- Email: [email protected]
Seizure prophylaxis in an animal model of epilepsy by dietary fluoxetine supplementation
Epilepsy Research (2007) 74, 19—27
journal homepage: www.elsevier.com/locate/epilepsyres
Seizure prophylaxis in an animal model of epilepsy by dietary fluoxetine supplementation Alyssa Richman, Stephen C. Heinrichs ∗ Department of Psychology, 140 Commonwealth Avenue, Chestnut Hill, MA 02467, United States Received 13 August 2006; received in revised form 20 November 2006; accepted 30 November 2006 Available online 9 January 2007
KEYWORDS Antidepressant; Tail suspension handling; Seizure; Locomotor activity; Serotonin
Summary Clinical and animal model evidence suggests that selective serotonin reuptake inhibitors (SSRIs) act as anticonvulsants. The present studies tested the possibility that the El mouse model of genetically predisposed/handling-triggered epilepsy would exhibit fewer seizures following SSRI treatment via dietary fluoxetine adulteration. In particular, potential bioenergetic and neural mechanisms for anticonvulsant efficacy of fluoxetine were explored using food intake/body weight monitoring and quantification of brain serotonin transporter protein. El mice consuming a chow diet ad libitum or yoked in quantity to fluoxetine diet intake exhibited seizure incidence of 40% in response to tail-suspension handling, whereas seizures were abolished (0%) among El mice consuming a fluoxetine-adultered diet over 7 days. A 3 day period of fluoxetine administration was insufficient to exert anticonvulsant efficacy and all treatment groups exhibited the same circadian locomotor activity patterns at the time of seizure susceptibility testing. Bioenergetic factors could not account for the anticonvulsant efficacy of fluoxetine since yoked diet controls with matched food intake, body weight change and blood glucose levels exhibited the same 40% seizure incidence as ad libitum chow controls. Importantly, the 7 day period of dietary fluoxetine exposure was effective in selectively reducing cell density in the parietal cortex and increasing serotonin transporter protein content in the nucleus accumbens. Taken together, these results suggest that dietary fluoxetine supplementation abolishes handling-induced seizure susceptibility in El mice via a neural remodeling mechanism independent of energy balance. © 2006 Elsevier B.V. All rights reserved.
Introduction
∗ Corresponding author at: VA Medical Center, Research 151Neuropharmacology, 150 South Huntington Avenue, Boston, MA 02130, United States. Tel.: +1 857 364 5617; fax: +1 617 278 4540. E-mail address: [email protected] (S.C. Heinrichs).
While selective serotonin reuptake inhibitors (SSRIs) are a first line pharmacotherapy used to treat clinical depression (Dailey and Naritoku, 1996), fluoxetine and other antidepressants also exert anticonvulsant effects (Hernandez et al., 2002; Ferrero et al., 2005; Kecskemeti et al., 2005; Pericic et al., 2005; Tupal and Faingold, 2006). Serotonergic signaling appears to have a direct role in modulating
0920-1211/$ — see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.eplepsyres.2006.11.007
20 seizure susceptibility as both 5-HT2C and 5-HT4 knockout mice exhibit enhanced sensitivity to pentylenetetrazoleinduced convulsions (Tecott et al., 1995; Compan et al., 2004). Similarly, resting state levels of serotonin (5-HT) are lower in seizure prone mice, as compared to 11 inbred strains (King and LaMotte, 1989). Consistent with a causal role for 5-HT deficiency in seizure incidence, administration of 5-hydroxytrytophan (a 5-HT precursor) with MK-486 (a peripheral decarboxylation inhibitor which limits 5-HT synthesis to the brain) increases brain 5-HT levels in seizure prone mice, and significantly decreases the occurrence of seizures (Hiramatsu, 1981). At the receptor level, the 5-HT1A receptor agonist, WAY100635, modifies epileptic activity depending on the type of seizure (Lopez-Meraz et al., 2005). Chronic treatment with fluoxetine decreases epileptic threshold in a manner correlated with levels of hippocampal glutamate release (Ferrero et al., 2005) and fluoxetine reduces spontaneous seizures in a model of temporal lobe epilepsy (Hernandez et al., 2002). This evidence suggests that serotonergic neurotransmission inhibits neuronal network excitability (Dailey et al., 1992) although consensus has been slow to develop on this point due to confusion between manifestations of illness versus treatment effects (Jobe and Browning, 2005). One aim of the present studies was to provide a better understanding of the central mechanisms of action for the anticonvulsant effects of fluoxetine. The following experiments employed the ‘epilepsy’ (El) mouse model to monitor seizure susceptibility after subchronic oral administration of the antidepressant fluoxetine. The El progenitor mouse resulted from a spontaneous mutation which was developed subsequently into the El strain through selective breeding to maintain and amplify the seizure prone phenotype (Murashima et al., 2002). Seizures originate in the parietal cortex of El mice, and subsequently spread to other brain regions such as the hippocampus (Ishida et al., 1993). The seizure locus may shift to the hippocampus in seizure-experienced El mice (Mutoh et al., 1993) although this migration is a dynamic process dependent upon age and handling history (Leussis and Heinrichs, in press). Susceptibility to seizures in El mice increases with age, and with repeated sensory stimulations involving tail-suspension, or ‘tossing-up’ (Suzuki and Nakamoto, 1982). The El mouse thus constitutes an animal model for idiopathic complex partial seizures (Seyfried and Glaser, 1985; Murashima et al., 2004), for temporal lobe epilepsy (Suzuki et al., 1983) and for simple reflex epilepsy (Sarkisian, 2001). El mice are genetically predisposed to seize although episodes are not intractable since pharmacological and dietary studies have demonstrated the adult plasticity of seizure susceptibility (Heinrichs and Seyfried, 2006). For example, caffeine augments seizures in El mice by altering the pharmacokinetics of a concurrently administered anticonvulsant drug (Hashiguchi et al., 2001). Similarly, high fat, low carbohydrate ketogenic diet exposure exerts anticonvulsant effects in El mice (Rho et al., 1999; Seyfried et al., 2004). These findings provide evidence that pharmacological and dietary interventions are capable of altering seizure susceptibility in adult El mice. In addition to antidepressant and anticonvulsant efficacy of SSRIs, drugs such as fluoxetine also decrease symptoms of anxiety, including social phobia, and impact bioenergetic variables such as appetite, body weight and glucose avail-
A. Richman, S.C. Heinrichs ability (Schatzberg, 2000). It is therefore noteworthy that major characteristics of El mice which differentiate this strain from controls include social withdrawal and low body weight (McFadyen-Leussis and Heinrichs, 2005; Turner et al., in press). Antidepressants could then attenuate seizure activity in El mice to the extent that enrichment of social interactions exert anticonvulsant effects, a finding which has been reported recently (Schridde and van Luijtelaar, 2005). Similarly and in relation to energy balance, fluoxetine is employed therapeutically in obese patients to achieve weight loss (Kordik and Reitz, 1999). Several studies provide evidence for linkage between brain mechanisms of appetite regulation and seizure susceptibility since 5-HT4 mutant mice, for example, exhibit dysregulation in both arenas (Compan et al., 2004). Moreover, work with diabetic mice reveals that fluoxetine can alter blood glucose levels which is an important consideration in prescribing antidepressant medication to diabetics (Gomez et al., 2001). Thus, it will be important in the context of the present studies to examine the potential role of fluoxetine-induced changes in appetite, body weight and glucose bioavailabililty on seizure susceptibility in El mice. No published study has examined anticonvulsant efficacy of fluoxetine in El mice. While most animal studies employ peripheral injection as the route of SSRI drug administration (Pericic et al., 2005), such a procedure is prohibitive given the hyper-reactivity to handling exhibited by El mice (Heinrichs and Seyfried, 2006). The present studies therefore employed oral self-administration via diet adulteration as a means of long-term passive exposure to fluoxetine. Regarding the effective anticonvulsant dose of the drug, fluoxetine exerted a protective effect at 5 mg/kg in a limbic motor seizure model, whereas in the genetically epilepsyprone rat the effective dose was 16 mg/kg (Dailey et al., 1996). Additional studies found a 20 mg/kg dose of fluoxetine to be effective in reducing pentylenetetrazole-induced seizures (Pericic et al., 2005), a 10 mg/kg dose to be effective in attenuating cocaine-induced seizures (Macedo et al., 2004) and a 25 mg/kg dose suppressed audiogenic seizure severity (Tupal and Faingold, 2006). Regarding the duration of fluoxetine exposure, one study showed that when fluoxetine is administered repeatedly for 5 days, anticonvulsant effects of a 20 mg/kg/day sub-chronic dose were equivalent to those exerted by an acute 20 mg/kg dose (Pericic et al., 2005). Thus, the present studies employed an intermediate 10 mg/kg/day dose to examine the hypothesis that oral administration of fluoxetine over either 3 or 7 consecutive days would exert an anticonvulsant effect in seizure susceptible El mice.
Methods Animals El mice were bred and maintained at Boston College from stock donated originally by Dr. Thomas Seyfried. Seizure susceptible El mice (n = 57) were at least 6 months old at the time testing since older mice of the El strain are more susceptible to seizures. Seizure phenotype in response to tail suspension handling for routine husbandry was recorded for each cage of El mice after weaning in order to achieve treatment group matching. Animals had free access to drinking water at all times and were fed ad libitum chow
Seizure prophylaxis in an animal model of epilepsy
21
(ProLab3000, LabDiets, Richmond, IN, USA) prior to the study. The composition of the chow was 59.5% carbohydrate, 22% protein, 5% fat, 11% fiber, and 2.5% added minerals. Group housed mice were re-housed singly 1 week prior to testing in a colony room with a controlled temperature of 72 ◦ C, a humidity of 48%, and a 12-h reverse light cycle (lights off 1000—2200). All experiments were performed in accordance with the guidelines of the Institutional Animal Care and Use Committee of Boston College.
around the perimeter of a standard mouse housing cage. Four infrared photocell beams located along the long axis and eight beams across the width were positioned 2 cm above the bedding at 16 cm intervals. Mice continued to consume their respective experimental diets and were tested in the locomotor apparatus for 24 h following the HISS test.
Chronic fluoxetine self-administration
Mice were anesthetized with the inhaled anesthetic isoflurane (Aerrane, Baxter Healthcare Corporation, Deerfield, IL) prior to intracardial perfusion with 0.1 M phosphate-buffered saline (PBS, pH 7.4), followed by 4% paraformaldehyde in 0.9% saline. Brains were removed immediately and post-fixed overnight at 4 ◦ C. Brains were placed in a 30% sucrose solution in PBS at 4 ◦ C until sinking (48—72 h). Free-floating sections (40 m) were obtained with a cryostat, and stored in cryoprotectant (pH 7.2) at −20 ◦ C until immunohistochemical analysis.
