Effects of vigabatrin treatment on status epilepticus-induced neuronal damage and mossy fiber sprouting in the rat hippocampus

Epilepsy Research 33 (1999) 67 – 85

Effects of vigabatrin treatment on status epilepticus-induced neuronal damage and mossy fiber sprouting in the rat hippocampus Asla Pitka¨nen *, Jari Nissinen, Esa Jolkkonen, Jarkko Tuunanen, Toivo Halonen A.I. Virtanen Institute, Uni6ersity of Kuopio, P.O. Box 1627, FIN-70 211 Kuopio, Finland Received 6 May 1998; received in revised form 16 June 1998; accepted 17 June 1998

Abstract Selective neuronal damage and mossy fiber sprouting may underlie epileptogenesis and spontaneous seizure generation in the epileptic hippocampus. It may be beneficial to prevent their development after cerebral insults that are known to be associated with a high risk of epilepsy later in life in humans. In the present study, we investigated whether chronic treatment with an anticonvulsant, vigabatrin (g-vinyl GABA), would prevent the damage to hilar neurons and the development of mossy fiber sprouting. Vigabatrin treatment was started either 1 h, or 2 or 7 days after the beginning of kainic acid-induced (9 mg/kg, i.p.) status epilepticus and continued via subcutaneous osmotic minipumps for 2 months (75 mg/kg per day). Thereafter, rats were perfused for histological analyses. One series of horizontal sections was stained with thionine to estimate the total number of hilar neurons by unbiased stereology. One series was prepared for somatostatin immunohistochemistry and another for Timm histochemistry to detect mossy fiber sprouting. Our data show that vigabatrin treatment did not prevent the decrease in the total number of hilar cells, nor the decrease in hilar somatostatin-immunoreactive (SOM-ir) neurons when SOM-ir neuronal numbers were averaged from all septotemporal levels. However, when vigabatrin was administered 2 days after the onset of status epilepticus, we found a mild neuroprotective effect on SOM-ir neurons in the septal end of the hippocampus (92% SOM-ir neurons remaining; PB0.05 compared to the vehicle group). Vigabatrin did not prevent mossy fiber sprouting regardless of when treatment was started. Rather, sprouting actually increased in the septal end of the hippocampus when vigabatrin treatment began 1 h after the onset of status epilepticus (P B0.05 compared to the vehicle group). Our data show that chronic elevation of brain GABA levels after status epilepticus does not have any substantial effects on neuronal loss or mossy fiber sprouting in the rat hippocampus. © 1999 Elsevier Science B.V. All rights reserved. Keywords: Anticonvulsant; Epilepsy; Kainic acid; Neuroprotection; Plasticity; Somatostatin; Stereological cell counting; g-Vinyl-GABA

* Corresponding author. Tel: +358 17 163296; fax: + 358 17 163025; e-mail: [email protected] 0920-1211/99/$ - see front matter © 1999 Elsevier Science B.V. All rights reserved. PII S0920-1211(98)00074-6

68

A. Pitka¨nen et al. / Epilepsy Research 33 (1999) 67–85

1. Introduction Hippocampal damage is found in approximately 60 – 70% of patients with temporal lobe epilepsy undergoing surgery due to drug-refractory seizures (Babb and Pretorius, 1993). The neuronal damage includes the loss of pyramidal cells in the hippocampus proper, the loss of granule cells in the dentate gyrus, and the loss of various populations of hilar neurons, including mossy cells and neurons containing somatostatin or neuropeptide Y (Babb and Pretorius, 1993). Damage to hippocampal principal cells is often associated with memory impairment both in animals (Zola-Morgan et al., 1992; Moser et al., 1993) and humans (Zola-Morgan et al., 1986), whereas the loss of hilar cells may be the underlying cause for the lowered seizure threshold in the dentate gyrus (Sloviter, 1994). In addition to this neuronal damage, axons of the remaining neurons sprout both in experimental models as well as in the human epileptic hippocampus (Tauck and Nadler, 1985; Sutula et al., 1989; Houser et al., 1990). This has been best characterized in the dentate gyrus where granule cell axons, that is, mossy fibers, sprout into the inner molecular layer. The functional significance of axonal sprouting has remained controversial. Some neuroanatomical studies, however, show both in the rat (Represa et al., 1993; Okazaki et al., 1995; Buckmaster et al., 1996) and human (Zhang and Houser, 1995) epileptic hippocampus that the sprouted mossy fibers within the inner molecular layer make synaptic contacts preferentially with the proximal dendrites of granule cells. Moreover, in a recent electrophysiological study, Wuarin and Dudek (1996) found increased granule cell excitability in the hippocampus associated with mossy fiber sprouting. These data suggest that mossy fiber sprouting may facilitate the development of excitatory epileptogenic circuitry in the hippocampus. Alternatively, it has been suggested that sprouting mossy fibers, which make contacts with inhibitory neurons, may be involved in diminishing the spread of excitation (Sloviter, 1992). Recent studies on experimental status epilepticus indicate that, in addition to acute neuronal damage, there is delayed neuronal damage that takes place over several days after the initial insult

