Epilepsy Research (2008) 80, 119—131

journa l homepage: www.e lsev ier .com/ locate /ep i lepsyres

Effects of a static magnetic field on audiogenicseizures in black Swiss mice

Michael J. McLeana,∗, Stefan Engstroma, Zhang Qinkuna,Christopher Spankovichc, Daniel Polleyb,c

a Department of Neurology, Vanderbilt University Medical Center, Nashville, TN, USAb Department of Hearing and Speech Sciences, Vanderbilt Bill Wilkerson Center for Otolaryngology andCommunication Sciences, Vanderbilt University, Nashville, TN, USAc Vanderbilt Kennedy Center for Human Development, Vanderbilt University, Nashville, TN, USA

Received 28 June 2007; received in revised form 3 January 2008; accepted 13 March 2008Available online 9 June 2008

KEYWORDSStatic magnetic field;Permanent magnets;Auditory brainstemresponse;Animal seizuremodel;Audiogenic seizures;EEG

Summary Effects of a static magnetic field (SMF) with strong gradient components werestudied in black Swiss mice. Exposure to SMF (100—220 mT, 15—40 T/m for 1 h) did not affectthe threshold for detecting auditory brainstem responses. Serial seizures elevated the hear-ing threshold at some frequencies, but there was no difference between SMF-exposed andunexposed control mice. EEG changes were recorded during audiogenic seizures. Pretreatmentwith SMF prolonged seizure latency in response to stimulation with white noise of increasingintensity from 74 to 102 dBA (1 min interval between 2 and 4 dBA increments) without signifi-cant effects on seizure severity. Gender-related differences were not statistically significant.Stimulation with 10 min sound steps revealed prolongation of latency and reduction of seizureseverity in SMF-exposed, but not unexposed, mice. Pretreatment with phenytoin (5 mg/kg) incombination with SMF had significantly greater effects on seizure latency and severity than

either pretreatment alone. These findings indicate that the SMF studied here under differentconditions elevated seizure threshold and had anticonvulsant properties in Black Swiss miceand increased the efficacy of a conventional anticonvulsant drug.

ts re

© 2008 Elsevier B.V. All righ

∗ Corresponding author at: Vanderbilt/Stallworth RehabilitationHospital, Rm 1222J, Nashville, TN 37212, USA. Tel.: +1 615 936 0220;fax: +1 615 321 5247.

E-mail address: michael.mclean@vanderbilt.edu (M.J. McLean).

I

Fdpc2ib

0920-1211/$ — see front matter © 2008 Elsevier B.V. All rights reserved.doi:10.1016/j.eplepsyres.2008.03.022

served.

ntroduction

ailure to control seizures with pharmacological agentsrives the quest for novel therapies. More than 30% of

atients with epilepsy are estimated to have inadequateontrol while taking AEDs (Kwan and Brodie, 2000; Schmidt,002; French, 2006). Patients with inadequate response tonitial therapy or those who experienced multiple seizuresefore treatment were likely to have refractory seizures

1

(ipd1abtSsrt(2

ateuHV2taedcKifiotaoTm

cstsLL2vaAsaaic(iS1B2Fua

DZ

HofeltiasiTmrgercah(chfcms

ppitbgmituvawMapm

M

A

BhlHb

20

Kwan and Brodie, 2000). Results of large comparative stud-es of new and old AEDs consistently reveal that manyatients with partial epilepsy continue to have seizuresespite treatment (Mattson et al., 1985, 1992; Smith et al.,987; Brodie et al., 1995; Chadwick et al., 1998; Rowan etl., 2005). Ten AEDs were approved by regulatory agenciesetween 1993 and 2005. Many patients remain refractory tohe newer medications in monotherapy and polytherapy (seechmidt, 2002). Polypharmacy, even with the newer AEDs,eems unlikely solve the issue of refractoriness in the longun (Lhatoo et al., 2000; Schmidt, 2002). Because refrac-ory seizures reduce quality of life of patients with epilepsyGuekht et al., 2007; Theodore et al., 2006; Piazzini et al.,007), there is continuing need for novel treatments.

In addition to the continuing search for novel AEDs,nother approach has been to develop non-pharmacologicalherapies such as epilepsy surgery (e.g., for temporal lobepilepsy; see Spencer, 2002, for review), vagus nerve stim-lation (Helmers et al., 2003; Theodore et al., 2006; Deerdt et al., 2007), and the ketogenic diet (Freeman andining, 1999; Huffman and Kossoff, 2006; Gasior et al.,006). Each of these modalities has potential and limita-ions. Many patients require continued treatment with AEDsfter epilepsy surgery (Schmidt and Loscher, 2003; Schmidtt al., 2004) or experience recurrent seizures upon with-rawal of AEDs postoperatively, i.e., the patients were notured of their epilepsy by surgery (Schmidt et al., 2004;im et al., 2005). Vagus nerve stimulation (VNS) is lessnvasive. Early implantation of VNS resulted in seizure-reedom for 1 year in 11.8% of patients compared to 4.5%n patients with late implantation after more than 6 yearsf epilepsy (Helmers et al., 2003). The ketogenic diet, orhe less restrictive Atkins diet, is useful as adjunctive ther-py for some patients, but many patients do not benefitr are unable to comply with the regimen (Kossoff, 2004).hus, there is still a need for additional non-pharmacologicalodalities for the treatment of refractory seizures.As in the development of AEDs, demonstration of effi-

acy in disease-related animal models is an essential firsttep in the development of potential non-pharmacologicalreatments. Testing in animal seizure models is the main-tay of antiepileptic drug development (Krall et al., 1978;oscher, 1984; Kupferberg, 1989; White et al., 1998, 2002;oscher and Leppik, 2002; Sables et al., 2002; Smith et al.,007). A variety of models are used to screen for anticon-ulsant potential, predict the clinical spectrum of utilitynd justify investigation of potential agents in humans.udiogenic seizures (AGS) in rodents are a form of sen-ory reflex seizures produced by loud sound. The seizuresre considered to be a model of generalized seizures (Rossnd Coleman, 2000). The pathway for generation of AGSnvolving initiation in the colliculi and spread to the spinalord has been worked out in genetically epilepsy-prone ratsGEPR; see Faingold, 1999, for review). The same pathways assumed to be operant in other AGS susceptible strains.everal strains of mice, including Frings (Frings and Frings,953; Skradski et al., 1998, 2001; McMillan and White, 2004),

UB/BnJ (Johnson et al., 2005) and DBA/2 (Johnson et al.,006), GEPR (Dailey et al., 1989, 1996; Jobe et al., 1994;aingold et al., 1990) and other rat strains, are available forse in the screening process. Age-dependent hearing lossffects some strains of AGS-prone mice, including Fring’s,

etsww

M.J. McLean et al.

BA/2 and BUB/BnJ mice, and GEPR (Faingold et al., 1990;heng et al., 1999; Klein et al., 2005; Zhou et al., 2006).

Black Swiss mice were derived from National Institutes ofealth Swiss and C57BL/6 mice. Serendipitously, offspringf certain crosses between knockout mice and Black Swissemales were found to have AGS at 2—4 weeks of age (Misawat al., 2001, 2002). The seizures consisted of wild running,oss of righting, tonic flexion and extension and a brief pos-ictal period (Misawa et al., 2002). There was no changen hearing threshold with age as determined by brainstemuditory evoked response testing (Misawa et al., 2002). Theusceptibility to AGS in Black Swiss mice was found to benherited as a simple recessive trait (Misawa et al., 2002).he AGS susceptibility gene, jams1 (juvenile audiogeniconogenic seizures), localized to a 1.6 ± 0.5 centimorgan

egion on mouse chromosome 10 delimited by the basiginene and marker D10Mit140 (Misawa et al., 2002). Misawat al. (2002) note that the jams1 locus is syntenic to aegion on human chromosome 19p13.3 that has been impli-ated in a familial form of febrile convulsions (Puttagunta etl., 2000). Also, the gene for cystatin B, a mutant of whichas been implicated in progressive Baltic myoclonic epilepsyEPM1; Pennacchio et al., 1996), is near jams1 on mousehromosome 10. However, the identity of the gene productas not been ascertained. Misawa et al. (2002) observed uni-orm seizure susceptibility in their mice which were bred in alosed colony of homozygotes for multiple markers on chro-osome 10. We selected Black Swiss mice for the present

tudies.We have previously observed anticonvulsant effects of

retreatment with static magnetic fields (SMF). In AGS-rone DBA/2 mice, pretreatment with SMF reduced seizurencidence and potentiated anticonvulsant effects of pheny-oin (McLean et al., 2003). The SMF were time-invariantut spatially inhomogeneous with regions containing steepradients. Exposure to SMF produced by the same electro-agnetic array decreased severity of seizures elicited by

ntracerebroventricular administration of NMDA and addedo the anticonvulsant effect of phenytoin (McLean et al.,npublished). These are the only studies involving anticon-ulsant effects of static magnetic fields of which we areware and no controlled clinical trials have been reportedith static magnetic treatment devices (for review, seecLean et al., 2001; queries of on-line databases were neg-tive). Here we report effects of SMF produced by a type ofermanent magnet on sound-induced seizures in Black Swissice.