One experimental trial was completed over a 3 day span and three trials were completed over a 7 day span. On Day 1 of each trial, all mice were weighed and singly housed in weight matched groups. Each mouse was provided with a pre-weighed food container with a chow or fluoxetine wet mash ration at the beginning of each 24 h period. An additional empty cage was also employed for each of the four trials to quantify evaporation per unit mass of wet mash and so that this constant could be deducted from total daily food intake. On subsequent days, the food containers were removed, weighed and then refilled with fresh food. Mice in a yoked condition were fed the average amount of food consumed by the fluoxetine group over the previous 24 h period. This procedure assured that non-appetitive actions of fluoxetine were responsible for anticonvulsion efficacy, not simply food restriction, since the fluoxetine group consumed less than the ad libitum chow control group. All feeding/weighing procedures were performed between 1000 and 1600 during the nocturnal portion of the circadian cycle when mice were normally awake and active.
Perfusion and tissue sectioning
Blood glucose monitoring Just prior to intracardial perfusion, a sample of cardiac blood was employed for assessing blood glucose levels at a single time point 7 days following the beginning of fluoxetine exposure and 2 h following HISS testing. The TrueTrack Smart System (Home Diagnostics, Inc. Fort Lauderdale) employed amperametric technology to quantify an oxidation reaction for real time blood glucose level determinations. This device provides a 20—600 mg/dl range of sensitivity and a 3—5% intra-assay precision when assessing blood glucose levels.
Diet preparation Nissl staining Pelleted chow was blended to create a powder to which fluoxetine (Spectrum Chemical, Gardena, CA) was added to yield a roughly 10 mg/kg/day dose of fluoxetine based upon the initial body weight and daily average food intake of experimental mice. Water was added to the dry mixture in order to create a wet mash that was placed in the bottom of a cylindrical chamber in order to avoid spillage. A ration of diet sufficient to result in leftover food the following day was provided to each animal in the ad libitum treatment groups.
Handling-induced seizure susceptibility (HISS) test On the final day of diet exposure, a seizure susceptibility test was performed in which mice were picked up by their tails and held 10—20 cm above the floor of the home cage for 30 s. Mice were then placed in a clean cage for 2 min and then lifted by the tail for another 15 s before being returned to their home cage. Behavioral characteristics of seizures in El mice are evident during three specific phases: (1) prodromal: squeaking and transient immobility, followed by ‘‘running fits’’, (2) ictal: convulsions starting with the hind limbs but rapidly becoming generalized tonic-clonic, loss of postural equilibrium, tail cocked over head (Straub tail), salivation, defecation, and urination and (3) post-ictal: lethargy, ‘‘kangaroo posture’’ (sitting on hind legs with forepaws drawn up), turning of the head in alternate directions, and occasional hyperirritability. Seizures were determined to have occurred in mice that exhibited behavioral signs such as squeaking, tonic-clonic jerking, cocked tail over the head, loss of posture and post-ictal ataxia. An all-or-none criterion was employed in detecting seizures and no attempt was made to score seizure severity.
Locomotor activity test The locomotor apparatus (San Diego Instruments) consisted of photocell grids, each measuring 25 cm × 48 cm, which were positioned
Tissue sections were mounted onto gelatinized slides, rinsed in ethanol and distilled water and then stained in 0.13% cresyl violet in an acetic acid buffer. Stained slides were rinsed in ethanol, cleared in xylene, and coverslipped.
Serotonin transporter (5-HTT) immunohistochemistry Immunohistochemical detection of 5-HTT protein was performed using a primary antibody obtained from Calbiochem (San Carlos, CA). After rinsing in 50% ethanol and quenching of endogenous peroxidase activity with a 3% hydrogen peroxide solution, tissue sections were placed in a 15% normal goat serum blocking solution. Sections were incubated with the rabbit anti-5-HTT antibody (1:1000) overnight at 4 ◦ C. On the second day, sections were incubated with a goat, anti-rabbit secondary antibody (1:1000) for 1 h and then a streptavidin-HRP conjugate (ABC kit) for 30 min (both from Vector Laboratories, San Carlos, CA). Sections were stained using a DAB chromagen and a hematoxylin counterstain. Immunohistochemical labeling was eliminated when either the primary or secondary antibody was omitted from the assay.
Brain mapping by microscopy Sections were photographed using an RT color Spot camera (Diagnostic Instruments Inc., Sterling Heights, MI) mounted on a Zeiss bright-field microscope. Nissl positive and 5-HTT immunoreactive cells were counted using the IP Lab image analysis software (Scanalytics, Fairfax, VA). In order to avoid potential experimenter and laterality biases, analyses were completed without knowledge of treatment condition using brain tissue from both sides of the sagittal plane. Sites were selected for analysis based upon a series of cFos mapping studies performed previously in El strain mice (McFadyenLeussis and Heinrichs, 2004), by coarse examination for areas
22 of intense staining by a treatment-blind observer, and by the known distribution of serotonin transporter protein in rodent brain (Cabrera-Vera and Battaglia, 1998). In particular, published electrophysiological and activity mapping studies have identified parietal cortex as a likely source of seizures in the El brain (Nakamoto et al., 1990; Ishida et al., 1993). Based upon studies in our own laboratory suggesting that immediate-early gene expression emerges in the parietal cortex of El brain early in development, the present studies tested the a priori hypothesis that this known seizure locus would by modulated by fluoxetine exposure. Moreover, sub-cortical diencephalic and basal forebrain regions are reported to constitute brain sites for seizure facilitation and interactivity with other brain seizure substrates (Eells et al., 2004). In particular, published studies demonstrate that active cells are more abundant in paraventricular thalamus after seizure induction (Mraovitch and Calando, 1999) and it has been suggested that recurring activation of the paraventricular thalamus implies that this region stores memories of previous, stress-related experiences (Fenoglio et al., 2006). As the El mouse model is known to be hyper-reactive to stressor exposure (Drage and Heinrichs, 2005), the present studies tested the a priori hypothesis that the paraventricular thalamus would be modulated by fluoxetine exposure. Finally, 5-HTT is enriched in brain areas such as amygdala and nucleus accumbens which are reciprocally connected and constitute sites vital for goal-directed responses and plasticity (Sur et al., 1996; Stevenson and Gratton, 2003). In preliminary studies, long-term fluoxetine exposure was found to selectively enhance 5-HTT expression in the nucleus accumbens (unpublished observations) and the present studies therefore tested the a priori hypothesis that these brain regions would be modulated in El mice by fluoxetine exposure. Thus, Nissl and 5-HTT labeled cells were counted in the parietal cortex, paraventricular nucleus of thalamus, nucleus accumbens, and basolateral nucleus of the amygdala of El mouse brain in the present studies. The selection of these four brain regions of interest based upon the prior research findings just described effectively reduces the likelihood that the present cell density and activation measures would reflect non-specific actions of systemically administered fluoxetine. Specific brain regions and nuclei of interest were identified using a mouse brain atlas (Paxinos and Franklin, 2001).
A. Richman, S.C. Heinrichs
Figure 1 Frequency of tail suspension-induced seizures in El mice following either three or seven consecutive 24-h periods of exposure to ad libitum chow diet, ad libitum fluoxetineadultered diet, or yoked chow diet conditions. * p < 0.05 relative to chow and yoked controls.
of dietary condition with mice eating more ad libitum chow diet than either the fluoxetine or yoked diets (Fig. 2). A significant main effect of dietary condition on body weight [F(2,12) = 4.5, p < 0.05] resulted from significant loss of body weight in the fluoxetine-treated group relative to the chow and yoked diet groups (Fig. 2). Note that body weight loss resulting from fluoxetine treatment likely reflects negative energy balance actions of the drug other than appetite suppression (Dubuc and Peterson, 1990) since mice in the yoked group ate the same quantity of food over the 3 day trial without experiencing body weight loss. Mice were immediately aware of placement of food rations in their home cages as 50, 40 and 20% of animals in the chow, fluoxetine and yoked diet groups interacted with the
Statistical analysis For all experiments, mixed factor analyses of variance (ANOVA) were performed for food intake, body weight, locomotor activity and local brain site cell density with treatment as a between subjects factor and experimental hour or day as a within subjects factor. The 3 day trial employed n = 5 mice/treatment group whereas the 7 day trial employed 12—15 mice/treatment group; a subset of n = 4 mice was selected from each group in each trial for histochemical analysis. Simple main effect analyses were conducted when appropriate to determine individual group differences, and comparisons were considered significant when p < 0.05. Seizure frequency results were examined using non-parametric Kruskal—Wallis tests.
Results Three day fluoxetine self-administration trial HISS testing after 3 days of dietary exposure produced a relatively high, and not significantly different (Kruskal—Wallis < 1, ns), rate of seizures in all treatment groups: 40% (2 of 5) in the chow group, 40% (2 of 5) in the yoked group, and 60% (3 of 5) in the fluoxetine group (Fig. 1). Analysis of food intake over the 3 day dietary trial revealed a significant [F(1,8) = 33.4, p < 0.001] main effect
Figure 2 Mean ± S.E.M. daily food intake of wet mash with evaporation taken into account (top) and body weight change (bottom) in El mice over three consecutive 24-h periods of exposure to ad libitum chow diet, ad libitum fluoxetine-adultered diet or yoked chow diet conditions.
Seizure prophylaxis in an animal model of epilepsy
23
Table 1 Cell density assessed using a Nissl stain in El mice exposed over the previous 3 days to chow, fluoxetine or yoked diets Diet condition
NAc
PVT
Amyg
Cortex
Chow Fluoxetine Yoked
569 ± 78 661 ± 93 396 ± 22
490 ± 85 523 ± 69 390 ± 64
601 ± 89 528 ± 49 489 ± 8
483 ± 31 567 ± 41 520 ± 53
Values are mean ± S.E.M. based upon n = 4 mice in each treatment group. Amyg: basolateral amygdalar nucleus, Cortex: parietal cortex, NAc: nucleus accumbens, PVT: paraventricular nucleus of thalamus.
food receptacle without delay. These results suggest that short-term fluoxetine-induced anorexia and body weight loss did not impact susceptibility to handling-induced seizures. Examination of cell density assessed by Nissl staining in nucleus accumbens, paraventricular nucleus of thalamus, amygdala and parietal cortex following a 3 day period of exposure to fluoxetine or chow diet treatments revealed no significant effects of condition (Table 1).
