsions at 26–29 days of life (Morimoto et al., 1991). Therefore, to study the effects ofLEV on febrile seizures we chose to induce seizures by hyperthermia inimmature rats at the age of 28 days. We demonstrated that hyperther-mia-induced seizures occurred in 28 day-old-ratswith CDwith a highermortality rate compared to those of intact rats and the mortality inthe former setting was decreased dramatically by LEV. Our findings are

Fig. 2. Photomicrographs of hematoxylin and eosin stained brain sections from intact animalhorizontal lamination and loss of orientation of pyramidal cells.

supported by Germano et al. (1996) who reported that hyperthermia-induced seizures in MAM-treated animals cause higher mortality ratescompared to normal animals. On the other hand, human studies haveindicated that there is a link between prolonged febrile seizures andtemporal lobe epilepsy despite a long lag period (Cendes et al., 1993;Hamati-Haddad and Abou-Khalil, 1998). However, it has been pointedout that this proposition needs to be verified by strong evidence fromprospective studies (Shinnar et al., 2001, 2008). The other issues to beconsidered are why some children have febrile seizures and why somefebrile seizures last long. In many temporal lobe epilepsy cases withmesial temporal sclerosis on the MRI and history of febrile seizures,a ‘dual pathology’ is found in the pathologic specimen at the time ofsurgery (VanLandingham et al., 1998). In this context, the current studyalso supports the possibility of a dual pathology. Our observations are

s (A), and animals with CD (B) showing neuronal dispersion in hippocampus, disrupted

Page 5

Fig. 3. GFAP immunoreactivity in hippocampal regions in control (A), CD (B), hyperthermia (C), CD plus LEV (D), hyperthermia plus LEV (E), CD plus hyperthermia (F) and CD plus LEVplus hyperthermia (G) groups. A slight increase in GFAP immunoreactivity was noted in CD plus hyperthermia (F) which decreased to levels comparable to that of controls by LEV (G).Scale bars=25 μm.

613B. Ahishali et al. / Life Sciences 87 (2010) 609–619

also in linewith aprevious study indicating that LEVcompletely controlsseizures during thefirst 3 days of treatment in intact rats (vanVliet et al.,2008). Furthermore, human and animal studies have indicated that LEV

can reduce seizure frequency and activity in normal and malformedbrain (Löscher and Hönack, 1993, 2000; Klitgaard et al., 1998; Collins etal., 2006; Cilio et al., 2009). On the contrary, LEV was found to be

Page 6

Fig. 4. Occludin immunoreactivity (arrows) in hippocampal regions in control (A), CD (B), hyperthermia (C), CD plus LEV (D), LEV plus hyperthermia (E), CD plus hyperthermia (F) andCDplus LEV plus hyperthermia (G) groups. Note that the decreased occludin immunoreactivity in CDplus hyperthermia (F) in the capillaries is increased by LEV treatment (G). Nomarkedalterations were observed in the endothelial TJs of capillaries in other groups. Scale bars=10 μm.

614 B. Ahishali et al. / Life Sciences 87 (2010) 609–619

ineffective in electroshock and pentylenetetrazole-induced seizures inintact animals (Gower et al., 1995; Klitgaard et al., 1998). Therefore, ithas been suggested that antiepileptic and neuroprotective effects of LEV

might be dose-dependent and the differences in pharmacokineticsmight lead to more rapid elimination of the drug in rats than humans(Doheny et al., 1999; Brandt et al., 2007; Ueda et al., 2009). Our findings

Page 7

Fig. 5. Expression of occludin in hippocampus documented by western blotting analysis. Four occludin bands in the same experimental groups represent samples from 4 differentanimals (A). Note the significant increase in the slightly decreased occludin band intensity in CD plus hyperthermia by LEV (B). *Pb0.05 versus CD plus hyperthermia and control.

615B. Ahishali et al. / Life Sciences 87 (2010) 609–619

are also in concordance with data indicating that experimentally-induced CD renders immature rats more susceptible to hyperthermia-induced seizures (Germano et al., 1996; Scantlebury et al., 2004).