ethods

nimal care

lack Swiss mice (Tac:N:NIHS-BC; Taconic, Germantown, NY) wereoused under a 12 h/12 h light/dark cycle with food and water adibitum. Animal care met standards of the National Institutes ofealth. The experimental protocols were approved by the Vander-ilt University Institutional Animal Care and Use Committee. For

xperiments, mice were transported to the laboratory from a cen-ral animal care facility on the day of experiments. Mice with theame birth date were said to belong to the same batch. Experimentsere conducted between 08:00 and 16:00 h. Concomitant controlsere included in each experiment. Experiments were duplicated or

Effects of a static magnetic field on audiogenic seizures in black Swiss mice 121

Table 1 Summary of experiments, statistical comparison between groups

Experiment Exp day Nctrl Nexposed D p (exp-ctrl)

A.SMF vs. No SMF;protocol #1

Exp1 86 86 0.40 1.6 × 10−6

Exp2 71 75 0.44 8.6 × 10−7

Exp3 60 65 0.32 0.002

B.SMF vs. No SMF;protocol #2

Exp1 33 26 0.69 6.8 × 10−7

Exp2 24 23 0.71 5.2 × 10−7

Exp3 10 10 0.9 1.7 × 10−4

C.PHT ± SMFprotocol #1

Exp1 45 46 0.58 1.8 × 10−7

Exp2 42 46 0.33 0.0142 −4

me bring d

N2d

Exp3

Exp# refers to sequential day in a series of experiments on the sacontrol and exposed distributions and p is the probability (compa

triplicated. A total of 322 mice (164 in control groups; 158 in SMF-exposed groups) of both genders were used in this study. See Table 1for disposition by treatment group.

Magnetic devices

Mice were exposed to magnetic fields produced by a sin-gle cuboidal permanent magnet (Neotexx; Berlin, DE) made of

tnnis

Figure 1 The permanent magnet was analytically modeled by aindicated by the thick arrow. The smaller lines indicate field vectorsfield and gradients in a 10 mm × 10 mm × 2 mm volume over the sidecomponent parallel to the local field vector and Gq is the componeof its magnetization, the gradient is dominated by the parallel comit is the perpendicular component that dominates as can be seen inobserved but they are centered on zero and the dominating componein 2 mm slices at increasing distances from the surface of the magnet,in 2 mm steps.

45 0.44 2.0 × 10

atches of mice. D is the maximal vertical distance between theata from exposed mice to those from unexposed controls).

52 neodymium—iron—boron (NdFeB). The magnets measured5 mm × 10 mm × 10 mm and were magnetized through the 10 mmimension. These magnets produced static magnetic fields (SMF;

ime-invariant). Flux density was measured at the surface of a mag-et with a teslameter (FW Bell model 4048; Orlando, FL) to confirmumerical models of the field around the magnets. The dosimetrys displayed graphically in Fig. 1. At a distance of 5 mm from theurface of the magnet, the flux density (B, Tesla) and perpendicu-

ssuming uniform magnetization across the thinner dimension(direction and amplitude). Panel (a) shows the distribution ofwhere field lines exit normal to the surface. Gp is the gradientnt perpendicular to it. Away from the magnet in the directionponent (b). In a direction perpendicular to the magnetizationthe bottom graph (c). A wide spread of parallel components isnt is the perpendicular one. Both panels show the distributionstarting at 5 mm and ending at 15 mm (lowest fields/gradients)

1

lrFfinuwtt

E

Midhutpbitotfibt

PGaomiipduptct

AAatTaUiIs

4W#iuTSssls

n

taw(

tumDdwt2hfii

APfiwftfTAa2

febitimt(mst

sitdaww

A

Treesctpe

22

ar gradient component (Gq, T/m) were about 220 mT and 40 T/m,espectively. At 15 mm, the values were about 30 mT and 4 T/m.or most of the exposure time, the head of the mouse in the tubelled a volume between 6 and 12 mm from the surface of the mag-et. Dosimetry was calculated using an analytical model assumingniform magnetization in the NdFeB magnet. The field distributionas estimated in 10 mm × 10 mm × 2 mm slices at 2 mm intervals

hrough the position of the head of a mouse restrained in a syringeube.

xperimental procedure

ice were studied between 21 and 30 days of age, the period of max-mum susceptibility to audiogenic seizures. To minimize movementuring exposure to SMF, mice were restrained in 35 ml syringes thatad been perforated for ventilation and the plunger was advancedntil the nose of the mouse was near a hole produced by cuttinghe hub off the end of the syringe. Exposure to the SMF was accom-lished by attaching the magnet to the syringe tube with a rubberand. The magnet was positioned above the head of the mousenside. The long axis of the magnet was parallel to the long axis ofhe tube. The magnetization vector was tangential to the surfacef the tube to expose the mouse to field gradients perpendicularo the local field vector. Geomagnetic and stray electromagneticelds (e.g., from lights and power circuits) were not eliminated,ut such fields are at least two orders of magnitude smaller thanhe experimental SMF used here.

retreatmentroups of mice were pretreated in tubes (tube-restrained, TR) withnd without magnets for 1 h prior to auditory stimulation. The timef exposure was determined in previous experiments with otherouse strains (see McLean et al., 2005) and in preliminary exper-

ments with the black Swiss mice. For experiments testing thenteraction between the magnetic field and an anticonvulsant drug,henytoin was administered at a range of doses (0.5—40 mg/kg) toifferent groups of mice intraperitoneally 2 h prior to auditory stim-lation to achieve maximum blood levels. The mice were weighedrior to injection. Aliquots (0.1 ml per 10 g body weight) of pheny-oin (Sigma; St. Louis USA; dissolved in 0.9% saline) stock solutionalculated to produce the desired doses were injected intraperi-oneally.

uditory stimulationfter pretreatment, mice were removed from the syringe tubes onet a time and placed in a plexiglass cylinder (height, 7.5 in.; diame-er, 6 in.) with a speaker (model 40—1233; Radio Shack; Fort Worth,X, USA) centered in the lid. The speaker was driven by a publicddress amplifier (Optimus MPA-50; Radio Shack, Fort Worth, TX,SA) to which a DVD player was connected. A disc with a record-

ng of white noise generated with Mathematica (Wolfram Research,nc.; Champaign, IL) was played at different levels for auditorytimulation.

Sound intensity was measured with a handheld Extech Model07740 digital sound level meter (Extech Instruments Corporation;altham, MA). Two stimulation protocols were used. In protocol1, the sound intensity was increased manually in steps from an

nitial intensity of 74 dBA to 79, 82, 85, 87, 92, 96, 100 and 102 dBAntil a seizure occurred or the maximum sound level was reached.he first sound was played for 2 min to avoid startle responses.eizure-associated deaths were markedly reduced by eliminatingtartle responses in response to the onset of sound. Subsequent

teps were of 1 min duration (total duration 10 min). Thus, theatency to seizure initiation serves as a timebase and indicates theound intensity required to trigger seizures.