Seven day fluoxetine self-administration trial HISS testing after 7 days of dietary exposure revealed a relatively high rate of seizures in the two control groups: 40% (6 of 15) in the chow group and 42% (5 of 12) in the yoked group relative to a significantly (Kruskal—Wallis = 8.09, p < 0.02) decreased 0% (0 of 15) in the fluoxetine group (Fig. 1). Analysis of food intake revealed a main effect of treatment [F(2,41) = 43.8, p < 0.001] with mice in the ad libitum chow group consuming more food than mice in the fluoxetine and yoked groups (Fig. 3 ). Moreover, animals in the fluoxetine and yoked groups lost significantly [F(1,25) = 6.5, p < 0.005] more body weight than the ad libitum chow control group (Fig. 3). Mice in the three diet conditions were observed to begin eating immediately after placement of the food receptacle (chow, 65%; yoked, 44%; fluoxetine, 67%) indicating that none of the diet conditions were aversive to the extent that the animal would avoid the diets or the locations in which they were served. These results demonstrate that a 7 day diet yoking procedure was effective in matching daily food intake and body weight changes of mice not exposed to fluoxetine to those exhibited by the ad libitum fluoxetine group. Analysis of locomotor activity revealed a significant main effect of time [F(23,598) = 7.4, p < 0.001] without an effect of treatment. The effect of time reflects a circadian activity rhythm in which mice are more active during the nocturnal hours (Fig. 4). These results suggest that consumption over 7 days of a fluoxetine-adultered diet abolishes susceptibility to handling-induced seizures in El mice without disrupting the normal profile of circadian locomotor activity. At the time of sacrifice for brain harvesting, blood glucose levels were not affected by the dietary treatments administered over 7 days—–chow group: 279 ± 15 mg/dl, yoked group: 279 ± 9 mg/dl fluoxetine: 271 ± 10 mg/dl. However, cell density assessed by Nissl staining was altered in nucleus accumbens, amygdala and parietal cortex by diet condition [F(2,5) = 5.8, 8.2 and 8.6, respectively, p < 0.05] with lower densities in nucleus accumbens and amygdala in
Figure 3 Mean ± S.E.M. daily food intake of wet mash with evaporation taken into account (top) and body weight change (bottom) in El mice over seven consecutive 24-h periods of exposure to ad libitum chow diet, ad libitum fluoxetine-adultered diet or yoked chow diet conditions.
the yoked group, as well as lower densities in the parietal cortex in the fluoxetine group (p < 0.05) relative to the chow control group (Fig. 5). Examination of 5-HTT immunoreactivity in four representative brain regions revealed a significant [F(2,7) = 3.7, p < 0.05] main effect of treatment in the nucleus accumbens with 7 days of fluoxetine exposure elevating the concentration of the serotonin transporter protein relative to the chow control group (Fig. 6). These
Figure 4 Locomotor activity (mean ± S.E.M.) of El mice over a 24-h period of continued exposure to chow diet, fluoxetineadultered diet or yoked diet conditions beginning after Day 7 handling-induced seizure susceptibility testing. The white bar reflects the diurnal portion of the circadian cycle and the dark bar reflects the nocturnal portion.
24
A. Richman, S.C. Heinrichs
Figure 5 Cell density (mean ± S.E.M.) assessed using a Nissl stain in El mice exposed over the previous 7 days to chow, fluoxetine or yoked diets (top panel). Representative sections of parietal cortex are provided for mice in the chow diet (left) and fluoxetine diet (right) treatment groups. Amyg: basolateral amygdalar nucleus, Cortex: parietal cortex, NAc: nucleus accumbens, PVT: paraventricular nucleus of thalamus, SC: subarachnoid cavity, SSp: primary somatosensory area. * p < 0.05 relative to chow diet control group.
results suggest that a 7 day exposure to fluoxetine-adultered diet is sufficient to induce neural adaptations to the drug which did not include changes in resting blood glucose.
Discussion The main finding of the present studies is that a 7 day period of dietary exposure to a 10 mg/kg/day dose of fluoxetine was effective in abolishing precipitated seizures in El mice. The anticonvulsant effect of fluoxetine was time-of-exposure dependent in that a shorter 3 day trial demonstrated a high 60% seizure frequency in fluoxetine-exposed El mice which was comparable to the 40% seizure frequency of chow diet controls. Moreover, a yoked diet control group was used to assess the alternative hypothesis that efficacy of fluoxetine was dependent on diminished food intake and body weight loss. The fact that 40% of yoked group mice seized in the 7 day trial documented that non-specific consequences of diet restriction were not sufficient to suppress seizure susceptibility in El mice. Instead, the ability of fluoxetine treatment to enhance the serotonin transporter protein target for SSRI drugs in the basal forebrain and reduce cell density in a cortical region implicated in seizure etiology in El mice suggests that fluoxetine attenuated seizures by remodeling
brain seizure susceptibility circuits. This conclusion is supported by the fact that the change in cortical cell density produced by 7 day fluoxetine exposure was not reproduced by a 3 day period of fluoxetine exposure. Thus, changes produced by 7 day SSRI treatment were neuroadaptive in nature rather than simple non-specific, pharmacological effects of exposure to the drug. Future studies can test this hypothesis using local intra-cerebral administration of fluoxetine into the present candidate nuclei. The weight of scientific evidence suggests that increasing extracellular norepinephrine and/or serotonin by blockade of the reuptake of these monoamines is the pharmacodynamic action responsible for the anticonvulsant effects of antidepressants (Dailey and Naritoku, 1996; Jobe and Browning, 2005). The present results support this claim in that 1 week of exposure to a fluoxetine-adultered diet resulted in increased serotonin transporter protein content in the nucleus accumbens and a correlated elimination of handling-induced seizures in genetically susceptible El mice. While the serotonin transporter immunohistochemistry endpoint serves as an important manipulation check in the present studies to insure that dietary fluoxetine exerted direct actions on serotonergic neurotransmission, the finding of increased serotonin transporter content in the nucleus accumbens may provide insight into the mech-
Seizure prophylaxis in an animal model of epilepsy
25
Figure 6 Serotonin transporter protein (5-HTT) content (mean ± S.E.M.) assessed immunohistochemically in El mice exposed over the previous 7 days to chow, fluoxetine or yoked diets (top panel). Representative sections of nucleus accumbens are provided for mice in the chow diet (left) and fluoxetine diet (right) treatment groups. aca: anterior commisure, Amyg: basolateral amygdalar nucleus, Cortex: parietal cortex, LV: lateral ventricle, NAc: nucleus accumbens, PVT: paraventricular nucleus of thalamus. * p < 0.05 relative to chow diet control group.
anism of seizure susceptibility in the El mouse model. Brain reward circuitry, which includes the nucleus accumbens as a forebrain terminus of an affective regulatory pathway in rodents (Berridge, 1996), has been identified as one component of the seizure initiation pathway in the epileptic brain (Jobe and Browning, 2005). In complementary fashion, the suspension of seizure susceptibility in El mice via fluoxetine exposure in the present studies supports the contention that increasing serotonergic neurotransmission serves as a ‘‘exterior defense shield’’ (Jobe and Browning, 2005) against tail suspension handling-induced convulsions in El mice. Similarly, evidence from neuroimaging studies suggests that vagal nerve stimulation therapy for epilepsy acts via brainstem serotonergic and noradrenergic projections to limbic and cortical structures that are involved in mood regulation (Nemeroff et al., 2006). Given the fact that the affective dysregulation profile of El mice is robust and varied with behavioral, neural and physiological components (Drage and Heinrichs, 2005), future studies could pursue the claimed comorbidity of epilepsy and affective disorders (Nemeroff et al., 2006) using this animal model. It is well known that clinically relevant dietary restriction or provision of alternative diets exert anticonvulsive effects in animal models of epilepsy including the El mouse (Stafstrom and Bough, 2003; Mantis et al., 2004; Seyfried et
al., 2004). An autonomous anticonvulsant mechanism based entirely on nutritional change is potentially troublesome in the present studies for two reasons: (1) fluoxetine was administered via diet adulteration in the present studies such that attempts to self-regulate intake of the drug also impact food intake and vice versa and (2) fluoxetine facilitates serotonergic neurotransmission which is well known to alter appetite and body weight in multiple species (Heisler et al., 1997; Harvey and Bouwer, 2000). The oral route of fluoxetine self-administration was selected for the present studies since the El mouse phenotype of stress hyperreactivity is prohibitive for daily systemic or parenteral administration procedures of the sort typically employed for chronic administration of fluoxetine in rodents (Kecskemeti et al., 2005; Pericic et al., 2005). In particular, such drug administration procedures are stressful in and of themselves (Balcombe et al., 2004) and would be expected either to trigger seizures in El mice or to produce a seizure refractory period which would reduce HISS test sensitivity. In addition, several procedures were employed in the present studies to guard against potential non-specific appetitive actions of fluoxetine on seizure susceptibility. First, mice in the various treatment groups were weight matched prior to beginning the experimental trials. Second, a diet yoking procedure insured that this control group would exhibit the
26 same changes in food intake as fluoxetine self-administering mice. Third, a post-mortem blood glucose measure was employed to assess this bioenergetic component on the day of seizure susceptibility testing. Finally, a behavioral measure of motivation to eat was employed to assess the eagerness of mice in the various treatment groups to seek out their daily dietary ration. The home cage measure of immediate eating revealed that a significant proportion of mice in all three diet groups including the fluoxetine group were eager to come into contact with the food dish and begin eating without delay. This suggests that there were no general motivational confounds in fluoxetine-treated mice and this conclusion was supported by the lack of effect of fluoxetine on overall home cage motor activity beginning at the time of seizure susceptibility testing. Finally, single time point resting blood glucose levels were equivalent among the three dietary groups at the time of seizure susceptibility testing. Taken together with the lack of prophylaxis observed for seizure susceptibility in the yoked diet group, these results suggest that non-specific bioenergetic variables are not able to account for anticonvulsant efficacy of fluoxetine in the present studies. The present results add the El mouse to the growing list of animal models of epilepsy in which fluoxetine has been found to exert anticonvulsant actions (Dailey et al., 1996; Macedo et al., 2004; Ugale et al., 2004; Kecskemeti et al., 2005; Pericic et al., 2005). One mechanism for reduced seizure frequency in El mice treated with fluoxetine may be remodeling of the somatosensory region of the parietal cortex which has been identified previously as a seizure trigger zone (Ishida et al., 1993; Murashima et al., 1996). However, the present results also support an alternative hypothesis that protective serotonergic circuits in affective regulatory regions of the brain serve as seizure modulators. Consistent with the later alternative, high emotionality characteristics of the El mouse separate from seizure susceptibility also suggest efficacy of fluoxetine. For example, El mice are hyper-reactive to stressor exposure (Drage and Heinrichs, 2005), exhibit an anxiogenic-like phenotype in response to novelty (Pascual and Heinrichs, in press), and these characteristics could potentially be ameliorated by the anxiolytic efficacy of fluoxetine (Silva and Brandao, 2000). Similarly, El mice exhibit social withdrawal (Turner et al., in press) and an impairment in social recognition memory (Lim et al., in press) and these behavioral and cognitive deficiencies are reported to be normalized by fluoxetine (El Hage et al., 2004). Future studies can address whether seizure protective effects of fluoxetine in El mice extend into other realms of affective or anti-social psychopathology.
Acknowledgements We thank Chen Lim, Jennifer Pascual and Laura Turner for their help with tissue harvesting. This research was supported by C.U.R.E. and a Research Incentive Grant from Boston College to SCH.
References Balcombe, J.P., Barnard, N.D., et al., 2004. Laboratory routines cause animal stress. Contemp. Top Lab. Anim. Sci. 43 (6), 42—51.