(Filipkowski et al., 1994; Pollard et al., 1994a,b). Moreover, mossy fiber sprouting is progressive, continuing for weeks or months after the initial insult. These data raise the question whether the delayed neuronal damage and progressive mossy fiber sprouting after the initial insult could be diminished or prevented by pharmacotherapy. Such intervention would enable us to further efforts in preventing the epileptogenesis that is associated with various human conditions, such as encephalitis, stroke, head trauma, and prolonged complex febrile seizures. So far, such attempts with antiepileptic drugs have been unsuccessful (Hernandez, 1997). Vigabatrin (g-vinyl GABA) is an antiepileptic drug that elevates brain GABA levels several-fold by irreversibly inhibiting the GABA-metabolizing enzyme, GABA-transaminase (Grant and Heel, 1991). Our previous studies showed that if we administer vigabatrin to rats before inducing status epilepticus by electrical stimulation of the perforant pathway (Ylinen et al., 1991; Pitka¨nen et al., 1996) or by kainic acid treatment (Halonen et al., 1995), we could largely prevent subsequent hippocampal neuronal damage. In the present study, we address the question: can epileptogenic cell loss and progressive mossy fiber sprouting caused by status epilepticus be prevented with vigabatrin treatment that is started after the onset of status epilepticus?

2. Methods

2.1. Animals Male Han–Wistar rats (n= 69) (300–360 g) were used in this study. The rats were housed in individual cages at a temperature of 2091°C, with humidity maintained at 50–60% and lights on from 07:00 to 19:00 h. Standard food pellets and water were freely available.

2.2. Status epilepticus induced by kainic acid with subsequent administration of 6igabatrin To determine the potential effects of vigabatrin treatment on kainic acid-induced status epilepti-

A. Pitka¨nen et al. / Epilepsy Research 33 (1999) 67–85

cus, 39 rats were injected intraperitoneally with kainic acid (KA, 9 mg/kg in 0.9% NaCl; Sigma K-0250). The control group received 0.9% NaCl. To ascertain the development of status epilepticus, the rats were observed for at least 4 h after kainic acid injection. Only the rats that developed status epilepticus (i.e. recurrent seizures with bilateral forelimb clonus following kainate injection) and survived for 2 months thereafter were included in the subsequent analyses. Vigabatrin (VGB) treatment was started either 1 h, or 2 or 7 days after the onset of status epilepticus (i.e. the time when the rat had the first kainate-induced bilateral forelimb clonus).

69

The treatment groups included in the study are described in detail in Fig. 1. The reason to administer vigabatrin intraperitoneally during the first 2 days to the rats in the KA+ VGB 1-h group was to avoid the use of anesthesia early after the induction of status epilepticus. Starting on day 2 after the beginning of status epilepticus, vigabatrin was administered via osmotic pumps. This chronic administration used a vigabatrin dose (75 mg/kg per day) that in previous studies was reported to prevent the appearance of seizures in experimental models of epilepsy (for review, see Grant and Heel (1991)) without causing vacuolization in rat brain (Gibson et al., 1990).

Fig. 1. Schematic diagram of the study design. Six experimental groups were included in the study. The number of rats in each group is in parentheses.

70

A. Pitka¨nen et al. / Epilepsy Research 33 (1999) 67–85

Alzet 2ML1 osmotic pumps were filled with vigabatrin dissolved in 0.9% NaC1 (pumping rate 10.61 ml/h, information provided by manufacturer). Two days after the beginning of status epilepticus, the animals were anesthetized for 10 – 20 min with an intraperitoneal injection of an anesthetic cocktail (40 mg/kg sodium pentobarbital and 60 mg/kg chloral hydrate) and the pumps were implanted subcutaneously in the back slightly posterior to the scapulae according to the protocol of the manufacturer. The pumps were replaced under aseptic conditions with new pumps every 2 weeks until the end of the study. After each pump removal, the volume of vigabatrin solution left in the pump was measured to confirm that the solution had been delivered from the pump to the animal.