Our data indicated that the degree of BBB disruption resulting fromhyperthermia-induced seizures may be related to the development ofCD in rats. In MAM-treated rats, aberrant vessel morphology andserum albumin leakage in themalformed hippocampuswere reportedand BBB leakage was observed to be exacerbated by seizures (Marchiet al., 2006). It has also been shown that hyperperfusion in lesionaltissue of focal CD occurs during ictal events suggesting that thishemodynamic factor may contribute to an increase of albuminextravasation into brain (Roch et al., 2002). Our study clearlydemonstrated that LEV decreased the increased BBB permeability inhippocampus induced by hyperthermic seizures in rats with CD. Thisfinding is in well agreement with a recent study in which anantiepileptic drug, topiramate, was morphologically shown to exertprotective effects on BBB components in the ammonal cortex in an

experimental model of febrile seizure in rats (Łotowska et al., 2008).In our study, many capillaries in the hippocampus of rats with CDexposed to hyperthermia-induced seizures were surrounded bymarked perivascular edema as evidenced ultrastructurally by capil-laries with narrow lumens compressed by the grossly swollensurrounding astrocytic foot processes. We also observed that duringhyperthermia-induced seizures, TJs were occasionally opened andpinocytotic vesicles were increased in capillary endothelial cells in theCA1 region of the hippocampus of rats with CD and theseultrastructural alterations were attenuated upon LEV pretreatment.It is well-known that decreased expression of occludin protein inendothelial cells leads to increased BBB permeability. Interestingly,we revealed that LEV treatment increased occludin immunoreactivityand expression in brain capillary endothelial cells of rats with CDduring hyperthermic seizures. On the other hand, enhanced barrierfunction in chronic epilepsy may also account for the preservation ofexpression patterns of TJ protein along with intact ultrastructural

Page 8

616 B. Ahishali et al. / Life Sciences 87 (2010) 609–619

appearance of BBB (Kasantikul et al., 1983). Taken together, our dataindicate that enhancement of TJ protein expression by LEV may alsohelp the drug exert a protective effect on paracellular pathway byacting on the functional and the morphological properties of the BBBduring hyperthermic seizures in animals with CD.

We observed no significant effects of LEV on enhanced bloodpressure during hyperthermic seizures of kindled rat with CD. On theother hand, it is well-known that acute hypertension stimulates theproduction of pinocytotic/caveolar vesicles in endothelial cells. In ourstudy, following hyperthermic seizures in rats with CD which led to anincrease in arterial blood pressure, we observed increased pinocytoticvesicles in the endothelial cells suggesting increased transcytoticactivity which was attenuated by LEV treatment. In addition, theconglomerates of vesicles that we observed in the endothelial cells

Fig. 6. Electron micrographs of capillaries in the CA1 region of hippocampus. Note the pecompression of capillaries and narrowing of lumens (F1 and F3), an opened TJ (inset; openendothelial cells of rats with CD exposed to hyperthermia-induced seizures. Following hypenormal with only a minimal perivascular astrocytic edema (asterisk; C). The administratiohyperthermia (E) groups exhibited intact ultrastructural features of BBB components simil

following hyperthermia-induced seizures in rats with CD disappearedupon LEV treatment. The exact mechanism of the action of LEV in theinhibition of transcellular pathways might be related to an effect of thedrug on endothelial cells of BBB through a process similar to that insynaptic release involving a putative synaptic vesicle protein (SV2). LEVhas been shown to bind an integral membrane protein, SV2, which ismostly localized in neurons (Lynch et al., 2004; Kaminski et al., 2008)and is also abundantly present in endothelial cells of retinal bloodvessels (Stojic et al., 2007). It is likely that SV2 proteins modulate theformation of the soluble N-ethylmaleimide-sensitive factor attachmentprotein receptor (SNAREs)-complexes that mediate vesicle fusion bybridging between opposed lipid bilayers (Lezzi et al., 2005). Freetransport vesicles in endothelial cells are targeted by the vesicleidentifiers, v-SNAREs, which can bind to their cognate t-SNAREs,

ricapillary edema with intensive swelling of astrocytic end-feet (asterisks) leading toarrow; F3), frequent pinocytotic vesicles (F2) and a conglomerate of vesicles (F4) inrthermia-induced seizures in intact animals, BBB components are seen to be relativelyn of LEV led to reversal of these findings (G). CD (B), CD plus LEV (D) and LEV plus

ar to controls (A).