A. Richman, S.C. Heinrichs Berridge, K.C., 1996. Food reward: brain substrates of wanting and liking. Neurosci. Biobehav. Rev. 20 (1), 1—25. Cabrera-Vera, T.M., Battaglia, G., 1998. Prenatal exposure to fluoxetine (Prozac) produces site-specific and age-dependent alterations in brain serotonin transporters in rat progeny: evidence from autoradiographic studies. J. Pharmacol. Exp. Ther. 286 (3), 1474—1481. Compan, V., Zhou, M., et al., 2004. Attenuated response to stress and novelty and hypersensitivity to seizures in 5-HT4 receptor knock-out mice. J. Neurosci. 24 (2), 412—419. Dailey, J.W., Mishra, P.K., et al., 1992. Serotonergic abnormalities in the central nervous system of seizure-naive genetically epilepsyprone rats. Life Sci. 50 (4), 319—326. Dailey, J.W., Naritoku, D.K., 1996. Antidepressants and seizures: clinical anecdotes overshadow neuroscience. Biochem. Pharmacol. 52 (9), 1323—1329. Dailey, J.W., Yan, Q.S., et al., 1996. Neurochemical correlates of antiepileptic drugs in the genetically epilepsy-prone rat (GEPR). Life Sci. 58 (4), 259—266. Drage, M.G., Heinrichs, S.C., 2005. Phenotyping the untouchables: environmental enhancement of behavioral and physiological activation in seizure-prone El mice. Epilepsy Behav. 6 (1), 35—42. Dubuc, P.U., Peterson, C.M., 1990. Ineffectiveness of parenteral fluoxetine or RU-486 to alter long-term food intake, body weight or body composition of genetically obese mice. J. Pharmacol. Exp. Ther. 255 (3), 976—979. Eells, J.B., Clough, R.W., et al., 2004. Comparative fos immunoreactivity in the brain after forebrain, brainstem, or combined seizures induced by electroshock, pentylenetetrazol, focally induced and audiogenic seizures in rats. Neuroscience 123 (1), 279—292. El Hage, W., Peronny, S., et al., 2004. Impaired memory following predatory stress in mice is improved by fluoxetine. Prog. Neuropsychopharmacol. Biol. Psychiatry 28 (1), 123—128. Fenoglio, K.A., Chen, Y., et al., 2006. Neuroplasticity of the hypothalamic—pituitary—adrenal axis early in life requires recurrent recruitment of stress-regulating brain regions. J. Neurosci. 26 (9), 2434—2442. Ferrero, A.J., Cereseto, M., et al., 2005. Chronic treatment with fluoxetine decreases seizure threshold in naive but not in rats exposed to the learned helplessness paradigm: Correlation with the hippocampal glutamate release. Prog. Neuropsychopharmacol. Biol. Psychiatry 29 (5), 678—686. Gomez, R., Huber, J., et al., 2001. Acute effect of different antidepressants on glycemia in diabetic and non-diabetic rats. Braz. J. Med. Biol. Res. 34 (1), 57—64. Harvey, B.H., Bouwer, C.D., 2000. Neuropharmacology of paradoxic weight gain with selective serotonin reuptake inhibitors. Clin. Neuropharmacol. 23 (2), 90—97. Hashiguchi, W., Nagatomo, I., et al., 2001. Influences of caffeine to nitric oxide production and zonisamide concentration in the brain of seizure-susceptible EL mice. Psychiatry Clin. Neurosci. 55 (4), 319—324. Heinrichs, S.C., Seyfried, T.N., 2006. Behavioral seizure correlates in animal models of epilepsy: a road map for assay selection, data interpretation, and the search for causal mechanisms. Epilepsy Behav. 8 (1), 5—38. Heisler, L.K., Kanarek, R.B., et al., 1997. Fluoxetine decreases fat and protein intakes but not carbohydrate intake in male rats. Pharmacol. Biochem. Behav. 58 (3), 767—773. Hernandez, E.J., Williams, P.A., et al., 2002. Effects of fluoxetine and TFMPP on spontaneous seizures in rats with pilocarpineinduced epilepsy. Epilepsia 43 (11), 1337—1345. Hiramatsu, M., 1981. Brain monoamine levels and E1 mouse convulsions. Folia Psychiatr. Neurol. Jpn. 35 (3), 261—266. Ishida, N., Kasamo, K., et al., 1993. Epileptic seizure of El mouse initiates at the parietal cortex: depth EEG observation in freely
Seizure prophylaxis in an animal model of epilepsy moving condition using buffer amplifier. Brain Res. 608 (1), 52—57. Jobe, P.C., Browning, R.A., 2005. The serotonergic and noradrenergic effects of antidepressant drugs are anticonvulsant, not proconvulsant. Epilepsy Behav. 7 (4), 602—619. Kecskemeti, V., Rusznak, Z., et al., 2005. Norfluoxetine and fluoxetine have similar anticonvulsant and Ca2+ channel blocking potencies. Brain Res. Bull. 67 (1/2), 126—132. King, J.T., LaMotte, C.C., 1989. El mouse as a model of focal epilepsy: a review. Epilepsia 30 (3), 257—265. Kordik, C.P., Reitz, A.B., 1999. Pharmacological treatment of obesity: therapeutic strategies. J. Med. Chem. 42 (2), 181—201. Leussis, M.P., Heinrichs, S.C., in press. Temporal ontogeny of seizure circuit activation prior to the onset of seizure susceptibility in EL mice. Neuroscience. Lim, C.E., Turner, L.H., et al., in press. Short-term social recognition memory deficit and altered social and physiological stressor reactivity in seizure-susceptible El mice. Seizure. Lopez-Meraz, M.L., Gonzalez-Trujano, M.E., et al., 2005. 5-HT1A receptor agonists modify epileptic seizures in three experimental models in rats. Neuropharmacology 49 (3), 367—375. Macedo, D.S., Santos, R.S., et al., 2004. Effect of anxiolytic, antidepressant, and antipsychotic drugs on cocaine-induced seizures and mortality. Epilepsy Behav. 5 (6), 852—856. Mantis, J.G., Centeno, N.A., et al., 2004. Management of multifactorial idiopathic epilepsy in EL mice with caloric restriction and the ketogenic diet: role of glucose and ketone bodies. Nutr. Metab. (Lond.) 1 (1), 11. McFadyen-Leussis, M., Heinrichs, S.C., 2004. Handling of EL/Suz mice elicits differential fos activation in the locus ceruleus during development. In: Society for Neuroscience Annual Meeting, San Diego, California. McFadyen-Leussis, M.P., Heinrichs, S.C., 2005. Seizure-prone EL/Suz mice exhibit physical and motor delays and heightened locomotor activity in response to novelty during development. Epilepsy Behav. 6 (3), 312—319. Mraovitch, S., Calando, Y., 1999. Interactions between limbic, thalamo-striatal-cortical, and central autonomic pathways during epileptic seizure progression. J. Comp. Neurol. 411 (1), 145—161. Murashima, Y.L., Kassamo, K., et al., 1996. Developmental and seizure-related regional differences in immediate early gene expression and GABAergic abnormalities in the brain of EL mice. Epilepsy Res. 26 (1), 3—14. Murashima, Y.L., Suzuki, J., et al., 2004. In: Benjamin, S.M. (Ed.), Ictogenesis and Epileptogenesis in Epileptic Mutant EL Mice. Focus on Epilepsy Research. Nova Science Publishers, Inc., New York, pp. 139—198. Murashima, Y.L., Yoshii, M., et al., 2002. Ictogenesis and epileptogenesis in EL mice. Epilepsia 43 (Suppl. 5), 130—135. Mutoh, K., Ito, M., et al., 1993. Depth EEG in mutant epileptic E1 mice: demonstration of secondary generalization of the seizure from the hippocampus. Electroencephalogr. Clin. Neurophysiol. 86 (3), 205—212. Nakamoto, Y., Nakayama, S., et al., 1990. Cerebral uptake of [14C]deoxyglucose during the entire seizure and the recovery period in an El mouse. Epilepsy Res. 5 (1), 43—48. Nemeroff, C.B., Mayberg, H.S., et al., 2006. VNS Therapy in treatment-resistant depression: clinical evidence and putative neurobiological mechanisms. Neuropsychopharmacology. 31 (7), 1345—1355.
27 Pascual, J., Heinrichs, S.C., in press. Olfactory neophobia and seizure susceptibility phenotypes in an animal model of epilepsy are normalized by impairment of brain corticotropin releasing factor. Epilepsia. Paxinos, G., Franklin, K.B.J., 2001. The Mouse Brain in Stereotaxic Coordinates. Academic Press, San Diego. Pericic, D., Lazic, J., et al., 2005. Anticonvulsant effects of acute and repeated fluoxetine treatment in unstressed and stressed mice. Brain Res. 1033 (1), 90—95. Rho, J.M., Kim, D.W., et al., 1999. Age-dependent differences in flurothyl seizure sensitivity in mice treated with a ketogenic diet. Epilepsy Res. 37 (3), 233—240. Sarkisian, M.R., 2001. Overview of the current animal models for human seizure and epileptic disorders. Epilepsy Behav. 2 (3), 201—216. Schatzberg, A.F., 2000. New indications for antidepressants. J. Clin. Psychiatry 61 (Suppl. 2), 9—17. Schridde, U., van Luijtelaar, G., 2005. The role of the environment on the development of spike-wave discharges in two strains of rats. Physiol. Behav. 84 (3), 379—386. Seyfried, T.N., Glaser, G.H., 1985. A review of mouse mutants as genetic models of epilepsy. Epilepsia 26 (2), 143—150. Seyfried, T.N., Greene, A.E., et al., 2004. Caloric restriction and epilepsy: historical perspectives, relationship to the ketogenic diet, and analysis in epileptic EL mice. In: Stafstrom, C.E., Rho, J.M. (Eds.), Epilepsy and the Ketogenic Diet. Humana Press, pp. 247—264. Silva, R.C.B., Brandao, M.L., 2000. Acute and chronic effects of gepirone and fluoxetine in rats tested in the elevated plusmaze: an ethological analysis. Pharmacol. Biochem. Behav. 65 (2), 209—216. Stafstrom, C.E., Bough, K.J., 2003. The ketogenic diet for the treatment of epilepsy: a challenge for nutritional neuroscientists. Nutr. Neurosci. 6 (2), 67—79. Stevenson, C.W., Gratton, A., 2003. Basolateral amygdala modulation of the nucleus accumbens dopamine response to stress: role of the medial prefrontal cortex. Eur. J. Neurosci. 17 (6), 1287—1295. Sur, C., Betz, H., et al., 1996. Immunocytochemical detection of the serotonin transporter in rat brain. Neuroscience 73 (1), 217— 231. Suzuki, J., Nakamoto, Y., 1982. Abnormal plastic phenomena of sensory-precipitated epilepsy in the mutant El mouse. Exp. Neurol. 75 (2), 440—452. Suzuki, J., Nakamoto, Y., et al., 1983. Local cerebral glucose utilization in epileptic seizures of the mutant El mouse. Brain Res. 266 (2), 359—363. Tecott, L.H., Sun, L.M., et al., 1995. Eating disorder and epilepsy in mice lacking 5-HT2c serotonin receptors. Nature 374 (6522), 542—546. Tupal, S., Faingold, C.L., 2006. Evidence supporting a role of serotonin in modulation of sudden death induced by seizures in DBA/2 mice. Epilepsia 47 (1), 21—26. Turner, L.H., Lim, C.E., et al., in press. Anti-social and seizure susceptibility phenotypes in an animal model of epilepsy are normalized by impairment of brain corticotropin releasing factor. Epilepsy Behav. Ugale, R.R., Mittal, N., et al., 2004. Essentiality of central GABAergic neuroactive steroid allopregnanolone for anticonvulsant action of fluoxetine against pentylenetetrazole-induced seizures in mice. Brain Res. 1023 (1), 102—111.