2.3. Video-EEG monitoring of kainate-induced seizures: effects of 6igabatrin administration Another group of rats (n = 18) was monitored by video-EEG continuously for 24 h following kainate injection to measure the severity and duration of status epilepticus. Initially, the animals were anesthetized with a mixture of sodium pentobarbital (60 mg/kg) and chloral hydrate (100 mg/kg) and placed in a Kopf stereotaxic frame. Epidural screw electrodes were positioned bilaterally into the skull above the frontal cortex (3.0 mm anterior and 2.5 mm lateral to bregma) and the parietal cortex (4.1 mm posterior and 2.6 mm lateral to bregma). In addition, reference and ground electrodes were positioned in the skull above the cerebellum. The electrodes were fixed with dental acrylate. The rats were allowed to recover for 10–12 days before intraperitoneal kainic acid injection (9 mg/kg). The duration and severity of status epilepticus was monitored using a video-digital EEG system (Nervus EEG Recording System, Tautagreining, Iceland; ISO-1032 Amplifier, Braintronics, The Netherlands; SVTS3000P Time Lapse 168 video recorder, Hitachi, Japan; Panasonic WV-CL350 Video Camera, Panasonic, Japan). Ten of the rats received vigabatrin (500 mg/kg, intraperitoneally) 1 h after the first bilateral forelimb clonus (i.e. the onset of status epilepticus). The other eight control rats

were treated with vehicle (0.9% NaCl) instead of vigabatrin.

2.4. Determination of serum 6igabatrin le6els and hippocampal GABA le6els In order to confirm the constant release of vigabatrin from the minipumps and the elevation of GABA levels in the brain, we measured serum levels of vigabatrin and GABA concentrations in the hippocampus at various times during vigabatrin treatment. Therefore, a separate group of rats (n= 12) was anesthetized and Alzet 2ML1 osmotic pumps, filled either with vigabatrin (250 mg/ml resulting in a daily dose of 75 mg/kg, n=9) or vehicle (0.9% NaCl, n= 3), were implanted subcutaneously in the back slightly posterior to the scapulae according to the protocol of the manufacturer. In order to measure vigabatrin and GABA levels, the rats were decapitated 2 days, or 1 or 2 weeks after pump implantation. Vigabatrin concentrations in the serum were measured by high-performance liquid chromatography (HPLC) as previously described by Halonen et al. (1990). GABA levels in the hippocampus were measured by HPLC as described by Valtonen et al. (1995).

2.5. Histological processing of brain tissue 2.5.1. Fixation Rats were perfusion-fixed transcardially 2 months after kainic acid-induced status epilepticus. Rats were first perfused with buffered 0.37% sulfide solution (300 ml, 30 ml/min) followed by 4% paraformaldehyde in 0.1 M phosphate buffer, pH 7.4 (300 ml, 30 ml/min) (Sloviter, 1982). The brains were removed from the skull, postfixed in buffered paraformaldehyde for 4 h and placed in 20% glycerol in 0.02 M potassium phosphatebuffered saline, pH 7.4 (KPBS) for 36 h. The brains were then blocked, frozen in dry ice, and stored at − 70°C until cut. They were sectioned in the horizontal plane (1-in-5 series) at a thickness of 30 mm with a sliding microtome. The sections were stored in a cryoprotectant tissue-collecting solution (TCS) (30% ethylene glycol, 25% glycerol in 0.05 M sodium phosphate buffer) at −20°C

A. Pitka¨nen et al. / Epilepsy Research 33 (1999) 67–85

until processed. Adjacent series of sections were used for either Nissl staining, somatostatin immunohistochemistry, or Timm staining.

2.5.2. Immunohistological staining for somatostatin To count the number of SOM-ir neurons, a 1-in-5 series of free-floating horizontal sections throughout the septotemporal axis of the hippocampus was collected from TCS and stained using the avidin–biotin technique for somatostatin immunohistochemistry as described previously (Pitka¨nen et al., 1996), except that the primary antibody dilution was 1:250. 2.5.3. Timm histochemistry Synaptic reorganization was analyzed from sections stained using the Timm sulfide/silver method (Sloviter, 1982). This staining recognizes mossy fibers due to their high zinc content. For staining, the horizontal sections (1-in-5 series) along the whole septotemporal axis of the hippocampus were mounted on gelatin-coated slides and dried at 37°C. The sections were developed in a darkroom in a freshly made solution of gum arabic (300 g/l), sodium citrate buffer (25.5 g/l citric acid monohydrate and 23.5 g/l sodium citrate), hydroquinone (17.04 g/l), and silver nitrate (850 mg/l) until an appropriate staining intensity was achieved (60–75 min). After the sections had been developed, they were rinsed in water (30 min) and placed in 5% sodium thiosulfate solution for 12 min. Finally, they were dehydrated through an ascending series of ethanol, cleared in xylenes, and coverslipped with Depex. 2.5.4. Other staining An adjacent series of sections was stained with thionine to identify the cytoarchitectonic boundaries of the relevant brain regions and to estimate the total number of hilar neurons using unbiased stereology. 2.6. Analysis of sections The sections were analyzed light microscopically with both brightfield and darkfield optics. All analyses were performed in a blind manner.