In protocol #2, 10 min steps were used to elicit seizures, begin-ing at 79 dBA. If no seizure occurred by the end of the first step,

umt(b

M.J. McLean et al.

he sound intensity was increased by one level for another 10 min,nd so on, until a seizure occurred. After three 10-min steps, miceithout seizures were exposed to 1 min of maximum intensity sound

102 dBA; total duration of sound exposure, 31 min).Because mice were tested for AGS multiple times, numbered ear

ags were used to identify each mouse throughout the series of stim-lations. Mice received the same pretreatment, e.g., TR control oragnetic field exposure, before each auditory stimulation session.aily stimulation resulted in many mice not having AGS, presumablyue to refractoriness; the same mice had AGS after several daysithout auditory stimulation. Stimulation every other day during

he period of maximum audiogenic seizure susceptibility between1 and 30 days of age was generally well-tolerated and resulted in aigh incidence of AGS each time. Treatment groups including micerom the same batch usually underwent 3 stimulation sessions dur-ng this period. Results from each session are presented separatelyn the figures.

udiogenic seizure stages and severityreconvulsive behaviors included excessive grooming that intensi-ed as seizure threshold was approached, a brief running episodeith or without hops and cessation of exploratory activity. This was

ollowed by a second episode of wild running (WR) that progressedo loss of righting (LOR) with truncal extension or neck flexion withorelimb extension followed by clonic jerking of the limbs (CLO).he seizure was defined by progression from WR to LOR and CLO.small number of mice also had tonic hindlimb extension (THE)

nd respiratory arrest (death, DEA). Seizure duration was generally0—30 s.

As seizures evolved, the sequential appearance of behaviors waselt to represent progression to greater severity. Mice that wereffectively treated by SMF or phenytoin exhibited no seizure-relatedehaviors (seizure severity score = 0). Wild running (seizure sever-ty score = 1) evolved into loss of righting accompanied by tonicruncal changes (seizure severity score = 2). Clonus (seizure sever-ty score = 3) occurred during the tonic changes was routinely theost severe stage to which the seizures evolved. Further evolution

o tonic hindlimb extension (seizure severity score = 4) and deathseizure severity score = 5) was encountered infrequently. For theost part, mice either achieved the stage of clonus or had no

eizure-related behavioral changes. Severity was scored based onhe most severe stage manifested during the event.

Seizure latency (time from onset of auditory stimulation toeizure initiation) and seizure stages exhibited by each mouse dur-ng each session of auditory stimulation were tabulated duringhe experiment by an observer. These data were entered into aatabase for subsequent analysis. For purposes of documentationnd later review, experiments performed in the sound chamberere recorded with a digital video camera (Sony Handycam) andere compressed and stored on the lab server.

uditory brainstem responses

he auditory brainstem response (ABR) was recorded from miceestrained in syringe tubes (TR) with (exposed) or without (control)xposure to SMF. Measurements were obtained either after a singlexposure to the SMF with restraint or tube-restraint alone withouteizure threshold testing (SMF, N = 9; TR, N = 8); or, after the fullomplement of pretreatment (control or SMF-exposed) and seizureests (SMF, N = 4,; TR, N = 4). ABR measurements made withoutrior seizure testing were obtained immediately following TR ± SMFxposure for 60 min. ABR measurements made from mice that had

ndergone the complete regimen of seizure threshold testing wereade within 24 h of the third and final seizure test. Mice were anes-

hetized with a mixture of ketamine (100 mg/kg) and medetomidine0.3 mg/kg) and positioned atop a heating pad with a customizedite bar that fixed the head position but left the external audi-

lack

eI8eFolfios

D

SAsC

STlawssdmogtoO

SDebtw

R

A

Afwo(daate(sted

Effects of a static magnetic field on audiogenic seizures in b

tory meati unimpeded. Subdermal electrodes (Grass Instruments,Quincy, MA; or, A—M systems, Carlsborg, WA) were placed at thevertex (ground), mastoids ipsilateral and contralateral to soundsource (active and reference, respectively). The contralateral earwas plugged with sound-attenuating putty.

ABR measurements were obtained within a sound attenuatedchamber (Acoustic Systems; Sequim, WA) lined with Sonex acous-tic foam (Illbruck Inc.; Minneapolis, MN). Auditory stimuli weresynthesized and produced using TDT (Tucker-Davis Technologies;Alachua, FL) hardware and software and a free field tweeter (Vifa;Rødovre, DK), which was observed to have a relatively flat frequencyresponse function between 2 and 50 kHz. Preamplifier filters wereset at 100 Hz high pass and 3000 Hz low pass. The speaker outputwas calibrated prior to ABR testing by positioning a 1/4’’ micro-phone (Bruel & Kjær; Naerum, DK connected to a Pulse 11 soundmeasurement system (Bruel & Kjær; Naerum, DK) at the positionthat would be occupied by the mouse’s ear canal and adjustingthe amplitude of the stimulus waveform such that each stimulustype was presented at a maximum of 80 dB SPL at a distance of20 cm. ABR measurements were made in response to tone pips (1ms rise/fall; 3 ms duration; 8, 16, and 32 kHz) and clicks (50 �smonopolar square wave). Each stimulus was presented 1000× perintensity at rate of 27.7 s−1. Stimuli were initially presented at 80 dBSPL and then reduced in 10 dB increments until an evoked responsewas no longer observed. A final measurement was then made atthe midpoint between the lowest sound level at which a responsewas observe and the highest sound level at which a response wasobserved. ABR threshold was estimated subjectively as the lowestsound level at which the evoked response pattern was preserved.ABR threshold estimates were made by an observer blind to theexperimental history of the mouse.

ABR measurements were recorded to determine the extent towhich the elevation in stimulus intensity needed to produce aseizure in SMF-exposed mice could be explained by a simple eleva-tion in hearing threshold. To address this possibility, ABR recordingswere obtained from exposed mice (TR and exposed to SMF) and con-trol mice (TR, no SMF exposure) before and after the full regimen ofseizure testing. If the magnetic field alone was capable of produc-ing a lasting difference in hearing threshold, we hypothesized thatSMF exposure would result in differences in ABR threshold for oneor more different types of auditory stimuli prior to seizure testing.If the higher sound levels needed to elicit a seizure in SMF-exposedmice were either attributable to — or caused by — an elevationin hearing threshold, we hypothesized that ABR thresholds in SMF-exposed mice would be higher than those observed in control miceacross one or more stimulus types.

Electroencephalography

Mice were placed in a restrainer to record the electroencephalo-gram (EEG). The restrainer consisted of a syringe tube cut in halfalong the long axis except for a 3 cm band extending from a yoke toimmobilize the head to the mid-trunk of the mouse. The yoke hada concave cutout and was inserted in a slot along the leading edgeof the band (corresponding in position to the base of the skull) andheld in place with a rubber band running around the bottom of therestrainer. Holes were place in the bottom of the restrainer to allowthe legs to dangle and facilitate observation of movements. With amouse in place, the entire assembly was lowered into the chamberfor auditory stimulation. Sound intensity was increased graduallyuntil a seizure occurred.

Two electrode placements were used for EEG recording. In each

configuration, an electrode coated with conductive paste was tapedto the tail to serve as ground. In the first configuration, electrodeswith flat plates (1 cm in diameter; Harvard Apparatus, Inc.; Hol-liston, MA) were coated with electrode paste and attached to theears. Leads from each ear with reference to a vertex subdermal

stafc

Swiss mice 123

lectrode were connected to a Grass (Quincy, MA) model IGMEB-NT36 minielectrode board attached to a Grass (Quincy, MA) model-10D EEG machine. In the second configuration, paired subdermallectrodes (model 8227410; Medtronic Xomed, Inc., Jacksonville,L) were positioned on each side of the head with one electrodever the frontal region and the other over the occiput. A pair ofeads on one side of the head served as the dual inputs to an ampli-er. Inputs to different amplifiers were selected to allow displayf recordings from the left and right side of the head at differentensitivities.

ata analysis and statistics

eizure data was analyzed for latency, severity, and gender effects.ll data were entered into a database. Graphical and tabular analy-is were performed with Mathematica 5.2 (Wolfram Research, Inc.;hampaign, IL).