journal homepage: www.elsevier.com/locate/epilepsyres
Seizure prophylaxis in an animal model of epilepsy by dietary fluoxetine supplementation Alyssa Richman, Stephen C. Heinrichs ∗ Department of Psychology, 140 Commonwealth Avenue, Chestnut Hill, MA 02467, United States Received 13 August 2006; received in revised form 20 November 2006; accepted 30 November 2006 Available online 9 January 2007
KEYWORDS Antidepressant; Tail suspension handling; Seizure; Locomotor activity; Serotonin
Summary Clinical and animal model evidence suggests that selective serotonin reuptake inhibitors (SSRIs) act as anticonvulsants. The present studies tested the possibility that the El mouse model of genetically predisposed/handling-triggered epilepsy would exhibit fewer seizures following SSRI treatment via dietary fluoxetine adulteration. In particular, potential bioenergetic and neural mechanisms for anticonvulsant efficacy of fluoxetine were explored using food intake/body weight monitoring and quantification of brain serotonin transporter protein. El mice consuming a chow diet ad libitum or yoked in quantity to fluoxetine diet intake exhibited seizure incidence of 40% in response to tail-suspension handling, whereas seizures were abolished (0%) among El mice consuming a fluoxetine-adultered diet over 7 days. A 3 day period of fluoxetine administration was insufficient to exert anticonvulsant efficacy and all treatment groups exhibited the same circadian locomotor activity patterns at the time of seizure susceptibility testing. Bioenergetic factors could not account for the anticonvulsant efficacy of fluoxetine since yoked diet controls with matched food intake, body weight change and blood glucose levels exhibited the same 40% seizure incidence as ad libitum chow controls. Importantly, the 7 day period of dietary fluoxetine exposure was effective in selectively reducing cell density in the parietal cortex and increasing serotonin transporter protein content in the nucleus accumbens. Taken together, these results suggest that dietary fluoxetine supplementation abolishes handling-induced seizure susceptibility in El mice via a neural remodeling mechanism independent of energy balance. © 2006 Elsevier B.V. All rights reserved.
Introduction
∗ Corresponding author at: VA Medical Center, Research 151Neuropharmacology, 150 South Huntington Avenue, Boston, MA 02130, United States. Tel.: +1 857 364 5617; fax: +1 617 278 4540. E-mail address: [email protected] (S.C. Heinrichs).
While selective serotonin reuptake inhibitors (SSRIs) are a first line pharmacotherapy used to treat clinical depression (Dailey and Naritoku, 1996), fluoxetine and other antidepressants also exert anticonvulsant effects (Hernandez et al., 2002; Ferrero et al., 2005; Kecskemeti et al., 2005; Pericic et al., 2005; Tupal and Faingold, 2006). Serotonergic signaling appears to have a direct role in modulating
0920-1211/$ — see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.eplepsyres.2006.11.007
20 seizure susceptibility as both 5-HT2C and 5-HT4 knockout mice exhibit enhanced sensitivity to pentylenetetrazoleinduced convulsions (Tecott et al., 1995; Compan et al., 2004). Similarly, resting state levels of serotonin (5-HT) are lower in seizure prone mice, as compared to 11 inbred strains (King and LaMotte, 1989). Consistent with a causal role for 5-HT deficiency in seizure incidence, administration of 5-hydroxytrytophan (a 5-HT precursor) with MK-486 (a peripheral decarboxylation inhibitor which limits 5-HT synthesis to the brain) increases brain 5-HT levels in seizure prone mice, and significantly decreases the occurrence of seizures (Hiramatsu, 1981). At the receptor level, the 5-HT1A receptor agonist, WAY100635, modifies epileptic activity depending on the type of seizure (Lopez-Meraz et al., 2005). Chronic treatment with fluoxetine decreases epileptic threshold in a manner correlated with levels of hippocampal glutamate release (Ferrero et al., 2005) and fluoxetine reduces spontaneous seizures in a model of temporal lobe epilepsy (Hernandez et al., 2002). This evidence suggests that serotonergic neurotransmission inhibits neuronal network excitability (Dailey et al., 1992) although consensus has been slow to develop on this point due to confusion between manifestations of illness versus treatment effects (Jobe and Browning, 2005). One aim of the present studies was to provide a better understanding of the central mechanisms of action for the anticonvulsant effects of fluoxetine. The following experiments employed the ‘epilepsy’ (El) mouse model to monitor seizure susceptibility after subchronic oral administration of the antidepressant fluoxetine. The El progenitor mouse resulted from a spontaneous mutation which was developed subsequently into the El strain through selective breeding to maintain and amplify the seizure prone phenotype (Murashima et al., 2002). Seizures originate in the parietal cortex of El mice, and subsequently spread to other brain regions such as the hippocampus (Ishida et al., 1993). The seizure locus may shift to the hippocampus in seizure-experienced El mice (Mutoh et al., 1993) although this migration is a dynamic process dependent upon age and handling history (Leussis and Heinrichs, in press). Susceptibility to seizures in El mice increases with age, and with repeated sensory stimulations involving tail-suspension, or ‘tossing-up’ (Suzuki and Nakamoto, 1982). The El mouse thus constitutes an animal model for idiopathic complex partial seizures (Seyfried and Glaser, 1985; Murashima et al., 2004), for temporal lobe epilepsy (Suzuki et al., 1983) and for simple reflex epilepsy (Sarkisian, 2001). El mice are genetically predisposed to seize although episodes are not intractable since pharmacological and dietary studies have demonstrated the adult plasticity of seizure susceptibility (Heinrichs and Seyfried, 2006). For example, caffeine augments seizures in El mice by altering the pharmacokinetics of a concurrently administered anticonvulsant drug (Hashiguchi et al., 2001). Similarly, high fat, low carbohydrate ketogenic diet exposure exerts anticonvulsant effects in El mice (Rho et al., 1999; Seyfried et al., 2004). These findings provide evidence that pharmacological and dietary interventions are capable of altering seizure susceptibility in adult El mice. In addition to antidepressant and anticonvulsant efficacy of SSRIs, drugs such as fluoxetine also decrease symptoms of anxiety, including social phobia, and impact bioenergetic variables such as appetite, body weight and glucose avail-
A. Richman, S.C. Heinrichs ability (Schatzberg, 2000). It is therefore noteworthy that major characteristics of El mice which differentiate this strain from controls include social withdrawal and low body weight (McFadyen-Leussis and Heinrichs, 2005; Turner et al., in press). Antidepressants could then attenuate seizure activity in El mice to the extent that enrichment of social interactions exert anticonvulsant effects, a finding which has been reported recently (Schridde and van Luijtelaar, 2005). Similarly and in relation to energy balance, fluoxetine is employed therapeutically in obese patients to achieve weight loss (Kordik and Reitz, 1999). Several studies provide evidence for linkage between brain mechanisms of appetite regulation and seizure susceptibility since 5-HT4 mutant mice, for example, exhibit dysregulation in both arenas (Compan et al., 2004). Moreover, work with diabetic mice reveals that fluoxetine can alter blood glucose levels which is an important consideration in prescribing antidepressant medication to diabetics (Gomez et al., 2001). Thus, it will be important in the context of the present studies to examine the potential role of fluoxetine-induced changes in appetite, body weight and glucose bioavailabililty on seizure susceptibility in El mice. No published study has examined anticonvulsant efficacy of fluoxetine in El mice. While most animal studies employ peripheral injection as the route of SSRI drug administration (Pericic et al., 2005), such a procedure is prohibitive given the hyper-reactivity to handling exhibited by El mice (Heinrichs and Seyfried, 2006). The present studies therefore employed oral self-administration via diet adulteration as a means of long-term passive exposure to fluoxetine. Regarding the effective anticonvulsant dose of the drug, fluoxetine exerted a protective effect at 5 mg/kg in a limbic motor seizure model, whereas in the genetically epilepsyprone rat the effective dose was 16 mg/kg (Dailey et al., 1996). Additional studies found a 20 mg/kg dose of fluoxetine to be effective in reducing pentylenetetrazole-induced seizures (Pericic et al., 2005), a 10 mg/kg dose to be effective in attenuating cocaine-induced seizures (Macedo et al., 2004) and a 25 mg/kg dose suppressed audiogenic seizure severity (Tupal and Faingold, 2006). Regarding the duration of fluoxetine exposure, one study showed that when fluoxetine is administered repeatedly for 5 days, anticonvulsant effects of a 20 mg/kg/day sub-chronic dose were equivalent to those exerted by an acute 20 mg/kg dose (Pericic et al., 2005). Thus, the present studies employed an intermediate 10 mg/kg/day dose to examine the hypothesis that oral administration of fluoxetine over either 3 or 7 consecutive days would exert an anticonvulsant effect in seizure susceptible El mice.
Methods Animals El mice were bred and maintained at Boston College from stock donated originally by Dr. Thomas Seyfried. Seizure susceptible El mice (n = 57) were at least 6 months old at the time testing since older mice of the El strain are more susceptible to seizures. Seizure phenotype in response to tail suspension handling for routine husbandry was recorded for each cage of El mice after weaning in order to achieve treatment group matching. Animals had free access to drinking water at all times and were fed ad libitum chow
Seizure prophylaxis in an animal model of epilepsy
21
(ProLab3000, LabDiets, Richmond, IN, USA) prior to the study. The composition of the chow was 59.5% carbohydrate, 22% protein, 5% fat, 11% fiber, and 2.5% added minerals. Group housed mice were re-housed singly 1 week prior to testing in a colony room with a controlled temperature of 72 ◦ C, a humidity of 48%, and a 12-h reverse light cycle (lights off 1000—2200). All experiments were performed in accordance with the guidelines of the Institutional Animal Care and Use Committee of Boston College.
around the perimeter of a standard mouse housing cage. Four infrared photocell beams located along the long axis and eight beams across the width were positioned 2 cm above the bedding at 16 cm intervals. Mice continued to consume their respective experimental diets and were tested in the locomotor apparatus for 24 h following the HISS test.
Chronic fluoxetine self-administration
Mice were anesthetized with the inhaled anesthetic isoflurane (Aerrane, Baxter Healthcare Corporation, Deerfield, IL) prior to intracardial perfusion with 0.1 M phosphate-buffered saline (PBS, pH 7.4), followed by 4% paraformaldehyde in 0.9% saline. Brains were removed immediately and post-fixed overnight at 4 ◦ C. Brains were placed in a 30% sucrose solution in PBS at 4 ◦ C until sinking (48—72 h). Free-floating sections (40 m) were obtained with a cryostat, and stored in cryoprotectant (pH 7.2) at −20 ◦ C until immunohistochemical analysis.
One experimental trial was completed over a 3 day span and three trials were completed over a 7 day span. On Day 1 of each trial, all mice were weighed and singly housed in weight matched groups. Each mouse was provided with a pre-weighed food container with a chow or fluoxetine wet mash ration at the beginning of each 24 h period. An additional empty cage was also employed for each of the four trials to quantify evaporation per unit mass of wet mash and so that this constant could be deducted from total daily food intake. On subsequent days, the food containers were removed, weighed and then refilled with fresh food. Mice in a yoked condition were fed the average amount of food consumed by the fluoxetine group over the previous 24 h period. This procedure assured that non-appetitive actions of fluoxetine were responsible for anticonvulsion efficacy, not simply food restriction, since the fluoxetine group consumed less than the ad libitum chow control group. All feeding/weighing procedures were performed between 1000 and 1600 during the nocturnal portion of the circadian cycle when mice were normally awake and active.