71

2.6.1. Estimation of neuronal numbers in the hilus using stereology Stereological analyses were conducted blindly with respect to the treatment status of the animal. The optical fractionator method was implemented using Stereo Investigator software in a Neuro Lucida morphometry system (MicroBrightField, USA) with guidelines described by West et al. (1991). Neuronal counts were derived from a minimum of eight thionine-stained sections (median 10), spaced at 450-mm intervals through the entire septotemporal extent of the left hippocampus from each brain. A color video camera (Hitachi HV-C20, Japan), interfaced with an Olympus BX50 microscope, was used to view sections on a high-resolution monitor, and neuroanatomical borders of the hilus were digitized under lowpower magnification. Subsequent cell counting was confined within these borders. The sections were inspected according to a systematic random sampling scheme such that counts were derived from a known and representative fraction of the hilus. Specifically, the motorized stage of the microscope system was under computer control, and the hilar fields in every histological section were surveyed at evenly spaced x–y intervals of 180× 180 mm. For each x–y step, cell counts were derived from a known fraction of the total area by using an unbiased counting frame that was 36×36 mm. Counting was performed throughout the section avoiding the neurons that were in focus at the surface of the section. Neuronal nuclei were counted only as they first came into focus within each optical dissector. Glia, identified by size and cytological characteristics, were excluded from the counts. Using these sampling parameters, the mean number of optical dissectors examined in sham-operated animals was 132 and the mean number of cells counted per hippocampus was 144. Finally, the total neuron number was estimated by multiplying the sum of the neurons counted by the reciprocal of the fraction of the hippocampus that was sampled (i.e. a multiple of the fraction of the histological sections examined, the fraction of the x–y step interval covered by the counting frame, and the fraction of the total section thickness examined). To provide a standardized statistic for evaluating the

72

A. Pitka¨nen et al. / Epilepsy Research 33 (1999) 67–85

precision of our neuronal counts, we determined the coefficient of error to be 9.4%.

2.6.2. Calculation of the number of hilar SOM-ir neurons The number of hilar SOM-ir neurons in the hilus of the left hippocampus was calculated manually at a magnification of × 200 with the aid of an ocular grid (Pitka¨nen et al., 1996). Only immunostained neurons with at least one dendrite emanating from the soma were counted. Due to status epilepticus-induced alterations in the area of the dentate hilus, the cells were counted as the number of SOM-ir cells per hilus. The number of SOM-ir neurons in the hilus were first expressed as the mean number of SOM-ir neurons per section averaged across all septotemporal levels of the hilus. Furthermore, since the number of SOM-ir neurons varies at different septotemporal levels of the hilus, the sections were divided into five equivalent septotemporal sublevels (levels I – V), level I (3.86 mm from the bregma) being the most septal and level V (6.82 mm ventral from the bregma) being the most temporal (each level contained three to five sections) (Jolkkonen et al., 1997). Therefore, we also expressed hilar SOM-ir neuron counts as a mean number of SOM-ir neurons per section at each of five different septotemporal levels, calculated by averaging the cell counts in all sections within each level. The same levels were used to estimate the density of mossy fiber sprouting (see below). 2.6.3. Scoring of Timm staining Sprouting of mossy fibers was evaluated from Timm-stained horizontal sections of the left hippocampus adjacent to those used for somatostatin immunohistochemistry. Since systemic administration of kainic acid induces more severe sprouting than does electrical kindling, the original rating scale described by Cavazos et al. (1991) for the kindling model was extended from 5 to 6 in order to include the most extensive sprouting into the scale (Jolkkonen et al., 1997). The density of Timm granules was estimated in the supragranular region and in the

inner molecular layer of the dentate gyrus. The score was estimated from the middle third of the suprapyramidal and infrapyramidal blades throughout the septotemporal axis of the hippocampus. The density of mossy fiber sprouting was first expressed as the mean Timm score counted throughout the hippocampus. In addition, the mean Timm score was determined at each of the five septotemporal levels, calculated by averaging the Timm scores in all sections within each level. Photomicrographs were taken with a Leica DM RD camera system.

2.7. Statistical analysis We previously showed that neither the number of SOM-ir neurons in the dentate hilus nor the Timm staining differed between the left and right hemispheres in rats treated with a systemic injection of kainic acid (Jolkkonen et al., 1997). Thus, the results (mean9 standard deviation) were calculated from the left hippocampus. The percentage of damage to all hilar neurons or to hilar SOM-ir neurons was calculated as [1− (number of neurons remaining in each case/ mean number of neurons in the control group)]× 100 (see Fig. 3A,B). Overall group differences in the mean number of hilar cells, mean number of SOM-ir neurons, mean Timm score, and EEG parameters were initially analyzed by the Kruskall–Wallis test. If a significant effect was found, the Mann–Whitney U-test was then used for comparisons between experimental groups. The effect of treatments on the number of SOM-ir neurons and Timm scores at different septotemporal levels of the hippocampus was analyzed by the Kruskall–Wallis test (to compare the difference at each septotemporal level between the groups) or by analysis of variance (ANOVA, to compare the difference over the septotemporal axis of the hippocampus between the groups). Correlations between the cell counts or cell counts and mossy fiber sprouting were tested using linear regression analysis and ANOVA. The level of significance was set at PB 0.05.