eizure latencyhe Kolmogorov—Smirnov test was used to compare differences in

atency between treatment groups. This statistical method evalu-tes the maximal vertical distance between two distributions (D)hich in and of itself is a measure of effect size. The data are pre-

ented as cumulative distributions of the number of mice havingeizures (ordinate) at different latencies (abscissa). Animals thatid not have seizures during stimulation were included at the maxi-um end of the latency scale (abscissa) as a conservative estimate

f changes in latency and to account for all mice in each treatmentroup. Data are plotted separately for each of the three stimula-ion sessions because of changes in the distributions (latency, sloper number of mice without seizures) from one session to the next.nly those mice that died were eliminated from the analysis.

eizure severityistribution of seizure stages in different treatment groups (groupsxposed to magnetic fields and concomitant controls) were plottedased on the most severe seizure stage manifested in a session. Sta-istical significance of differences between seizure severity scoresas tested by contingency table analysis.

esults

uditory brainstem responses (ABR)

BR thresholds measured for all groups of mice and allour auditory stimulus types (click, 8, 16 and 32 kHz noise)ere relatively high (>50 dB SPL), in keeping with previ-us reports of generally poor hearing in black Swiss miceMisawa et al., 2002). Importantly, no main effect for con-ition (SMF-exposed vs. not exposed) was observed acrossll four stimulus types (ANOVA, F = 1.07, p = 0.38). The datare displayed in Fig. 2. The lack of a difference in ABRhreshold was upheld in comparisons made between SMF-xposed and control mice both prior to auditory stimulationANOVA, F = 0.97, p = 0.34) and after one or three auditorytimulation sessions (ANOVA, F = 0.0, p = 1.0). We observedhat ABR thresholds were slightly elevated after the micexperienced three seizures (ANOVA, F = 3.03, p = 0.1) whenata were pooled across both experimental groups and

timulus types. The elevation in hearing threshold was sta-istically significant for 16 kHz (post hoc contrasts, p = 0.04)nd 32 kHz (post hoc contrasts, p = 0.01) tones. These dif-erences are likely explained as a temporary hearing deficitaused by the relatively loud seizure-inducing stimulus.

124 M.J. McLean et al.

Figure 2 Auditory brainstem responses measurements in TR SMF-exposed and control (not exposed) mice. ABRs were obtainedwith click and tone pips (8, 16 and 32 kHz) at varying sound levels. (A and B) ABR waveforms obtained from an SMF (A) and TR mouse(B) in response to 16 kHz tone pips. ABR thresholds were estimated to be 55 dB SPL in both cases. The black downward triangleindicates the estimated arrival time of the sound wave at the external ear. (C) Mean ± S.E.M. threshold measurements for eachs y bart

Iscc

E

Thasw

rw[vids

timulus type in mice exposed to TR (open bars) and SMF (grahreshold testing.

mportantly, the shift in hearing threshold was expressedimilarly in SMF-exposed and control mice, and therefore,annot explain the changes in seizure threshold observedonsequent to SMF exposure during seizure experiments.

ffects of audiogenic seizures on EEG

he sound intensity required to trigger seizures in mice thatad been restrained for EEG recordings varied between 82nd 102 dBA. EEG recordings associated with sound-inducedeizures were obtained in 11 SMF-exposed (four [36%] diedith seizures elicited by sound at 82, 96, 102 and 102 dBA,

imnis

s) either before (un-hatched) or after (cross-hatched) seizure

espectively) and 24 unexposed control mice (six [25%] diedith seizures elicited by sound intensities of 96 [N = 3] or 102

N = 4] dBA). Seizure-related changes were reversible in sur-iving mice. The EEG recordings became flat in all instancesn which the mice died during seizures. Spikes were notetected unequivocally in the 35 mice subjected to auditorytimulation.

Representative examples of EEG recordings are shown

n Fig. 3. Fig. 3A shows a seizure triggered in a controlouse by auditory stimulation with 96 dBA white noise. Run-

ing movements, but not seizures, occurred at lower soundntensities in this mouse. Within seconds after rasing theound intensity to 96 dBA, clonic leg movements commenced

Effects of a static magnetic field on audiogenic seizures in black Swiss mice 125

Figure 3 Examples of EEG changes in Black Swiss mice in response to auditory stimulation. See text for additional details. (A) EEGof a control mouse that had a seizure in response to stimulation with white noise at a sound intensity of 96 dBA. Running movements,but not seizures, occurred at lower sound intensities in this mouse. Note reduction from baseline amplitude and slow wave activityduring the seizure. Death occurred after the sound was turned off and the EEG became flat. (B) Recordings from a mouse exposedto SMF for 1 h before sound stimulation. A seizure was triggered by white noise at 87 dBA. Note slowing and decreased amplitude of

e an

idiw

ttsiTdrwt

(S(ast(

dcr

the traces during the seizure and recovery to baseline amplitudto the lower right of each set of traces.

(note movement artifact obscures EEG during this phase).This was followed by tonic flexion of the trunk and tonichindlimb extension (THE) associated with slow activity andreduced amplitude on the EEG. Shortly after the sound wasturned off, the mouse stopped breathing, became tonelessand died. This was associated with a flat (except for EKGartifact) EEG trace.

Fig. 3B shows EEG recordings of a seizure produced by87 dBA white noise in a mouse that had been exposed to SMFfor 1 h prior to sound stimulation. A startle response (syn-chronous bilateral jerk of the body and limbs) followed byrunning movements of the limbs occurred at 82 dBA. Within2 s of stimulation at 87 dBA, synchronous clonic hindlimbjerks (traces obscured by movement artifact) and urinaryincontinence occurred. This was followed by tonic stiffen-ing of the trunk (without flexion or extension of the head)and tail. During this phase, there was slow wave activ-ity, occasional extensor jerks of the head and reductionof EEG amplitude compared to pre-seizure activity. After∼17 s of hypertonus without limb movement, motor activityresumed abruptly (escape behavior) and the EEG returned topre-seizure fast activity of the same amplitude as the pre-seizure baseline. These changes indicated the end of the∼30 s seizure.

Effect of pretreatment with SMF on seizure latencyand behaviors

Analyzing the data with respect to gender for an aggregateof all similarly exposed mice (n = 535) and their correspond-

rwwbS

d frequency at the end of the seizure. Calibrations are shown

ng controls (n = 469) in our database showed no significantifference in the response between male and female micen either group (p > 0.25). Data from male and female miceere combined to produce the graphs in Figs. 4—6.

In three of three auditory stimulation sessions, latencyo seizure initiation was prolonged significantly in mice pre-reated with the SMF compared to unexposed controls whenound stimulation was delivered by increments at shortntervals (protocol #1; Fig. 4, Experiments 1—3; Table 1A).he distribution of seizure stages and seizure severity scoresid not differ significantly between groups (Fig. 4; loweright panel). These findings suggest that seizure thresholdas elevated by pretreatment with SMF under conditions

hat produced little alteration of seizure severity.When 10 min steps were used for auditory stimulation

protocol #2), seizure latency was markedly prolonged byMF pretreatment compared to controls in all three sessionsFig. 5, Experiments 1—3; Table 1B). Under these conditions,

significant reduction of CLO (significantly lower seizureeverity score) was observed in SMF-exposed mice comparedo unexposed controls during the second and third sessionFig. 5).

A buildup in preconvulsive behavior was particularly evi-ent during stimulation with protocol #2. The buildup washaracterized by agitation with excessive grooming inter-upted by short periods of running in fits and starts until wild

unning led to the convulsive seizure manifestations (LORith tonic truncal postures, CLO). Such behaviors appearedithin the first few minutes of stimulation in control miceut were postponed or absent in mice pretreated with theMF.