Page 9

Fig. 6 (continued).

617B. Ahishali et al. / Life Sciences 87 (2010) 609–619

present on the target membrane (Schnitzer et al., 1995; Rippe et al.,2002). Therefore, it can be concluded that SV2 blockade by LEV mayhave led to the inhibition of SNAREs functions which may account forthe effect of the drug on BBB permeability during hyperthermic seizuresin the setting of CD.

Data from published literature suggest that integrity of BBB can beimpaired due to a decline in GFAP expression implying that anincrease in GFAP expression may exert stimulatory and protectiveeffects on BBB (Pekny et al., 1998). On the other hand, it is reportedthat temperature-dependent changes in albumin immunoreactivityin several brain structures tightly correlate with GFAP immunoreac-tivity (Kiyatkin and Sharma, 2009). Moreover, rapid GFAP expressionhas been reported under several conditions associated with robusthyperthermia and brain edema (Cervós-Navarro et al., 1998; Kiyatkinet al., 2007). We observed that GFAP immunoreactivity was slightlyincreased in rats with CD by hyperthermia-induced seizures, while

LEV treatment led to GFAP immunoreactivity comparable to that ofcontrols. Therefore, our data suggest that alterations in GFAPexpression, at least partly, may be related to the improvement ofBBB integrity during hyperthermia-induced seizures in rats with CDfollowing LEV treatment.

Conclusion

Our observations that LEV treatment led to complete alleviation ofthe increased BBB permeability in hyperthermic seizures in rats withCD along with the inability of the drug to completely abolish theRacine scores of hyperthermic seizures in these animals suggest thatLEV provides protective effects against hyperthermic seizure-inducedBBB disruption independent of the anticonvulsant action of thedrug in the setting of CD. Furthermore, LEV decreases the increasedBBB permeability not only by acting on paracellular pathway leading

Page 10

618 B. Ahishali et al. / Life Sciences 87 (2010) 609–619

to increased expression of occludin, but also on transcellular pathwayby decreasing pinocytotic activity in barrier type of brain capillaryendothelial cells.

Conflicts of interest statement

The authors declare that there are no conflicts of interest.

Acknowledgments

The present work was supported by the Research Fund of IstanbulUniversity (Project Number: 3927/2009) and UCB Pharma, Belgium.The sponsors of the study had no role in the study design, collection,analysis, and interpretation of data, in the writing of the report, or inthe decision to submit the paper for publication.

References

Aronica E, Gorter JA, Janseni GH, van Veelen CW, van Rijen PC, Leenstra S, et al.Expression and cellular distribution of multidrug transporter proteins in two majorcauses of medically intractable epilepsy: focal cortical dysplasia and glioneuronaltumors. Neuroscience 2003;118:417–29.

Brandt C, Glien M, Gastens AM, Fedrowitz M, Bethmann K, Volk HA, et al. Prophylactictreatment with levetiracetam after status epilepticus: lack of effect on epileptogen-esis, neuronal damage, and behavioral alterations in rats. Neuropharmacology2007;53:207–21.

Cendes F, Andermann F, Dubeau F, Gloor P, Evans A, Jones-Gotman M, et al. Earlychildhood prolonged febrile convulsions, atrophy and sclerosis of mesial structures,and temporal lobe epilepsy: an MRI volumetric study. Neurology 1993;43:1083–7.

Cervós-Navarro J, Sharma HS, Westman J, Bongcam-Rudloff E. Glial reactions in thecentral nervous system following heat stress. Prog Brain Res 1998;115:241–74.