A. Pitka¨nen et al. / Epilepsy Research 33 (1999) 67–85

73

Fig. 2. (A) Total number of hilar neurons estimated by unbiased stereology in different treatment groups (mean 9S.D.). (B) The mean number of somatostatin-immunoreactive (SOM-ir) neurons averaged across all septotemporal levels of the hilus in different treatment groups (mean 9S.D.). (C) The median density of mossy fiber sprouting (Timm score) averaged across all septotemporal levels in different treatment groups. Percentages in panels (A) and (B) refer to the percentage of neurons remaining compared to the control group. Statistical significance: *P B 0.05, **PB 0.01 (compared to the KA+ vehicle group), cP B0.05, c c PB0.01 (compared to the control group) (Mann–Whitney U-test).

A. Pitka¨nen et al. / Epilepsy Research 33 (1999) 67–85

74

Table 1 Appearance and duration of behavioral and electroencephalographic seizure activity after kainic acid administration in vehicle- and vigabatrin (KA+VGB 1-h group)-treated animals Behavioral/EEG parameter

KA+vehicle

First signs of epileptic 16.8 98.9 (8) activity in EEGa First behavioral partial 44.49 16.7 (8) seizurea First behavioral gener- 70.99 10.7 (8) alized seizurea Continuous epileptic 52.99 12.5 (8) spiking activity starteda Last HAFDa 5159 226 (7) End of epileptic spiking \24 h (7) activity in EEG Number of animals dy- 1/8 ing from status epilepticus Suppression of EEG background activity Started after vigaba- — trin injectiona Duration of suppres- — sion (min)

KA+VGB 1-h

16.4 9 5.5 (10) 45.8922.6 (10) 70.99 15.2 (10) 57.8 9 13.7 (10)

6309 252 (8) \24 h (8) 2/10

4649118 [310–658] 391 9 282 [90–760]

In KA+VGB 1-h group, vigabatrin (500 mg/kg, i.p.) was started 1 h after appearance of the first generalized seizure. In KA+vehicle group, rats received vehicle. Abbreviations: HAFD, high-amplitude and high-frequency discharges; EEG, electroencephalogram. Values are expressed as mean 9S.D. of the mean. Number of animals is in parentheses. Range of values is in brackets. a Minutes after kainate injection. There were no statistically significant differences in any of the parameters between the groups (Mann–Whitney U-test).

3. Results

3.1. Appearance of beha6ioral and electrographic seizures after kainate injection The development of seizure activity after intraperitoneal kainic acid injection was similar to that described previously (for review, see Sperk, 1994). As summarized in Table 1, the first signs of electrographic and behavioral seizure activity appeared about 17 and 44 min after kainate injection, respectively. The first generalized behav-

ioral seizure appeared about 71 min after the KA injection, and always occurred after the generation of continuous epileptic spiking activity, which is considered to be a marker for the onset of status epilepticus. Recurrent HAFDs (highamplitude and frequency discharges), which were typically associated with clonic partial or generalized seizures, disappeared about 9 h after KA injection. Thereafter, rhythmic discrete spiking activity continued as the amplitude and frequency decreased. Low amplitude spiking activity continued for more than 24 h after KA injection. Only those rats that developed generalized status epilepticus were included in the histological analysis.

3.2. Effect of 6igabatrin administration on electrographic seizure acti6ity In vigabatrin-treated animals, the first signs of electrographic and behavioral seizure activity appeared about 16 and 46 min after kainate injection, respectively. The first generalized behavioral seizure occurred at the same time as in vehicle-treated controls (about 71 min after the KA injection). Therefore, the generation of status epilepticus and the behavioral and EEG seizure activity at the time of vigabatrin injection were similar to that in the vehicle control group (Table 1). Vigabatrin injection did not change the EEG pattern immediately. However, 5–11 h after vigabatrin administration, background EEG activity became suppressed (decreased amplitude) in all animals and the suppression lasted for 1.5– 12 h. In two of the rats, cortical EEG activity was completely dampened (flat) for 4 and 23 min. Interestingly, in most (six of eight) of the vigabatrin-treated rats, the last HAFDs emerged during the strongly suppressed EEG background activity; however, the timing of their appearance did not differ between vehicle- and vigabatrintreated rats. Following the last HAFD, the frequency and amplitude of epileptic spiking that continued over the next 24 h were lower in vigabatrin-treated rats compared to vehicle-treated rats.