126 M.J. McLean et al.

Figure 4 Effects of SMF on seizure latency and severity. Auditory stimulation by protocol #1. Cumulative distributions of seizurelatency from the first through third separate experiments with the same five batches of mice. Mice that did not have seizuresduring the standard stimulation period (10 min) were added at the extreme right of the distribution to estimate changes in latencyconservatively and account for all mice in each treatment group. In all three experiments, latencies were significantly longer inSMF-exposed mice (stars) than in unexposed controls (diamonds; see Table 1A for p values). Lower right panel: Distribution of seizurestages by experiment (EXP1—3; C = unexposed controls; E = SMF-exposed) according to gray scale at right to assess seizure severity.Number of mice per group shown in parentheses above respective columns. The differences in seizure severity between C and Ewere not significant (n.s. at top of each pair of columns) in experiments 1—3 by contingency table analysis. Distribution of seizuresH

FlopCip

everity categorized by most severe stage achieved: None; WR = Windlimb Extension; DEA = Death.

igure 5 Effects of SMF on seizure latency and severity. Auditoryatency from the first through third separate experiments with thene batch in experiment 3. In all three experiments, latencies wereosed controls (diamonds; see Table 1B for p values). Lower right= unexposed controls; E = SMF-exposed) according to gray scale at r

n parentheses above respective columns. SMF-exposed mice had sigair of columns) in experiments 1—3 by contingency table analysis.

ild Running; LOR = Loss of Righting; CLO = Clonus; THE = Tonic

stimulation by protocol #2. Cumulative distributions of seizuresame two batches of mice in the first two experiments andsignificantly longer in SMF-exposed mice (stars) than in unex-

panel: Distribution of seizure stages by experiment (EXP1—3;ight to assess seizure severity. Number of mice per group shownnificantly reduced seizure severity (p values given above eachSeizure severity scale to right as in Fig. 4.

Effects of a static magnetic field on audiogenic seizures in black Swiss mice 127

Figure 6 Effects of phenytoin with and without SMF exposure on seizure latency and severity. Auditory stimulation by protocol #1.Cumulative distributions of seizure latency from the first through third separate experiments with the same three batches of mice.In all three experiments, latencies were significantly longer in SMF-exposed mice (stars) than in unexposed controls (diamonds; seeTable 1C for p values). Lower right panel: Distribution of seizure stages by experiment (EXP1—3; C = unexposed controls; E = SMF-

y. Nuverit

D

Asmihttsploiwoadas

fsmsatrs

exposed) according to gray scale at right to assess seizure severitcolumns. SMF-exposed mice had significantly lower seizure secolumns). Seizure severity scale to right as in Fig. 4.

Phenytoin dose—response characteristic

Groups of mice were pretreated with different doses ofphenytoin to identify a dose with minor effects on seizurelatency and seizure severity score for studies of interac-tions with SMF. The principal criterion for selection wasthat effects of the phenytoin dose should be small enoughto afford detection of additive or potentiative interactionsbetween phenytoin and SMF. A range of phenytoin doses,from 0.5 to 40 mg/kg, was studied (data not shown). Seizurelatencies after pretreatment with phenytoin doses of 0.5,1 and 2 mg/kg did not differ from those seen in control(untreated) groups. Seizure latency was prolonged in micewith seizures and most mice had no seizures after adminis-tration of 7.5 or 10 mg/kg of phenytoin. No mice had seizuresin groups pretreated with 20 or 40 mg/kg of phenytoin. Aphenytoin dose of 5 mg/kg had small effects: latency wasprolonged, but all mice had seizures at some sound level.This dose was selected for experiments involving pretreat-ment with phenytoin and SMF, alone or in combination.

Effects of pretreatment with phenytoin and SMF,alone and in combination, on seizure threshold

Three batches of mice were pretreated with 5 mg/kg pheny-toin i.p. (2 h before auditory stimulation) alone or the samedose of phenytoin in combination with SMF (1 h exposure)

prior to auditory stimulation. Combination pretreatmentsignificantly increased seizure latency compared to pheny-toin alone in all three auditory stimulation sessions (Fig. 6;Experiments 1—3; Table 1C) and reduced seizure severity(Fig. 6; lower right panel).

PstT

mber of mice per group shown in parentheses above respectivey in experiments 2 and 3 (p values given above each pair of

iscussion

udiogenic seizures were triggered during a period of seizureusceptibility between postnatal days 21—30 in black Swissice by 72—102 dBA white noise. Seizure manifestations

ncluded episodes of wild running, loss of righting withypertonus of the trunk in flexion or extension and clonus ofhe limbs. The mice were not responsive to tactile stimula-ion during these phases. Most mice tolerated at least threeound-induced seizures triggered every other day during theeriod of susceptibility. The seizure behaviors triggered byoud noise during the period of seizure susceptibility did notccur prior to about day 18 or after day 32 in our exper-ments (McLean et al., unpublished). At these ages, thereere startle responses in which the mice hopped, had oner a few whole body jerk or ran in brief spurts. However, atges outside the seizure susceptibility period, the behaviorsid not progress to loss of righting, tonic or clonic activitynd the mice did not lose responsiveness to environmentaltimuli.

Electroencephalographic (EEG) changes were recordedrom restrained mice with extracranial electrodes duringound-induced seizures. The mice were restrained to mini-ize motion artifacts. This could have resulted in elevated

eizure thresholds. Restraint during recordings was associ-ted with higher thresholds for AGS than was the case inhe studies of latency described here. Restraint has beeneported to have anticonvulsant effects in some mouseeizure models under a variety of conditions (e.g., see

ericic et al., 2000; Tchekalarova and Georiev, 2006). Highound intensities resulted in more severe seizures, e.g.,he appearance of tonic hindlimb extension, in some mice.he EEGs showed ictal slowing from faster baseline activ-

1

iatopttiatmieewa(cprschpd

ibcgiaiaicsibsceim(wttttast2tallmowb

wtbttdt

drmbewiiSi

stsscrhswptetpitomm

camlteo1ltom

Seeaw

28

ty and reduction of amplitude with recovery to baselinectivity after seizures (Fig. 3B). In mice that died duringonic hindlimb extension, complete suppression of the EEGccurred (Fig. 3A). In mice given PTZ, spikes were recordedrior to myoclonic jerks (data not shown). This indicateshat spikes can be recorded with the extracranial electrodeechnique used here. EEG spikes could have been obscuredn the present study by motion artifact during the runningnd clonic phases of seizures. Alternatively, activation ofhe brainstem circuitry resulting in audiogenic seizures inice may not have resulted in cortical spike-wave activ-

ty that could be detected with extracranial electrodes orven with intracranial electrodes (Vlgr1 knockout mice; Yagit al., 2005). Spikes and spike-wave activity were recordedith intracranial and depth electrodes first in the brainstemnd, later, in cortical loci in genetically epilepsy-prone ratsGEPR; Faingold, 1999; Moraes et al., 2005). The hierarchi-al progression of EEG changes and behaviors resulted inostural, tonic and clonic changes (see Faingold, 1999, foreview) including those seen in the black Swiss mice. Thetereotypical convulsive behaviors are similar in AGS sus-eptible mouse and rat strains. In addition, GEPR typicallyad tonic hindlimb extension also. Differences in electrodelacement, animal strain and animal size may contribute toifferences in the appearance of ictal EEG changes.

Latency to sound-induced behavioral changes was signif-cantly prolonged in mice pretreated with the SMF producedy permanent magnets compared to unexposed controls. Toontrol for and minimize differences between treatmentroups due to restraint stress, all mice were pretreatedn syringe tubes with or without attached magnets. Also,ll mice were removed from the tubes and from proxim-ty to the magnets before auditory stimulation. Because thebsence of a seizure at any given sound level could be due tonability to express seizures, e.g., because of developmentalhanges in the nervous system or refractoriness after a prioreizure episode, auditory stimulus intensity was increasedn an attempt to elicit seizures in as many mice as possi-le. At least 80% of unexposed control mice in every batchtudied could be induced to have seizures. Compared tooncomitant unexposed controls, significantly fewer SMF-xposed mice had seizures with maximal sound stimulationn a variety of experimental paradigms. This was true ofice of both genders, as observed previously in DBA/2 mice

McLean et al., 2003). Results from male and female miceere pooled in the data analysis presented here. During

he period of maximum seizure susceptibility (21—30 days),he same batch (same birth date) of mice was tested up tohree times at two to three day intervals to avoid refrac-oriness. In experiments with 10 min sound steps (Fig. 5)nd phenytoin in combination with SMF (Fig. 6), efficacyeemed to increase from one auditory stimulation sessiono the next. The shift in latency occurred with a shift of—13 dBA in auditory stimulation intensity. The decibel sys-em is logarithmic. Therefore, changes of this magnitudere akin to increasing auditory stimulus intensity from theevel of normal conversational speech to shouting to a col-

eague across the room. Changes of this magnitude may oray not be sufficient to produce clinical benefit in terms

f seizure control. However, a valid clinical test of efficacyill require envelopment of the seizure-generating networky effective regions of the SMF. Dosimetry for such studies

ietde

M.J. McLean et al.

ill have to be calculated with sure knowledge of loca-ion of the seizure focus. Also, properties of magnets muste optimized for delivering the necessary field metrics tohe vicinity of the seizure generator. This will require fur-her investigation of the field metrics and, possibly, theesign of magnetic devices specifically for certain seizuresypes.