Perfusion and tissue sectioning
Blood glucose monitoring Just prior to intracardial perfusion, a sample of cardiac blood was employed for assessing blood glucose levels at a single time point 7 days following the beginning of fluoxetine exposure and 2 h following HISS testing. The TrueTrack Smart System (Home Diagnostics, Inc. Fort Lauderdale) employed amperametric technology to quantify an oxidation reaction for real time blood glucose level determinations. This device provides a 20—600 mg/dl range of sensitivity and a 3—5% intra-assay precision when assessing blood glucose levels.
Diet preparation Nissl staining Pelleted chow was blended to create a powder to which fluoxetine (Spectrum Chemical, Gardena, CA) was added to yield a roughly 10 mg/kg/day dose of fluoxetine based upon the initial body weight and daily average food intake of experimental mice. Water was added to the dry mixture in order to create a wet mash that was placed in the bottom of a cylindrical chamber in order to avoid spillage. A ration of diet sufficient to result in leftover food the following day was provided to each animal in the ad libitum treatment groups.
Handling-induced seizure susceptibility (HISS) test On the final day of diet exposure, a seizure susceptibility test was performed in which mice were picked up by their tails and held 10—20 cm above the floor of the home cage for 30 s. Mice were then placed in a clean cage for 2 min and then lifted by the tail for another 15 s before being returned to their home cage. Behavioral characteristics of seizures in El mice are evident during three specific phases: (1) prodromal: squeaking and transient immobility, followed by ‘‘running fits’’, (2) ictal: convulsions starting with the hind limbs but rapidly becoming generalized tonic-clonic, loss of postural equilibrium, tail cocked over head (Straub tail), salivation, defecation, and urination and (3) post-ictal: lethargy, ‘‘kangaroo posture’’ (sitting on hind legs with forepaws drawn up), turning of the head in alternate directions, and occasional hyperirritability. Seizures were determined to have occurred in mice that exhibited behavioral signs such as squeaking, tonic-clonic jerking, cocked tail over the head, loss of posture and post-ictal ataxia. An all-or-none criterion was employed in detecting seizures and no attempt was made to score seizure severity.
Locomotor activity test The locomotor apparatus (San Diego Instruments) consisted of photocell grids, each measuring 25 cm × 48 cm, which were positioned
Tissue sections were mounted onto gelatinized slides, rinsed in ethanol and distilled water and then stained in 0.13% cresyl violet in an acetic acid buffer. Stained slides were rinsed in ethanol, cleared in xylene, and coverslipped.
Serotonin transporter (5-HTT) immunohistochemistry Immunohistochemical detection of 5-HTT protein was performed using a primary antibody obtained from Calbiochem (San Carlos, CA). After rinsing in 50% ethanol and quenching of endogenous peroxidase activity with a 3% hydrogen peroxide solution, tissue sections were placed in a 15% normal goat serum blocking solution. Sections were incubated with the rabbit anti-5-HTT antibody (1:1000) overnight at 4 ◦ C. On the second day, sections were incubated with a goat, anti-rabbit secondary antibody (1:1000) for 1 h and then a streptavidin-HRP conjugate (ABC kit) for 30 min (both from Vector Laboratories, San Carlos, CA). Sections were stained using a DAB chromagen and a hematoxylin counterstain. Immunohistochemical labeling was eliminated when either the primary or secondary antibody was omitted from the assay.
Brain mapping by microscopy Sections were photographed using an RT color Spot camera (Diagnostic Instruments Inc., Sterling Heights, MI) mounted on a Zeiss bright-field microscope. Nissl positive and 5-HTT immunoreactive cells were counted using the IP Lab image analysis software (Scanalytics, Fairfax, VA). In order to avoid potential experimenter and laterality biases, analyses were completed without knowledge of treatment condition using brain tissue from both sides of the sagittal plane. Sites were selected for analysis based upon a series of cFos mapping studies performed previously in El strain mice (McFadyenLeussis and Heinrichs, 2004), by coarse examination for areas
22 of intense staining by a treatment-blind observer, and by the known distribution of serotonin transporter protein in rodent brain (Cabrera-Vera and Battaglia, 1998). In particular, published electrophysiological and activity mapping studies have identified parietal cortex as a likely source of seizures in the El brain (Nakamoto et al., 1990; Ishida et al., 1993). Based upon studies in our own laboratory suggesting that immediate-early gene expression emerges in the parietal cortex of El brain early in development, the present studies tested the a priori hypothesis that this known seizure locus would by modulated by fluoxetine exposure. Moreover, sub-cortical diencephalic and basal forebrain regions are reported to constitute brain sites for seizure facilitation and interactivity with other brain seizure substrates (Eells et al., 2004). In particular, published studies demonstrate that active cells are more abundant in paraventricular thalamus after seizure induction (Mraovitch and Calando, 1999) and it has been suggested that recurring activation of the paraventricular thalamus implies that this region stores memories of previous, stress-related experiences (Fenoglio et al., 2006). As the El mouse model is known to be hyper-reactive to stressor exposure (Drage and Heinrichs, 2005), the present studies tested the a priori hypothesis that the paraventricular thalamus would be modulated by fluoxetine exposure. Finally, 5-HTT is enriched in brain areas such as amygdala and nucleus accumbens which are reciprocally connected and constitute sites vital for goal-directed responses and plasticity (Sur et al., 1996; Stevenson and Gratton, 2003). In preliminary studies, long-term fluoxetine exposure was found to selectively enhance 5-HTT expression in the nucleus accumbens (unpublished observations) and the present studies therefore tested the a priori hypothesis that these brain regions would be modulated in El mice by fluoxetine exposure. Thus, Nissl and 5-HTT labeled cells were counted in the parietal cortex, paraventricular nucleus of thalamus, nucleus accumbens, and basolateral nucleus of the amygdala of El mouse brain in the present studies. The selection of these four brain regions of interest based upon the prior research findings just described effectively reduces the likelihood that the present cell density and activation measures would reflect non-specific actions of systemically administered fluoxetine. Specific brain regions and nuclei of interest were identified using a mouse brain atlas (Paxinos and Franklin, 2001).
A. Richman, S.C. Heinrichs
Figure 1 Frequency of tail suspension-induced seizures in El mice following either three or seven consecutive 24-h periods of exposure to ad libitum chow diet, ad libitum fluoxetineadultered diet, or yoked chow diet conditions. * p < 0.05 relative to chow and yoked controls.
of dietary condition with mice eating more ad libitum chow diet than either the fluoxetine or yoked diets (Fig. 2). A significant main effect of dietary condition on body weight [F(2,12) = 4.5, p < 0.05] resulted from significant loss of body weight in the fluoxetine-treated group relative to the chow and yoked diet groups (Fig. 2). Note that body weight loss resulting from fluoxetine treatment likely reflects negative energy balance actions of the drug other than appetite suppression (Dubuc and Peterson, 1990) since mice in the yoked group ate the same quantity of food over the 3 day trial without experiencing body weight loss. Mice were immediately aware of placement of food rations in their home cages as 50, 40 and 20% of animals in the chow, fluoxetine and yoked diet groups interacted with the
Statistical analysis For all experiments, mixed factor analyses of variance (ANOVA) were performed for food intake, body weight, locomotor activity and local brain site cell density with treatment as a between subjects factor and experimental hour or day as a within subjects factor. The 3 day trial employed n = 5 mice/treatment group whereas the 7 day trial employed 12—15 mice/treatment group; a subset of n = 4 mice was selected from each group in each trial for histochemical analysis. Simple main effect analyses were conducted when appropriate to determine individual group differences, and comparisons were considered significant when p < 0.05. Seizure frequency results were examined using non-parametric Kruskal—Wallis tests.
Results Three day fluoxetine self-administration trial HISS testing after 3 days of dietary exposure produced a relatively high, and not significantly different (Kruskal—Wallis < 1, ns), rate of seizures in all treatment groups: 40% (2 of 5) in the chow group, 40% (2 of 5) in the yoked group, and 60% (3 of 5) in the fluoxetine group (Fig. 1). Analysis of food intake over the 3 day dietary trial revealed a significant [F(1,8) = 33.4, p < 0.001] main effect
Figure 2 Mean ± S.E.M. daily food intake of wet mash with evaporation taken into account (top) and body weight change (bottom) in El mice over three consecutive 24-h periods of exposure to ad libitum chow diet, ad libitum fluoxetine-adultered diet or yoked chow diet conditions.
Seizure prophylaxis in an animal model of epilepsy
23
Table 1 Cell density assessed using a Nissl stain in El mice exposed over the previous 3 days to chow, fluoxetine or yoked diets Diet condition
NAc
PVT
Amyg
Cortex
Chow Fluoxetine Yoked
569 ± 78 661 ± 93 396 ± 22
490 ± 85 523 ± 69 390 ± 64
601 ± 89 528 ± 49 489 ± 8
483 ± 31 567 ± 41 520 ± 53
Values are mean ± S.E.M. based upon n = 4 mice in each treatment group. Amyg: basolateral amygdalar nucleus, Cortex: parietal cortex, NAc: nucleus accumbens, PVT: paraventricular nucleus of thalamus.
food receptacle without delay. These results suggest that short-term fluoxetine-induced anorexia and body weight loss did not impact susceptibility to handling-induced seizures. Examination of cell density assessed by Nissl staining in nucleus accumbens, paraventricular nucleus of thalamus, amygdala and parietal cortex following a 3 day period of exposure to fluoxetine or chow diet treatments revealed no significant effects of condition (Table 1).