A. Pitka¨nen et al. / Epilepsy Research 33 (1999) 67–85

3.3. Le6els of GABA in the hippocampus after administration of 6igabatrin 6ia osmotic minipumps Serum levels of vigabatrin and GABA concentrations in the hippocampus following vigabatrin delivery via Alzet minipumps are summarized in Table 2. Serum vigabatrin concentrations were elevated up to about 50 mM by 2 days of treatment. Also, hippocampal GABA levels were increased by 1.9-fold 2 days after pump implantation. These elevated vigabatrin and GABA levels remained constant during the rest of the 2-week period following pump implantation.

3.4. Number of neurons in the dentate hilus in the different treatment groups The estimated total number of hilar neurons in each treatment group is shown in Fig. 2A. The total number of neurons was 5385096862 (n = 5) in the control group, 5587592121 (n =2) in the vigabatrin-only group, 44786915705 (n = 7) in the KA + vehicle group, 3958999902 (n =7) in the KA+ VGB 1-h group, 3616195707 (n = 7) in the KA+VGB 2-day group, and 375009 7909 (n= 8) in the KA+VGB 7-day group. Kruskall – Wallis analysis showed significant differences between the treatment groups (P B 0.05). Compared to the control group, the diminished number of hilar neurons remaining in the KA+ vehicle group (83% of that in controls) was not significantly different. In all of the KA+VGB-treated groups, Table 2 Levels of vigabatrin in serum and GABA in the hippocampus after delivering vigabatrin (75 mg/kg) via subcutaneous Alzet minipumps Time of sampling

Serum VGB (mM) GABA in the hippocampus (mmol/g tissue)

Vehicle 2 days 1 week 2 weeks

B2 53.7 91.20 (3) 35.39 16.0 (3) 52.79 7.1 (3)

2.209 0.29 4.099 0.46 5.089 0.44 4.859 0.26

(3) (3) (3) (3)

VGB, vigabatrin; GABA, g-aminobutyric acid. Values are expressed as mean 9S.D. Number of animals is in parentheses. Vigabatrin concentration in the vehicle group was below the detection limit (2 mM).

75

the number of hilar neurons was significantly lower than in the control group (P B 0.05). However, the number of hilar neurons in the vigabatrin-treated groups did not differ from each other or from the KA+ vehicle group. The severity of damage to hilar neurons correlated with the density of mossy fiber sprouting (averaged across all five levels) (Fig. 3A). The number of hilar neurons also correlated with the number of SOM-ir neurons (averaged across all five levels) in the hilus (Fig. 3C).

3.5. Number of SOM-ir neurons in the dentate hilus in the different treatment groups The effect of different treatments on the number of hilar SOM-ir neurons was analyzed initially as the mean number of SOM-ir neurons per section averaged across all five levels (Fig. 2B). We also analyzed the distribution of SOM-ir neurons at five different levels along the septotemporal axis of the hippocampus in each treatment group (Fig. 5). The mean number of hilar SOM-ir neurons differed significantly between the treatment groups (PB0.01, Kruskall–Wallis analysis). In the KA+ vehicle group, the mean number of SOM-ir neurons was significantly lower than in the control (53% remaining, PB 0.01) or the vigabatrin only (PB 0.05) groups. Also, in the KA+VGB 1-h (41% of SOM-ir neurons remaining), KA+VGB 2-day (67% remaining), and KA+ VGB 7-day (55% remaining) groups, the mean number of SOM-ir neurons was significantly decreased compared to the control group (PB0.05, PB 0.05 and PB 0.01, respectively). Even though the mean number of SOM-ir neurons was highest in the KA+ VGB 2-day group, it did not differ significantly from the other KA+ VGB-treated groups. Vigabatrin treatment alone did not affect the number of SOM-ir neurons compared to controls. In the KA+ vehicle group, we found damage to SOM-ir neurons along the entire septotemporal axis of the hippocampus (ANOVA, PB 0.001 compared to controls) (Fig. 4C and Fig. 5A). As in our previous studies, the damage was milder in the septal end of the hippocampus than in the temporal end: 63% of SOM-ir neurons were left in the septal end of the hippocampus (level I, PB 0.05

76

A. Pitka¨nen et al. / Epilepsy Research 33 (1999) 67–85

compared to controls, Mann–Whitney U-test) while only 26% remained in the temporal end (level V, PB0.05 compared to controls, Mann–Whitney U-test). Vigabatrin treatment that was started 2 days after status epilepticus (KA+ VGB 2-day group) significantly alleviated the loss of SOM-ir neurons in the septal end of the hippocampus (ANOVA, P B0.05 compared to the vehicle group; levels I–III were included in the analysis) (Fig. 4E and Fig. 5C). In fact, in the KA+ VGB 2-day group, the number of SOM-ir neurons in the most septal end (level I, 92% of SOM-ir neurons remaining) did not differ from that in controls. In the KA+ VGB 1-h and KA+VGB 7-day groups, however, the damage to SOM-ir neurons was as severe as in the KA+vehicle group at all levels (Fig. 4D,F and Fig. 5B,D). The severity of damage to SOM-ir neurons correlated with the density of mossy fiber sprouting (Fig. 3B).