Chen K, Baram TZ, Soltesz I. Febrile seizures in the developing brain result in persistentmodification of neuronal excitability in limbic circuits. Nat Med 1999;5:888–994.

Cilio MR, Bianchi R, Balestri M, Onofri A, Giovannini S, Di Capua M, et al. Intravenouslevetiracetam terminates refractory status epilepticus in two patients withmigrating partial seizures in infancy. Epilepsy Res 2009;86:66–71.

Collins JJ, Tudor C, Leonard JM, Chuck G, Franz DN. Levetiracetam as adjunctiveantiepileptic therapy for patients with tuberous sclerosis complex: a retrospectiveopen-label trial. J Child Neurol 2006;21:53–7.

Doheny HC, Ratnaraj N, Whittington MA, Jefferys JG, Patsalos PN. Blood andcerebrospinal fluid pharmacokinetics of the novel anticonvulsant levetiracetam(ucb L059) in the rat. Epilepsy Res 1999;34:161–8.

Dube CM, Baram TZ. Complex febrile seizures—an experimental model in immaturerodents. In: Pitkanen A, Schwartzkroin PA, Moshe SL, editors. Models of Seizuresand Epilepsy. Elsevier Academic Press; 2006. p. 33-340. Chapter 26.

Germano IM, Zhang YF, Sperber EF, Moshe SL. Neuronal migration disorders increasesusceptibility to hyperthermia-induced seizures in developing rats. Epilepsia1996;37:902–10.

Gower AJ, Hirsch E, Boehrer A, Noyer M, Marescaux C. Effects of levetiracetam, a novelantiepileptic drug, on convulsant activity in two genetic rat models of epilepsy.Epilepsy Res 1995;22:207–13.

Gurses C, Ekizoglu O, Orhan N, Ustek D, Arican N, Ahishali B, et al. Levetiracetamdecreases the seizure activity and blood-brain barrier permeability in pentylenete-trazole-kindled rats with cortical dysplasia. Brain Res 2009;1281:71–83.

Hamati-Haddad A, Abou-Khalil B. Epilepsy diagnosis and localization in patients withantecedent childhood febrile convulsions. Neurology 1998;50:917–22.

Ishimaru Y, Chiba S, Serikawa T, SasaM, Inaba H, Tamura Y, et al. Effects of levetiracetamon hippocampal kindling in Noda epileptic rats. Brain Res 2010;1309:104–9.

Kaminski RM, Matagne A, Leclercq K, Gillard M, Michel P, Kenda B, et al. SV2A protein isa broad-spectrum anticonvulsant target: functional correlation between proteinbinding and seizure protection in models of both partial and generalized epilepsy.Neuropharmacology 2008;54:715–20.

Kasantikul V, Brown WJ, Oldendorf WH, Crandall PC. Ultrastructural parameters oflimbic microvasculature in human psychomotor epilepsy. Clin Neuropathol1983;2:171–8.

Kaya M, Gurses C, Kalayci R, Ekizoglu O, Ahishali B, Orhan N, et al. Morphological andfunctional changes of blood-brain barrier in kindled rats with cortical dysplasia.Brain Res 2008;1208:181–91.

Kiyatkin EA, Sharma HS. Permeability of the blood-brain barrier depends on braintemperature. Neuroscience 2009;161:926–39.

Kiyatkin EA, Brown PL, Sharma HS. Brain edema and breakdown of the blood-brainbarrier during methamphetamine intoxication: critical role of brain hyperthermia.Eur J Neurosci 2007;6:1242–53.

Klitgaard H, Pitkänen A. Antiepileptogenesis, neuroprotection, and disease modificationin the treatment of epilepsy: focus on levetiracetam. Epileptic Disorder 2003;5(Supply 1):S9-S16.

Klitgaard H, Matagne A, Gobert J, Wülfert E. Evidence for a unique profile of levetiracetamin rodent models of seizures and epilepsy. Eur J Pharmacol 1998;353:191–206.

Kuteykin-Teplyakov K, Brandt C, Hoffmann K, Löscher W. Complex time-dependentalterations in the brain expression of different drug efflux transporter genes afterstatus epilepticus. Epilepsia 2009;50:887–97.