Seven day fluoxetine self-administration trial HISS testing after 7 days of dietary exposure revealed a relatively high rate of seizures in the two control groups: 40% (6 of 15) in the chow group and 42% (5 of 12) in the yoked group relative to a significantly (Kruskal—Wallis = 8.09, p < 0.02) decreased 0% (0 of 15) in the fluoxetine group (Fig. 1). Analysis of food intake revealed a main effect of treatment [F(2,41) = 43.8, p < 0.001] with mice in the ad libitum chow group consuming more food than mice in the fluoxetine and yoked groups (Fig. 3 ). Moreover, animals in the fluoxetine and yoked groups lost significantly [F(1,25) = 6.5, p < 0.005] more body weight than the ad libitum chow control group (Fig. 3). Mice in the three diet conditions were observed to begin eating immediately after placement of the food receptacle (chow, 65%; yoked, 44%; fluoxetine, 67%) indicating that none of the diet conditions were aversive to the extent that the animal would avoid the diets or the locations in which they were served. These results demonstrate that a 7 day diet yoking procedure was effective in matching daily food intake and body weight changes of mice not exposed to fluoxetine to those exhibited by the ad libitum fluoxetine group. Analysis of locomotor activity revealed a significant main effect of time [F(23,598) = 7.4, p < 0.001] without an effect of treatment. The effect of time reflects a circadian activity rhythm in which mice are more active during the nocturnal hours (Fig. 4). These results suggest that consumption over 7 days of a fluoxetine-adultered diet abolishes susceptibility to handling-induced seizures in El mice without disrupting the normal profile of circadian locomotor activity. At the time of sacrifice for brain harvesting, blood glucose levels were not affected by the dietary treatments administered over 7 days—–chow group: 279 ± 15 mg/dl, yoked group: 279 ± 9 mg/dl fluoxetine: 271 ± 10 mg/dl. However, cell density assessed by Nissl staining was altered in nucleus accumbens, amygdala and parietal cortex by diet condition [F(2,5) = 5.8, 8.2 and 8.6, respectively, p < 0.05] with lower densities in nucleus accumbens and amygdala in
Figure 3 Mean ± S.E.M. daily food intake of wet mash with evaporation taken into account (top) and body weight change (bottom) in El mice over seven consecutive 24-h periods of exposure to ad libitum chow diet, ad libitum fluoxetine-adultered diet or yoked chow diet conditions.
the yoked group, as well as lower densities in the parietal cortex in the fluoxetine group (p < 0.05) relative to the chow control group (Fig. 5). Examination of 5-HTT immunoreactivity in four representative brain regions revealed a significant [F(2,7) = 3.7, p < 0.05] main effect of treatment in the nucleus accumbens with 7 days of fluoxetine exposure elevating the concentration of the serotonin transporter protein relative to the chow control group (Fig. 6). These
Figure 4 Locomotor activity (mean ± S.E.M.) of El mice over a 24-h period of continued exposure to chow diet, fluoxetineadultered diet or yoked diet conditions beginning after Day 7 handling-induced seizure susceptibility testing. The white bar reflects the diurnal portion of the circadian cycle and the dark bar reflects the nocturnal portion.
24
A. Richman, S.C. Heinrichs
Figure 5 Cell density (mean ± S.E.M.) assessed using a Nissl stain in El mice exposed over the previous 7 days to chow, fluoxetine or yoked diets (top panel). Representative sections of parietal cortex are provided for mice in the chow diet (left) and fluoxetine diet (right) treatment groups. Amyg: basolateral amygdalar nucleus, Cortex: parietal cortex, NAc: nucleus accumbens, PVT: paraventricular nucleus of thalamus, SC: subarachnoid cavity, SSp: primary somatosensory area. * p < 0.05 relative to chow diet control group.
results suggest that a 7 day exposure to fluoxetine-adultered diet is sufficient to induce neural adaptations to the drug which did not include changes in resting blood glucose.
Discussion The main finding of the present studies is that a 7 day period of dietary exposure to a 10 mg/kg/day dose of fluoxetine was effective in abolishing precipitated seizures in El mice. The anticonvulsant effect of fluoxetine was time-of-exposure dependent in that a shorter 3 day trial demonstrated a high 60% seizure frequency in fluoxetine-exposed El mice which was comparable to the 40% seizure frequency of chow diet controls. Moreover, a yoked diet control group was used to assess the alternative hypothesis that efficacy of fluoxetine was dependent on diminished food intake and body weight loss. The fact that 40% of yoked group mice seized in the 7 day trial documented that non-specific consequences of diet restriction were not sufficient to suppress seizure susceptibility in El mice. Instead, the ability of fluoxetine treatment to enhance the serotonin transporter protein target for SSRI drugs in the basal forebrain and reduce cell density in a cortical region implicated in seizure etiology in El mice suggests that fluoxetine attenuated seizures by remodeling
brain seizure susceptibility circuits. This conclusion is supported by the fact that the change in cortical cell density produced by 7 day fluoxetine exposure was not reproduced by a 3 day period of fluoxetine exposure. Thus, changes produced by 7 day SSRI treatment were neuroadaptive in nature rather than simple non-specific, pharmacological effects of exposure to the drug. Future studies can test this hypothesis using local intra-cerebral administration of fluoxetine into the present candidate nuclei. The weight of scientific evidence suggests that increasing extracellular norepinephrine and/or serotonin by blockade of the reuptake of these monoamines is the pharmacodynamic action responsible for the anticonvulsant effects of antidepressants (Dailey and Naritoku, 1996; Jobe and Browning, 2005). The present results support this claim in that 1 week of exposure to a fluoxetine-adultered diet resulted in increased serotonin transporter protein content in the nucleus accumbens and a correlated elimination of handling-induced seizures in genetically susceptible El mice. While the serotonin transporter immunohistochemistry endpoint serves as an important manipulation check in the present studies to insure that dietary fluoxetine exerted direct actions on serotonergic neurotransmission, the finding of increased serotonin transporter content in the nucleus accumbens may provide insight into the mech-
Seizure prophylaxis in an animal model of epilepsy
25
Figure 6 Serotonin transporter protein (5-HTT) content (mean ± S.E.M.) assessed immunohistochemically in El mice exposed over the previous 7 days to chow, fluoxetine or yoked diets (top panel). Representative sections of nucleus accumbens are provided for mice in the chow diet (left) and fluoxetine diet (right) treatment groups. aca: anterior commisure, Amyg: basolateral amygdalar nucleus, Cortex: parietal cortex, LV: lateral ventricle, NAc: nucleus accumbens, PVT: paraventricular nucleus of thalamus. * p < 0.05 relative to chow diet control group.
anism of seizure susceptibility in the El mouse model. Brain reward circuitry, which includes the nucleus accumbens as a forebrain terminus of an affective regulatory pathway in rodents (Berridge, 1996), has been identified as one component of the seizure initiation pathway in the epileptic brain (Jobe and Browning, 2005). In complementary fashion, the suspension of seizure susceptibility in El mice via fluoxetine exposure in the present studies supports the contention that increasing serotonergic neurotransmission serves as a ‘‘exterior defense shield’’ (Jobe and Browning, 2005) against tail suspension handling-induced convulsions in El mice. Similarly, evidence from neuroimaging studies suggests that vagal nerve stimulation therapy for epilepsy acts via brainstem serotonergic and noradrenergic projections to limbic and cortical structures that are involved in mood regulation (Nemeroff et al., 2006). Given the fact that the affective dysregulation profile of El mice is robust and varied with behavioral, neural and physiological components (Drage and Heinrichs, 2005), future studies could pursue the claimed comorbidity of epilepsy and affective disorders (Nemeroff et al., 2006) using this animal model. It is well known that clinically relevant dietary restriction or provision of alternative diets exert anticonvulsive effects in animal models of epilepsy including the El mouse (Stafstrom and Bough, 2003; Mantis et al., 2004; Seyfried et
al., 2004). An autonomous anticonvulsant mechanism based entirely on nutritional change is potentially troublesome in the present studies for two reasons: (1) fluoxetine was administered via diet adulteration in the present studies such that attempts to self-regulate intake of the drug also impact food intake and vice versa and (2) fluoxetine facilitates serotonergic neurotransmission which is well known to alter appetite and body weight in multiple species (Heisler et al., 1997; Harvey and Bouwer, 2000). The oral route of fluoxetine self-administration was selected for the present studies since the El mouse phenotype of stress hyperreactivity is prohibitive for daily systemic or parenteral administration procedures of the sort typically employed for chronic administration of fluoxetine in rodents (Kecskemeti et al., 2005; Pericic et al., 2005). In particular, such drug administration procedures are stressful in and of themselves (Balcombe et al., 2004) and would be expected either to trigger seizures in El mice or to produce a seizure refractory period which would reduce HISS test sensitivity. In addition, several procedures were employed in the present studies to guard against potential non-specific appetitive actions of fluoxetine on seizure susceptibility. First, mice in the various treatment groups were weight matched prior to beginning the experimental trials. Second, a diet yoking procedure insured that this control group would exhibit the
26 same changes in food intake as fluoxetine self-administering mice. Third, a post-mortem blood glucose measure was employed to assess this bioenergetic component on the day of seizure susceptibility testing. Finally, a behavioral measure of motivation to eat was employed to assess the eagerness of mice in the various treatment groups to seek out their daily dietary ration. The home cage measure of immediate eating revealed that a significant proportion of mice in all three diet groups including the fluoxetine group were eager to come into contact with the food dish and begin eating without delay. This suggests that there were no general motivational confounds in fluoxetine-treated mice and this conclusion was supported by the lack of effect of fluoxetine on overall home cage motor activity beginning at the time of seizure susceptibility testing. Finally, single time point resting blood glucose levels were equivalent among the three dietary groups at the time of seizure susceptibility testing. Taken together with the lack of prophylaxis observed for seizure susceptibility in the yoked diet group, these results suggest that non-specific bioenergetic variables are not able to account for anticonvulsant efficacy of fluoxetine in the present studies. The present results add the El mouse to the growing list of animal models of epilepsy in which fluoxetine has been found to exert anticonvulsant actions (Dailey et al., 1996; Macedo et al., 2004; Ugale et al., 2004; Kecskemeti et al., 2005; Pericic et al., 2005). One mechanism for reduced seizure frequency in El mice treated with fluoxetine may be remodeling of the somatosensory region of the parietal cortex which has been identified previously as a seizure trigger zone (Ishida et al., 1993; Murashima et al., 1996). However, the present results also support an alternative hypothesis that protective serotonergic circuits in affective regulatory regions of the brain serve as seizure modulators. Consistent with the later alternative, high emotionality characteristics of the El mouse separate from seizure susceptibility also suggest efficacy of fluoxetine. For example, El mice are hyper-reactive to stressor exposure (Drage and Heinrichs, 2005), exhibit an anxiogenic-like phenotype in response to novelty (Pascual and Heinrichs, in press), and these characteristics could potentially be ameliorated by the anxiolytic efficacy of fluoxetine (Silva and Brandao, 2000). Similarly, El mice exhibit social withdrawal (Turner et al., in press) and an impairment in social recognition memory (Lim et al., in press) and these behavioral and cognitive deficiencies are reported to be normalized by fluoxetine (El Hage et al., 2004). Future studies can address whether seizure protective effects of fluoxetine in El mice extend into other realms of affective or anti-social psychopathology.
Acknowledgements We thank Chen Lim, Jennifer Pascual and Laura Turner for their help with tissue harvesting. This research was supported by C.U.R.E. and a Research Incentive Grant from Boston College to SCH.
References Balcombe, J.P., Barnard, N.D., et al., 2004. Laboratory routines cause animal stress. Contemp. Top Lab. Anim. Sci. 43 (6), 42—51.