3.6. Mossy fiber sprouting in the different treatment groups

Fig. 3. (A) Correlation between the damage to all hilar neurons (expressed as the percentage of neurons lost compared to control values) and the density of mossy fiber sprouting. (B) Correlation between the damage to hilar somatostatin-immunoreactive (SOM-ir) neurons (expressed as the percentage of neurons lost compared to control values) and the density of mossy fiber sprouting. (C) Correlation between the number of hilar SOM-ir neurons and the number of all hilar neurons. r, correlation coefficient; P, statistical significance; n, number of cases included in the analysis.

The effect of different treatments on the density of mossy fiber sprouting was analyzed initially by a mean Timm score for all sections stained (Fig. 2C). We also determined Timm scores at five different levels along the septotemporal axis of the hippocampus in each treatment group (Figs. 6 and 7). The mean Timm score for all sections stained differed significantly between the treatment groups (PB0.001, Kruskall–Wallis analysis) (Fig. 2C). In all kainate-treated groups, the density of mossy fiber sprouting was higher than in the control group. Interestingly, the mean density of sprouting in the KA + VGB 1-h group tended to be even higher than in the KA + vehicle group (P= 0.054), the KA+ VGB 2-day group (P=0.055), and the KA+ VGB 7-day group (P= 0.055). Vigabatrin treatment alone did not increase the density of Timm granules in the inner molecular layer compared to control rats. In line with previous observations, the analysis of mossy fiber sprouting along the septotemporal axis of the hippocampus showed that, in normal control rats, some mossy fibers were found in the

A. Pitka¨nen et al. / Epilepsy Research 33 (1999) 67–85

supragranular region and inner molecular layer, particularly in the temporal end of the hippocampus. Accordingly, in all treatment groups we found more mossy fiber sprouting in the temporal end than in the septal end (Fig. 7). Vigabatrin treatment alone did not affect the mossy fiber sprouting (data not shown). In the KA+ vehicle group, the density of mossy fiber sprouting was higher than in control rats at all septotemporal levels (ANOVA, P B 0.001) (Fig. 6B and Fig. 7A). Moreover, in the KA+VGB 1-h, KA+ VGB 2-day, and KA +VGB 7-day groups, the

77

density of mossy fiber sprouting was higher than in control rats (ANOVA, PB 0.001) (Figs. 6 and 7). Interestingly, in the KA+ VGB 1-h group, the severity of sprouting was even more intense than in the KA+ vehicle group, especially in the septal end of the dentate gyrus (ANOVA, P B 0.001; levels I–II and IV, PB 0.05 compared to the KA+ vehicle group, Mann–Whitney U-test) (Fig. 6D and Fig. 7B). In the KA+ VGB 2-day and KA + VGB 7-day groups, the density of sprouting did not differ significantly from that in the vehicle group (Fig. 6E,F and Fig. 7C,D).

Fig. 4. Brightfield photomicrographs of horizontal sections stained for somatostatin immunohistochemistry to demonstrate the damage to somatostatin-immunoreactive (SOM-ir) neurons in the septal end of the hilus (level I) in the different treatment groups. (A) Control rat. Arrows point to SOM-ir neurons in the hilus. (B) Vigabatrin-treated rat (number of SOM-ir neurons in this group is 95% of that in controls). (C) KA + vehicle-treated rat (63% of SOM-ir neurons remaining). (D) Section from a rat in the KA + VGB 1-h group (65% of SOM-ir neurons remaining). (E) Section from a rat in the KA+ VGB 2-day group (89% of SOM-ir neurons remaining). (F) Section from a rat in the KA+ VGB 7-day group (74% of SOM-ir neurons remaining). Box with dashed outlines indicates the region in the dentate gyrus, from which the photomicrographs were taken to illustrate the Timm staining (see Fig. 5). gr, granule cell layer; H, hilus; mol, molecular layer. Scale bars in all panels, 200 mm.

78

A. Pitka¨nen et al. / Epilepsy Research 33 (1999) 67–85

Fig. 5. The number (mean 9S.D.) of hilar somatostatin-immunoreactive (SOM-ir) neurons at various septotemporal levels (I is the most septal; V is the most temporal) in the different treatment groups. To help comparisons of neuronal numbers between treatment groups in Panels (B – D), the dashed grey lines show the number of SOM-ir neurons in the control and KA + vehicle groups (as in panel (A)). Note that vigabatrin treatment started 2 days after the beginning of status epilepticus prevented the loss of SOM-ir neurons in the septal end of hippocampus (panel (C)). Vigabatrin treatment started either 1 h or 7 days after the beginning of status epilepticus had no neuroprotective effects (panels (B) and (D)). Asterisks indicate the statistical significances compared to the KA + vehicle group (*PB0.05, **PB 0.01, Mann–Whitney U-test).