The black Swiss mice used in the present experimentsid not have age-dependent hearing loss. All mice wereestrained in syringe tubes for pretreatment before deter-ination of hearing threshold. There was no differenceetween hearing thresholds measured in controls and SMF-xposed mice (Fig. 2). Hearing thresholds to some stimuliere elevated after a series of seizures, but the changes

n control mice and SMF-exposed mice did not differ signif-cantly. Thus, changes in seizure latency and severity afterMF exposure were unlikely to be the result of altered hear-ng threshold.

The mice were removed from the SMF before auditorytimulation. This raises questions about how long the protec-ive effect of the field might last. Experiments with 10 minound steps (protocol 2) provide some insight. In the firstession (Fig. 5, Exp1), nearly 60% of SMF-exposed mice,ompared to only 25% of unexposed controls, had not expe-ienced a seizure by 10 min; almost all mice in both groupsad seizures by 20 min. During the second and third ses-ion, 50—60% of SMF-exposed mice had seizures at 30 min,hereas all the unexposed controls had seizures after shorteriods of auditory stimulation. To some extent, it appearshat the persistence of the protective effect depends onxperimental conditions and actually may persist longer inhe course of serial seizures. Additional experiments are inrogress to determine duration of persistence of the SMF-nduced prolongation of latency. In previous experiments,he anticonvulsant effect of SMF produced by a square arrayf electromagnets with alternating polarity persisted forore than an hour in the exposed group as a whole afterice were removed from the field (McLean et al., 2003).During sound stimulation with 1 min steps (protocol #1),

hanges in seizure latency after SMF exposure were notccompanied by significant changes in seizure severity aseasured by incidence of seizure-related behaviors (Fig. 4,

ower right panel). Once the threshold for seizure igni-ion was reached — even though latency was prolonged inxposed mice — similar seizure-related behavioral changesccurred In both control and SMF-exposed mice. Using0 min sound steps, seizure severity was reduced also (Fig. 5,ower right panel). These observations suggest that exposureo the SMF may raise seizure threshold and limit propagationf epileptiform activity variably under different conditions,ainly depending on the nature of the stimulus.It was possible to overwhelm the protective effect of the

MF. This was show here by increasing sound intensity untilach mouse had a seizure whenever possible. In previousxperiments, the anticonvulsant effect of SMF produced bysquare array of electromagnets with alternating polarityas overcome by increasing the dose of NMDA administered

ntracerebroventricularly to ICR/CD-1 outbred mice (McLeant al., unpublished). On the other hand, combined pre-reatment with phenytoin and SMF prolonged latency andecreased seizure severity more than pretreatment withither modality alone.

lack

timpgictiSttate(2

tpSatlspnotAwbmmw

R

B

B

C

C

C

D

Effects of a static magnetic field on audiogenic seizures in b

Pretreatment with phenytoin and SMF markedly reducedseverity of seizures, compared to either SMF or phenytoinpretreatment alone and tube-restrained controls. Based onthe number on mice that did not have a seizure by com-pletion of the protocol (maximum sound intensity 102 dBA),phenytoin appeared to be more effective with each seizure(see Fig. 6A—C). This could be due to phenytoin accumu-lation, refractoriness to seizures, unknown factors or somecombination of factors. There was no change in blood levelsof phenytoin in samples pooled from groups of mice injectedthree time with 5 mg/kg every other day, making phenytoinaccumulation an unlikely explanation (McLean et al., unpub-lished). In preliminary experiments, exposure to the SMFfor 1 h daily for 5 days before auditory stimulation did notincrease the effects of SMF on seizure latency (McLean et al.,unpublished). Thus, there is no evidence to date of a cumu-lative effect of repeated exposures to the magnetic field.Further experiments are necessary to determine the dura-tion of effects of the magnetic field and phenytoin, aloneand in combination.

Combined pretreatment with phenytoin and the SMFmore than doubled the number of mice in each batch thathad no seizures in all three experiments. Very little increasein the number of mice without seizures was observed afterpretreatment with SMF alone. The magnetic field couldincrease anticonvulsant efficacy by (a) producing an effectby an unknown mechanism that combines with the mech-anism(s) of phenytoin action, or (b) by enhancing one ormore cellular mechanisms of action of phenytoin. There aresome reports of effects of SMF on ion channels, but thesehave not been replicated (see McLean et al., 2001). So farwe have observed no change in effects of phenytoin onsodium, low threshold or high threshold calcium currentsmeasured under whole cell patch clamp conditions at roomtemperature or 37 ◦C during or after exposure to SMF pro-duced by the same type of magnets used in the experimentsdescribed here (McLean et al., unpublished). These find-ings make direct interactions between phenytoin and SMF onion channels seem to be unlikely. The current experimentshave not revealed the biophysical transduction mecha-nism(s) or the cellular effector mechanism(s) whereby SMF,alone or in combination with phenytoin, exert the observedeffects.

Potential biophysical transduction mechanisms of SMFhave been proposed (for review see Engstrom, 2004; Johnsenand Lohmann, 2005; Engstrom, 2006), but no investigationsof which we are aware have explained seizure control bySMF in terms of a specific mechanism. Changing the vec-tor of an imposed SMF resulted in changes in the positionof mole rat nests and altered c-Fos expression in the col-liculi (Nemec et al., 2001). This brain region is integrallyinvolved in the seizure pathway of AGS (Faingold, 1999).This coincidence suggests that brainstem neuronal networksrespond to imposed SMF. Efforts in this laboratory havefocused on the potential role of strong gradients withintime-invariant static magnetic fields. Others have consid-ered potential health effects of gradient fields but have not

studied seizures (e.g., see Barnothy, 1964) or have usedtime-varying gradients fields as employed in echo planarmagnetic spectroscopic imaging (Carlezon et al., 2005) ortranscranial magnetic stimulation (see McLean et al., 2001,for review). Unlike time-varying (pulsed) fields used for

D

Swiss mice 129

hese approaches, the permanent magnets used here do notnduce currents. The biophysical mechanism(s) coupling theagnetic field to seizure ignition is(are) not revealed by theresent findings. Though speculative, some insight can beleaned from results obtained during exposure to long, lowntensity sound (protocol #2). Preconvulsive behaviors arelearly present and mice gradually become more agitated inhe buildup to seizure ignition. These preconvulsive behav-ors were delayed or absent in mice pretreated with SMF.uch observations implicate widespread networks as poten-ial targets for SMF. We propose as a working hypothesis thathe threshold for synchronization in networks leading to AGSnd possibly other seizures might be elevated by exposureo SMF. Effects on synchrony might be detectable in studiesmploying a seizure warning system based on chaos theorysee, e.g., Litt and Lehnertz, 2002; Chaovalitwongse et al.,005; Lehnertz et al., 2007).