A. Richman, S.C. Heinrichs Berridge, K.C., 1996. Food reward: brain substrates of wanting and liking. Neurosci. Biobehav. Rev. 20 (1), 1—25. Cabrera-Vera, T.M., Battaglia, G., 1998. Prenatal exposure to fluoxetine (Prozac) produces site-specific and age-dependent alterations in brain serotonin transporters in rat progeny: evidence from autoradiographic studies. J. Pharmacol. Exp. Ther. 286 (3), 1474—1481. Compan, V., Zhou, M., et al., 2004. Attenuated response to stress and novelty and hypersensitivity to seizures in 5-HT4 receptor knock-out mice. J. Neurosci. 24 (2), 412—419. Dailey, J.W., Mishra, P.K., et al., 1992. Serotonergic abnormalities in the central nervous system of seizure-naive genetically epilepsyprone rats. Life Sci. 50 (4), 319—326. Dailey, J.W., Naritoku, D.K., 1996. Antidepressants and seizures: clinical anecdotes overshadow neuroscience. Biochem. Pharmacol. 52 (9), 1323—1329. Dailey, J.W., Yan, Q.S., et al., 1996. Neurochemical correlates of antiepileptic drugs in the genetically epilepsy-prone rat (GEPR). Life Sci. 58 (4), 259—266. Drage, M.G., Heinrichs, S.C., 2005. Phenotyping the untouchables: environmental enhancement of behavioral and physiological activation in seizure-prone El mice. Epilepsy Behav. 6 (1), 35—42. Dubuc, P.U., Peterson, C.M., 1990. Ineffectiveness of parenteral fluoxetine or RU-486 to alter long-term food intake, body weight or body composition of genetically obese mice. J. Pharmacol. Exp. Ther. 255 (3), 976—979. Eells, J.B., Clough, R.W., et al., 2004. Comparative fos immunoreactivity in the brain after forebrain, brainstem, or combined seizures induced by electroshock, pentylenetetrazol, focally induced and audiogenic seizures in rats. Neuroscience 123 (1), 279—292. El Hage, W., Peronny, S., et al., 2004. Impaired memory following predatory stress in mice is improved by fluoxetine. Prog. Neuropsychopharmacol. Biol. Psychiatry 28 (1), 123—128. Fenoglio, K.A., Chen, Y., et al., 2006. Neuroplasticity of the hypothalamic—pituitary—adrenal axis early in life requires recurrent recruitment of stress-regulating brain regions. J. Neurosci. 26 (9), 2434—2442. Ferrero, A.J., Cereseto, M., et al., 2005. Chronic treatment with fluoxetine decreases seizure threshold in naive but not in rats exposed to the learned helplessness paradigm: Correlation with the hippocampal glutamate release. Prog. Neuropsychopharmacol. Biol. Psychiatry 29 (5), 678—686. Gomez, R., Huber, J., et al., 2001. Acute effect of different antidepressants on glycemia in diabetic and non-diabetic rats. Braz. J. Med. Biol. Res. 34 (1), 57—64. Harvey, B.H., Bouwer, C.D., 2000. Neuropharmacology of paradoxic weight gain with selective serotonin reuptake inhibitors. Clin. Neuropharmacol. 23 (2), 90—97. Hashiguchi, W., Nagatomo, I., et al., 2001. Influences of caffeine to nitric oxide production and zonisamide concentration in the brain of seizure-susceptible EL mice. Psychiatry Clin. Neurosci. 55 (4), 319—324. Heinrichs, S.C., Seyfried, T.N., 2006. Behavioral seizure correlates in animal models of epilepsy: a road map for assay selection, data interpretation, and the search for causal mechanisms. Epilepsy Behav. 8 (1), 5—38. Heisler, L.K., Kanarek, R.B., et al., 1997. Fluoxetine decreases fat and protein intakes but not carbohydrate intake in male rats. Pharmacol. Biochem. Behav. 58 (3), 767—773. Hernandez, E.J., Williams, P.A., et al., 2002. Effects of fluoxetine and TFMPP on spontaneous seizures in rats with pilocarpineinduced epilepsy. Epilepsia 43 (11), 1337—1345. Hiramatsu, M., 1981. Brain monoamine levels and E1 mouse convulsions. Folia Psychiatr. Neurol. Jpn. 35 (3), 261—266. Ishida, N., Kasamo, K., et al., 1993. Epileptic seizure of El mouse initiates at the parietal cortex: depth EEG observation in freely
Seizure prophylaxis in an animal model of epilepsy moving condition using buffer amplifier. Brain Res. 608 (1), 52—57. Jobe, P.C., Browning, R.A., 2005. The serotonergic and noradrenergic effects of antidepressant drugs are anticonvulsant, not proconvulsant. Epilepsy Behav. 7 (4), 602—619. Kecskemeti, V., Rusznak, Z., et al., 2005. Norfluoxetine and fluoxetine have similar anticonvulsant and Ca2+ channel blocking potencies. Brain Res. Bull. 67 (1/2), 126—132. King, J.T., LaMotte, C.C., 1989. El mouse as a model of focal epilepsy: a review. Epilepsia 30 (3), 257—265. Kordik, C.P., Reitz, A.B., 1999. Pharmacological treatment of obesity: therapeutic strategies. J. Med. Chem. 42 (2), 181—201. Leussis, M.P., Heinrichs, S.C., in press. Temporal ontogeny of seizure circuit activation prior to the onset of seizure susceptibility in EL mice. Neuroscience. Lim, C.E., Turner, L.H., et al., in press. Short-term social recognition memory deficit and altered social and physiological stressor reactivity in seizure-susceptible El mice. Seizure. Lopez-Meraz, M.L., Gonzalez-Trujano, M.E., et al., 2005. 5-HT1A receptor agonists modify epileptic seizures in three experimental models in rats. Neuropharmacology 49 (3), 367—375. Macedo, D.S., Santos, R.S., et al., 2004. Effect of anxiolytic, antidepressant, and antipsychotic drugs on cocaine-induced seizures and mortality. Epilepsy Behav. 5 (6), 852—856. Mantis, J.G., Centeno, N.A., et al., 2004. Management of multifactorial idiopathic epilepsy in EL mice with caloric restriction and the ketogenic diet: role of glucose and ketone bodies. Nutr. Metab. (Lond.) 1 (1), 11. McFadyen-Leussis, M., Heinrichs, S.C., 2004. Handling of EL/Suz mice elicits differential fos activation in the locus ceruleus during development. In: Society for Neuroscience Annual Meeting, San Diego, California. McFadyen-Leussis, M.P., Heinrichs, S.C., 2005. Seizure-prone EL/Suz mice exhibit physical and motor delays and heightened locomotor activity in response to novelty during development. Epilepsy Behav. 6 (3), 312—319. Mraovitch, S., Calando, Y., 1999. Interactions between limbic, thalamo-striatal-cortical, and central autonomic pathways during epileptic seizure progression. J. Comp. Neurol. 411 (1), 145—161. Murashima, Y.L., Kassamo, K., et al., 1996. Developmental and seizure-related regional differences in immediate early gene expression and GABAergic abnormalities in the brain of EL mice. Epilepsy Res. 26 (1), 3—14. Murashima, Y.L., Suzuki, J., et al., 2004. In: Benjamin, S.M. (Ed.), Ictogenesis and Epileptogenesis in Epileptic Mutant EL Mice. Focus on Epilepsy Research. Nova Science Publishers, Inc., New York, pp. 139—198. Murashima, Y.L., Yoshii, M., et al., 2002. Ictogenesis and epileptogenesis in EL mice. Epilepsia 43 (Suppl. 5), 130—135. Mutoh, K., Ito, M., et al., 1993. Depth EEG in mutant epileptic E1 mice: demonstration of secondary generalization of the seizure from the hippocampus. Electroencephalogr. Clin. Neurophysiol. 86 (3), 205—212. Nakamoto, Y., Nakayama, S., et al., 1990. Cerebral uptake of [14C]deoxyglucose during the entire seizure and the recovery period in an El mouse. Epilepsy Res. 5 (1), 43—48. Nemeroff, C.B., Mayberg, H.S., et al., 2006. VNS Therapy in treatment-resistant depression: clinical evidence and putative neurobiological mechanisms. Neuropsychopharmacology. 31 (7), 1345—1355.
27 Pascual, J., Heinrichs, S.C., in press. Olfactory neophobia and seizure susceptibility phenotypes in an animal model of epilepsy are normalized by impairment of brain corticotropin releasing factor. Epilepsia. Paxinos, G., Franklin, K.B.J., 2001. The Mouse Brain in Stereotaxic Coordinates. Academic Press, San Diego. Pericic, D., Lazic, J., et al., 2005. Anticonvulsant effects of acute and repeated fluoxetine treatment in unstressed and stressed mice. Brain Res. 1033 (1), 90—95. Rho, J.M., Kim, D.W., et al., 1999. Age-dependent differences in flurothyl seizure sensitivity in mice treated with a ketogenic diet. Epilepsy Res. 37 (3), 233—240. Sarkisian, M.R., 2001. Overview of the current animal models for human seizure and epileptic disorders. Epilepsy Behav. 2 (3), 201—216. Schatzberg, A.F., 2000. New indications for antidepressants. J. Clin. Psychiatry 61 (Suppl. 2), 9—17. Schridde, U., van Luijtelaar, G., 2005. The role of the environment on the development of spike-wave discharges in two strains of rats. Physiol. Behav. 84 (3), 379—386. Seyfried, T.N., Glaser, G.H., 1985. A review of mouse mutants as genetic models of epilepsy. Epilepsia 26 (2), 143—150. Seyfried, T.N., Greene, A.E., et al., 2004. Caloric restriction and epilepsy: historical perspectives, relationship to the ketogenic diet, and analysis in epileptic EL mice. In: Stafstrom, C.E., Rho, J.M. (Eds.), Epilepsy and the Ketogenic Diet. Humana Press, pp. 247—264. Silva, R.C.B., Brandao, M.L., 2000. Acute and chronic effects of gepirone and fluoxetine in rats tested in the elevated plusmaze: an ethological analysis. Pharmacol. Biochem. Behav. 65 (2), 209—216. Stafstrom, C.E., Bough, K.J., 2003. The ketogenic diet for the treatment of epilepsy: a challenge for nutritional neuroscientists. Nutr. Neurosci. 6 (2), 67—79. Stevenson, C.W., Gratton, A., 2003. Basolateral amygdala modulation of the nucleus accumbens dopamine response to stress: role of the medial prefrontal cortex. Eur. J. Neurosci. 17 (6), 1287—1295. Sur, C., Betz, H., et al., 1996. Immunocytochemical detection of the serotonin transporter in rat brain. Neuroscience 73 (1), 217— 231. Suzuki, J., Nakamoto, Y., 1982. Abnormal plastic phenomena of sensory-precipitated epilepsy in the mutant El mouse. Exp. Neurol. 75 (2), 440—452. Suzuki, J., Nakamoto, Y., et al., 1983. Local cerebral glucose utilization in epileptic seizures of the mutant El mouse. Brain Res. 266 (2), 359—363. Tecott, L.H., Sun, L.M., et al., 1995. Eating disorder and epilepsy in mice lacking 5-HT2c serotonin receptors. Nature 374 (6522), 542—546. Tupal, S., Faingold, C.L., 2006. Evidence supporting a role of serotonin in modulation of sudden death induced by seizures in DBA/2 mice. Epilepsia 47 (1), 21—26. Turner, L.H., Lim, C.E., et al., in press. Anti-social and seizure susceptibility phenotypes in an animal model of epilepsy are normalized by impairment of brain corticotropin releasing factor. Epilepsy Behav. Ugale, R.R., Mittal, N., et al., 2004. Essentiality of central GABAergic neuroactive steroid allopregnanolone for anticonvulsant action of fluoxetine against pentylenetetrazole-induced seizures in mice. Brain Res. 1023 (1), 102—111.