Lezzi M, Theander S, Janz R, Loze C, Wollheim CB. SV2A and SV2C are not vesicular Ca2+

transporters but control glucose-evoked granule recruitment. J Cell Sci 2005;118:5647–60.

Lin DD, Roper SN. In utero irradiation as a model of cortical dysplasia. In: Pitkanen A,Schwartzkroin PA, Mosh SL, editors. Models of Seizures and Epilepsy. New York:Elsevier Academic Pres; 2006. p. 315–22.

Löscher W, Hönack D. Profile of ucb L059, a novel anticonvulsant drug, in modelsof partial and generalized epilepsy in mice and rats. Eur J Pharmacol 1993;232:147–58.

Löscher W, Hönack D. Development of tolerance during chronic treatment ofkindled rats with the novel antiepileptic drug levetiracetam. Epilepsia 2000;41:1499–506.

Löscher W, Hönack D, Rundfeldt C. Antiepileptogenic effects of the novel anticonvul-sant levetiracetam (ucb L059) in the kindling model of temporal lobe epilepsy.J Pharmacol Exp Ther 1998;284:474–9.

Łotowska JM, Sobaniec-Łotowska ME, Sendrowski K, Sobaniec W, Artemowicz B.Ultrastructure of the blood-brain barrier of the gyrus hippocampal cortex in anexperimental model of febrile seizures and with the use of a new generationantiepileptic drug—topiramate. Folia Neuropathol 2008;46:57–68.

Lynch BA, Lambeng N, Nocka K, Kensel-Hammes P, Bajjalieh SM, Matagne A, et al. Thesynaptic vesicle protein SV2A is the binding site for the antiepileptic druglevetiracetam. Proc Natl Acad Sci USA 2004;101:9861–6.

Marchi N, Guiso G, Caccia S, Rizzi M, Gagliardi B, Noé F, et al. Determinants of drug brainuptake in a rat model of seizure-associated malformations of cortical development.Neurobiol Dis 2006;24:429–42.

Marchi N, Angelov L, Masaryk T, Fazio V, Granata T, Hernandez N, et al. Seizure-promoting effect of blood-brain barrier disruption. Epilepsia 2007;48:732–42.

Morimoto T, Nagao H, Sano N, Takahashi M, Matsuda H. Electroencephalographic studyof rat hyperthermic seizures. Epilepsia 1991;32:289–93.

Oh HS, Lee MC, Kim HS, Lee JS, Lee JH, KimMK, et al. Pathophysiologic characteristics ofballoon cells in cortical dysplasia. Childs Nerv Syst 2008;24:175–83.

Pekny M, Stanness KA, Eliasson C, Betsholtz C, Janigro D. Impaired induction of blood-brain barrier properties in aortic endothelial cells by astrocytes from GFAP-deficient mice. Glia 1998;22:390–400.

Porter BE, Judkins AR, Clancy RR, Duhaime A, Dlugos DJ, Golden JA. Dysplasia: a commonfinding in intractable pediatric temporal lobe epilepsy. Neurology 2003;61:365–8.

Racine RJ. Modification of seizure activity by electrical stimulation: II. Motor seizure.Electroencephalogr Clin Neurophysiol 1972;32:281–94.

Ransohoff RM. Immunology: barrier to electrical storms. Nature 2009;457:155–6.Rippe B, Rosengren BI, Carlsson O, Venturoli D. Transendothelial transport: the vesicle

controversy. J Vasc Res 2002;39:375–90.Roch C, Leroy C, Nehlig A, Namer IJ. Magnetic resonance imaging in the study of the

lithium–pilocarpine model of temporal lobe epilepsy in adult rats. Epilepsia 2002;43:325–35.

Roper SN, Gilmore RL, Houser CR. Experimentally induced disorders of neuronalmigration produce an increased propensity for electrographic seizures in rats.Epilepsy Res 1995;21:205–19.