In summary, anticonvulsant properties of the experimen-al SMF alone or in combination with the well-known AED,henytoin were demonstrated in experiments with blackwiss mice. Many questions remain. Experiments in progressre designed to determine (a) the effect of varying the dura-ion of exposure to the SMF on seizure threshold; (b) howong elevation of seizure threshold persists after SMF expo-ure; (c) differences between exposure to perpendicular andarallel gradients within the SMF; (d) positions of the mag-et around the head or along the body where elevationf seizure threshold occurs; (e) the dose—response rela-ionship between SMF and seizure threshold; and (f) whichEDs interact with the SMF. Hopefully, these investigationsill help to determine the field metric coupling the SMF toiological systems and potential pharmacological effectorechanisms. Ultimately, it is hoped that this model systemay inform a method for studying effects of SMF in patientsith epilepsy.

eferences

arnothy, M.F. (Ed.), 1964. Biological Effects of Magnetic Fields.Plenum Press, New York.

rodie, M.J., Richens, A., Yuen, A.W.C, 1995. Double-blind com-parison of lamotrigine and carbamazepine in newly diagnosedepilepsy. Lancet 345, 476—479.

arlezon, W.A., Rohan, M.L., Mague, S.D., Meloni, E.G., Parsegian,A., Cayetano, K., Tomasciewicz, H.C., Rouse, E.D., Cohen, B.M.,Renshaws, P.F., 2005. Anti-depressant-like effects of cranialstimulation within a low-energy magnetic field in rats. Biol. Psy-chiatry 57, 571—576.

hadwick, D.W., Anhut, H., Greiner, M.J., Alexander, J., Mur-ray, G.H., Garofalo, E.A., Pierce, M.W., 1998. InternationalGabapentin Monotherapy Study Group 945-77 A double-blindtrial of gabapentin monotherapy for newly diagnosed partialseizures. Neurology 51, 1282—1288.

haovalitwongse, W., Iasemidis, L.D., Pardalos, P.M., Carney, P.R.,Shiau, D.-S., Sackellares, J.C., 2005. Performance of a seizurewarning algorithm based on the dynamics of intracranial EEG.Epilepsy Res. 64, 93—113.

ailey, J.W., Reigel, C.E., Mishra, P.K., Jobe, P.C., 1989. Neurobi-ology of seizure predisposition in the genetically epilepsy-prone

rat. Epilepsy Res. 3 (1), 3—17.

ailey, J.W., Yan, Q.S., Adams-Curtis, L.E., Ryu, J.R., Ko, K.H.,Mishra, P.K., Jobe, P.C., 1996. Neurochemical correlates ofantiepileptic drugs in the genetically epilepsyprone rat (GEPR).Life Sci. 58 (4), 259—266.

1

D

E

E

F

F

F

F

F

G

G

H

H

J

J

J

J

K

K

K

K

K

K

L

L

L

L

L

M

M

M

M

M

M

M

M

N

P

30

e Herdt, V., Boon, P., Ceulemans, B., Hauman, H., Lagae, L.,Legros, B., Sadzot, B., Van Bogaert, P., van Rijckevorsel, K., Ver-helst, H., Vonck, K., 2007. Vagus nerve stimulation for refractoryepilepsy: a Belgian multicenter study. Eur. J. Paediat. Neurol. 11,261—269.

ngstrom, S., 2004. Physical mechanisms of non-thermal extremelylow frequency magnetic field effects. Radio Sci. Bull. 311,96—106.

ngstrom, S., 2006. Magnetic field effects on free radical reactionsin biology. In: Barnes, F.S., Greenebaum, B. (Eds.), Bioengineer-ing and Medical Aspects of Electromagnetic Fields. CRC Press,Boca Raton, FL and New York, NY, pp. 157—168.

aingold, C.L., Walsh, E.J., Maxwell, J.K., Randall, M.E., 1990.Audiogenic seizure severity and hearing deficits in the geneti-cally epilepsy-prone rat. Exp. Neurol. 108, 55—60.

aingold, C.L., 1999. Neuronal networks in the genetically epilepsy-prone rat. In Jasper’s Basic Mechanisms of the Epilepsies, thirdedition: Advances in neurology, 79, pp. 311—321.

rench, J., 2006. Refractory epilepsy: one size does not fit all.Epilepsy Curr. 6 (6), 177—180.

reeman, J.M., Vining, E.P.G., 1999. Seizures decrease rapidly afterfasting: preliminary studies of the ketogenic diet. Arch. Pediatr.Adolesc. Med. 153, 946—949.

rings, H., Frings, M., 1953. Development of strains of albino micewith predictable susceptibilities to audiogenic seizures. Sci.,New Ser. 117 (3037), 283—284.

asior, M., Rogawski, M.A., Hartman, A.L., 2006. Neuroprotectiveand disease-modifying effects of the ketogenic diet. Behav. Phar-macol. 17 (5—6), 431—439.

uekht, A.B., Mitrokhina, T.V., Lebedeva, A.V., Dzugaeva, F.K.,Milchakova, L.E., Lokshina, O.B., Feygina, A.A., Gusev, E.I.,2007. Factors influencing on quality of life in people withepilepsy. Seizure 16 (2), 128—133.

elmers, S.L., Griesemer, D.A., Dean, J.C., Sanchez, J.D., Labar,D., Murphy, J.V., Bettis, D., Park, Y.D., Shuman, R.M., MorrisIII, G.L., 2003. Observations on the use of vagus nerve stimula-tion earlier in the course of pharmacoresistant epilepsy: patientswith seizures for 6 years or less. Neurologist 9, 160—164.

uffman, J., Kossoff, E.H., 2006. State of the ketogenic diet(s) inepilepsy. Curr. Neurol. Neurosci. Rep. 6 (4), 332—340.

obe, P.C., Mishra, P.K., Browning, R.A., Wang, C., Adams-Curtis,L.E., Ko, K.H., Dailey, J.W., 1994. Noradrenergic abnormalitiesin the genetically epilepsy-prone rat. Brain Res. Bull. 35 (5—6),493—504.

ohnsen, S., Lohmann, K.J., 2005. The physics and neurobiology ofmagnetoreception. Nat. Rev. Neurosci. 6, 703—712.

ohnson, K.R., Zheng, Q.Y., Weston, M.D., Ptacek, L.J., Noben-Trauth, K., 2005. The Mass1frings mutation underlies early onsethearing impairment in BUB/BnJ mice, a model for the auditorypathology of Usher syndrome IIC. Genomics I 85, 582—590.

ohnson, K.R., Zheng, Q.Y., Noben-Trauth, K., 2006. Strain back-ground effects and genetic modifiers of hearing in mice. BrainRes. 1091, 79—88.

im, Y.D., Heo, K., Park, S.C., Huh, K., Chang, J.W., Choi, J.U.,Chung, S.S., Lee, B.I., 2005. Antiepileptic drug withdrawal aftersuccessful surgery for intractable temporal lob epilepsy. Epilep-sia 46 (2), 251—257.

lein, B.D., Fu, Y.-H., Ptacek, L.J., White, H.S., 2005. Auditorydeficits associated with the Frings Mgr1 (Mass1) mutation inmice. Dev. Neurosci. 27, 321—332.

ossoff, E.H., 2004. More fat and fewer seizures: dietary therapiesfor epilepsy. Lancet Neurol. 3, 415—420.

rall, R.L., Penry, J.K., Kupferberg, H.J., Swinyard, E.A., 1978.

Antiepileptic drug development: I. History and a program forprogress. Epilepsia 19 (4), 393—408.

upferberg, H.J., 1989. Antiepileptic drug development program:a cooperative effort of government and industry. Epilepsia 30(Suppl. 1), S51—56; discussion S64—68.

P

P

M.J. McLean et al.

wan, P., Brodie, M.J., 2000. Early identification of refractoryepilepsy. N. Eng. J. Med. 342, 314—319.

ehnertz, K., Mormann, F., Osterhage, H., Muller, A., Prusseit, J.,Chernihovskyi, A., Staniek, M., Krug, D., Bialonski, S., Elger,C.E., 2007. J. Clin. Neurophysiol. 24 (2), 147—153.

hatoo, S.D., Wong, I.C.K., Polizzi, G., Sander, J.W.A.S.,2000. Long-term retention rates of lamotrigine, gabapentin,and topiramate in chronic epilepsy. Epilepsia 41 (12),1592—1596.

itt, B., Lehnertz, K., 2002. Seizure prediction and the preseizureperiod. Curr. Opin. Neurol. 15, 173—177.

oscher, W., 1984. Genetic animal models of epilepsy as aunique resource for the evaluation of anticonvulsant drugs.A review. Methods Find. Exp. Clin. Pharmacol. 6 (9),531—547.

oscher, W., Leppik, I.E., 2002. Critical re-evaluation of previouspreclinical strategies for the discovery and the development ofnew antiepileptic drugs. Epilepsy Res. 50, 17—20.

attson, R.H., Cramer, J.A., Collins, J.F., Smith, D.B., Delgado-Escueta, A.V., Browne, T.R., Williamson, P.D., Treiman, D.M.,McNamara, J.O., McCutchen, C.B., Homan, R.W., Crill, W.E.,Lubozynski, M.F., Rosenthal, N.P., Mayersdorf, A., 1985. Com-parison of carbamazepine, Phenobarbital, phenytoin, andprimidone in partial and secondarily generalized tonic-clonicseizures. N. Eng. J. Med. 313, 145—151.

attson, R.H., Cramer, J.A., Collins, J.F., 1992. A comparisonof valproate with carbamazepine for the treatment of com-plex partial seizures and secondaility generalized tonic-clonicseizures in adults. The Department of Veterans Affairs EpilepsyCooperative Study No. 264 Group. N. Engl. J. Med. 327 (11),765—771.

cLean, M.J., Engstrom, S., Holcomb, R.R., 2001. Magnetic fieldtherapy for epilepsy. Brain Behav. 2, s81—s87.

cLean, M.J., Engstrom, S., Holcomb, R.R., Sanchez, D., 2003. Astatic magnetic field modulates severity of audiogenic seizuresand anticonvulsant effects of phenytoin. Epilepsy Res. 55,105—116.

cMillan, D.R., White, P.C., 2004. Loss of the transmembraneand cytoplasmic domains of the very large G-protein-coupledreceptor-1 (VLGR1 or Mass1) causes audiogenic seizures in mice.Mol. Cell. Neurosci. 26, 322—329.

isawa, H., Kawasaki, Y., Melton, J., Sweeney, N., Jo, K., Nicoll,R.A., Bredt, D.S., 2001. Contrasting localizations of MALS/LIN-7PDZ proteins in brain and molecular compensation in knockoutmice. J. Biol. Chem. 276, 9264—9272.

isawa, H., Sherr, E.H., Lee, D.J., Chetkovich, D.M., Tan, A.,Schreiner, C.E., Bredt, D.S., 2002. Identification of a monogeniclocus (jams1) causing juvenile audiogenic seizures in mice. J.Neurosci. 22 (23), 10088—10093.

oraes, M.F.D., CHavali, M., Mishra, P.K., Jobe, P.C., Garcia-Cairasco, N., 2005. A comprehensive electrographic andbehavioral analysis of generalized tonic-clonic seizures of GEPR-9s. Brain Res. 1033, 1—12.

emec, P., Altmann, J., Marhold, S., Burda, H., Oelschlager,2001. Neuroanatomy of magneoreception: the superior collicu-lus involve in magnetic orientation in a mammal. Science 294,366—368.

ennacchio, L.A., Lehesjoki, A.E., Stone, N.E., Willour, V.L., Vir-taneva, K., Miao, J., D’Amato, E., Ramirez, L., Faham, M.,Koskiniemi, M., Warrington, J.A., Norio, R., de la Chapelle, A.,Cox, D.R., Myers, R.M., 1996. Mutations in the gene encodingcystatin B in progressive myoclonus epilepsy (EPMI). Science 271,1731—1734.

ericic, D., Svob, D., Jazvinscak, M., Mirkovic, K., 2000. Anticonvul-sant effect of swim stress in mice. Pharmacol. Biochem. Behav.66, 879—886.

iazzini, A, Ramaglia, G, Turner, K, Chifari, R, Kiky, EE, Canger, R,Canevini, M.P. Coping strategies in epilepsy: 50 drug-resistant

lack

S

S

S

T

T

W

W

Y

Z

Effects of a static magnetic field on audiogenic seizures in b

and 50 seizure-free patients. Seizure, 2007, Jan 5 [Epub aheadof print].

Puttagunta, R., Gordon, L.A., Meyer, G.E., Kapfhamer, D.,Lamerdin, J.E., Kantheti, P., Portman, K.M., Chung, W.K.,Jenne, D.E., Olsen, A.S., Burmeister, M., 2000. Comparativemaps of human 19p13. 3 and mouse chromosome 10 allow identi-fication of sequences at evolutionary breakpoints. Genome Res.10, 1369—1680.

Ross, K.C., Coleman, J.R., 2000. Developmental genetic audiogenicseizure models: behavior and biological substrates. Neurosci.Behav. Rev. 24, 639—653.

Rowan, A.J., Ramsay, R.E., Collins, J.F., Pryor, F., Boardman, K.D.,Uthman, B.M., Spitz, M., Frederick, T., Towne, A., Carter, G.S.,Marks, W., Felicetta, J., Tomyanovich, M.L., 2005. VA Coopera-tive Study 428 Group. New onset geriatric epilepsy A randomizedstudy of gabapentin, lamotrigine, and carbamazepine. Neurol-ogy 64, 1868—1873.

Schmidt, D., 2002. The clinical impact of new antiepileptic drugsafter a decade of use in epilepsy. Epilepsy Res., 1—12.

Schmidt, D., Loscher, W., 2003. How effective is surgery to cureseizures in drug-resistant temporal lobe epilepsy? Epilepsy Res.56, 85—91.

Schmidt, D., Baumgartner, C., Loscher, W., 2004. Seizure recurrenceafter planned discontinuation of antiepileptic drugs in seizure-free patients after epilepsy surgery: A review of current clinicalexperience. Epilepsia 45 (2), 179—186.

Skradski, S.L., Clark, A.M., Jiang, H., White, H.S., Fu, Y.H., Ptacek,L.J., 2001. A novel gene causing a Mendelian audiogenic mouseepilepsy. Neuron 31, 537—544.

Skradski, S.L., White, H.S., Ptacek, L.J., 1998. Genetic mapping ofa locus (mass1) causing audiogenic seizures in mice. Genomics49, 188—192.

Smith, D.B., Mattson, R.H., Cramer, J.A., Collins, J.F., Novelly, R.A.,Craft, B., 1987. Results of a nationwide Veterans AdministrationCooperative Study comparing the efficacy and toxicity of carba-mazepine, Phenobarbital, phenytoin and primidone. Epilepsia 28(Suppl. 3), S50—S58.

Z

Swiss mice 131

mith, M., Wilcox, K.S., White, H.S., 2007. Discovery of antiepilep-tic drugs. Neurotherapeutics 4 (1), 12—17.

pencer, S.S., 2002. When should temporal-lobe epilepsy be treatedsurgically? Lancet Neurol. 1, 375—382l.

tables, J.P., Bertram, E.H., White, H.S., Coulter, D.A., Dichter,M.A., Jacobs, M.P., Loscher, W., Lowenstein, D.H., Moshe, S.L.,Noebels, J.L., Davis, M., 2002. Models for epilepsy and epilep-togenisis: report from the NIH Workshop, Bethesda, Maryland.Epilepsia 43 (11), 1410—1420.

chekalarova, J., Georiev, V., 2006. Effect of acute versus chronictheophylline administration on acute restraint stress-inducedincrease of pentylenetetrazol seizure threshold in mice. BrainRes. Bull. 68, 464—468.l.

heodore, W.H., Spencer, S.S., Wiebe, S., Langfitt, J.T., Ali, A.,Shafer, P.O., Berg, A.T., Vickrey, B.G., 2006. ILAE Report:Epilepsy in North America: A report prepared under theauspices of the Global Campaign Against Epilepsy, the Inter-national Bureau for Epilepsy, the International League AgainstEpilepsy, and the World Health Organization. Epilepsia 47 (10),1700—1722.

hite, H.S., Wolf, H.H., Woodhead, J.H., Kupferberg, H.J., 1998.The National Institutes of Health Anticonvulsant Drug Devel-opment Program: screening for efficacy. Adv. Neurol. 76,29—39.

hite, H.S., Woodhead, J.H., Wilcox, K.S., Stables, J.P., Kupfer-berg, H.J., Wolf HH, 2002. General principles. Discovery andpreclinical development of antiepileptic drugs. In: AntiepilepticDrugs, Fifth Ed, Chapter 3.

agi, H., Takamura, Y., Yoneda, T., Konno, D., Akagi, Y., Yoshida,K., Sato, M., 2005. Vlgr1 knockout mice show audiogenic seizuresusceptibility. J. Neurochem. 92, 191—202.

heng, Q.Y., Johnson, K.R., Erway, L.C., 1999. Assessment of hear-

ing in 80 inbred strains of mice by ABR threshold analyses.Hearing Res. 130, 94—107.

hou, X., Jen, P.H.-S., Seburn, K.L., Frankel, W.N., Zheng, Q.Y.,2006. Auditory brainstem responses in 10 inbred strains of mice.Brain Res. 1091, 16—26.