Scantlebury MH, Ouellet PL, Psarropoulou C, Carmant L. Freeze lesion-induced focalcortical dysplasia predisposes to atypical hyperthermic seizures in the immaturerat. Epilepsia 2004;45:592–600.

Schmidt D, Löscher W. New developments in antiepileptic drug resistance: anintegrative view. Epilepsy Curr 2009;9:47–52.

Schnitzer JE, Liu J, Oh P. Endothelial caveolae have the molecular transport machineryfor vesicle budding, docking, and fusion including VAMP, NSF, SNAP, annexins, andGTPases. J Biol Chem 1995;270:14399–404.

Seiffert E, Dreier JP, Iven S, Bechmann I, Tomkins O, Heinemann U, et al. Lasting blood-brain barrier disruption induces epileptic focus in the rat somatosensory cortex.J Neurosci 2004;24:7829–36.

Shinnar S, Berg AT, Moshe SL, Shinnar R. How long do new-onset seizures in childrenlast? Ann Neurol 2001;49:659–64.

Shinnar S, Hesdorffer DC, Nordli Jr DR, Pellock JM, O'Dell C, Lewis DV, et al. FEBSTATStudy Team. Phenomenology of prolonged febrile seizures: results of the FEBSTATstudy. Neurology 2008;71:170–6.

Sisodiya SM, Martinian L, Scheffer GL, van der Valk P, Cross JH, Scheper RJ, et al. Majorvault protein, a marker of drug resistance, is upregulated in refractory epilepsy.Epilepsia 2003;44:1388–96.

Smyth MD, Barbaro NM, Baraban SC. Effects of antiepileptic drugs on inducedepileptiform activity in a rat model of dysplasia. Epilepsy Res 2002;50:251–64.

Stojic J, Stöhr H,Weber BH. Three novel ABCC5 splice variants in human retina and theirrole as regulators of ABCC5 gene expression. BMC Mol Biol 2007;8:42–51.

Suenaga N, Ichiyama T, Kubota M, Isumi H, Tohyama J, Furukawa S. Roles of matrixmetalloproteinase-9 and tissue inhibitors of metalloproteinases 1 in acute enceph-alopathy following prolonged febrile seizures. J Neurol Sci 2008;266:126–30.

Toth Z, Yan XX, Haftoglou S, Ribak CE, Baram TZ. Seizure-induced neuronal injury:vulnerability to febrile seizures in an immature rat model. J Neurosci 1998;18:4285–94.

Ueda Y, Doi T, Takaki M, Nagatomo K, Nakajima A,Willmore LJ. Levetiracetam enhancesendogenous antioxidant in the hippocampus of rats: in vivo evaluation by brainmicrodialysis combined with ESR spectroscopy. Brain Res 2009;1266:1–7.

van Eijsden P, Notenboom RG, Wu O, de Graan PN, van Nieuwenhuizen O, Nicolay K. Invivo 1H magnetic resonance spectroscopy, T2-weighted and diffusion-weightedMRI during lithium-pilocarpine-induced status epilepticus in the rat. Braun KP.Brain Res 2004;1030:11–8.

Page 11

619B. Ahishali et al. / Life Sciences 87 (2010) 609–619

van Vliet EA, van Schaik R, Edelbroek PM, da Silva FH, Wadman WJ, Gorter JA.Development of tolerance to levetiracetam in rats with chronic epilepsy. Epilepsia2008;49:1151–9.

VanLandinghamKE,HeinzER, Cavazos JE, LewisDV.Magnetic resonance imaging evidenceof hippocampal injury after prolonged focal febrile convulsions. Ann Neurol 1998;43:413–26.

Wong M. Mechanisms of epileptogenesis in tuberous sclerosis complex and relatedmalformations of cortical development with abnormal glioneuronal proliferation.Epilepsia 2008;49:8-21.

Zheng Y, Moussally J, Cash SS, Karnam HB, Cole AJ. Intravenous levetiracetam in therat pilocarpine-induced status epilepticus model: behavioral, physiological andhistological studies. Neuropharmacology 2010;58:793–8.