4. Discussion In the present study we investigated whether the neuronal damage and mossy fiber sprouting induced by status epilepticus in the rat hippocampus can be prevented by chronic vigabatrin treatment that was started after the beginning of status epilepticus. This study protocol was designed to mimic a clinical situation where treatment with

antiepileptic drugs is started after a brain insult, such as tumor surgery, head trauma, or prolonged complex febrile seizures, to prevent subsequent epileptogenesis. Within 2 days of vigabatrin administration via osmotic minipumps, serum concentrations of vigabatrin increased on average up to 50 mM and hippocampal GABA concentrations increased by 1.9-fold. The levels of vigabatrin and GABA re-

A. Pitka¨nen et al. / Epilepsy Research 33 (1999) 67–85

mained elevated throughout the 2-week period, after which a new VGB-filled pump was implanted in each rat. These data show that we can produce constantly elevated vigabatrin and GABA levels chronically by administering vigabatrin via subcutaneous osmotic minipumps. Therefore, it is unlikely that there were any substantial fluctuations in GABA levels in the hippocampus of animals undergoing VGB administration via this method during the 2-month follow-up period.

79

4.1. Vigabatrin treatment did not ha6e any major neuroprotecti6e effects on hilar neurons The number of neurons in the hilus of control rats was estimated to be about 53000. This corresponds well with the previous stereological cell counts by West et al. (1991) who found 53400 neurons in the hilus of 30-day-old Wistar rats. The mean neuronal numbers decreased below control values by about 20–30% in all treatment groups. In fact, the number of

Fig. 6. Brightfield photomicrographs of horizontal sections stained for Timm histochemistry to demonstrate the extent of mossy fiber sprouting into the supragranular region and inner molecular layer in the septal end of the dentate gyrus (level I) in the different treatment groups. (A) Control rat. Note the absence of Timm granules in the supragranular region and inner molecular layer (arrow points to a single Timm granule in the inner molecular layer). (B) Vigabatrin-treated rat, in which the staining did not differ from that in controls (arrow points to a single Timm granule in the inner molecular layer). (C) KA +vehicle-treated rat. Note dense lamina of sprouted mossy fibers in the inner molecular layer (open arrow). (D) Section from a rat belonging to the KA+VGB 1-h group. Note that mossy fiber sprouting is even more extensive (open arrow) than in a vehicle-treated rat (panel (C)). (E) Section from a rat in the KA +VGB 2-day group. The extent of mossy fiber sprouting (open arrow) was similar to that in a vehicle-treated rat. (F) Section from a rat in the KA + VGB 7-day group. Also in this group, the extent of mossy fiber sprouting (open arrow) was similar to that in a vehicle-treated rat. Abbreviations: gr, granule cell layer; H, hilus; imol, inner molecular layer; mmol, mid portion of the molecular layer; omol, outer molecular layer. Scale bars in all panels, 100 mm.

80

A. Pitka¨nen et al. / Epilepsy Research 33 (1999) 67–85

Fig. 7. Timm scores (median) at various septotemporal levels (I is the most septal and V is the most temporal) in the different treatment groups. To help comparisons of Timm scores between treatment groups in Panels (B – D), the dashed grey lines show the Timm scores in the control and KA +vehicle groups (as in panel (A)). Note that vigabatrin treatment started 1 h after the beginning of status epilepticus increased mossy fiber sprouting, especially septally, in the hippocampus compared to the KA + vehicle group (panel (B)). Vigabatrin treatment started either 2 or 7 days after the beginning of status epilepticus had no effect on sprouting compared to the KA + vehicle group (panels (C) and (D)). Asterisks indicate the statistical significances compared to the KA+ vehicle group (*PB 0.05, **PB 0.01, Mann–Whitney U-test).

hilar neurons in each of the KA + vigabatrin groups, but not the KA + vehicle group, was lower than in controls. Therefore, vigabatrin treatment started either 1 h, or 2 or 7 days after the beginning of status epilepticus could not prevent the overall loss of neurons in the hilus. The vulnerability of specific seizure-sensitive neurons, such as SOM-ir neurons, was also com-

pared between the different groups. Two months after status epilepticus, 53% of SOM-ir neurons remained in the KA+ vehicle group. In rat groups where vigabatrin treatment was started either 1 h or 2 or 7 days after the beginning of status epilepticus, the percentage of total SOM-ir neurons remaining were 41, 67 and 55%, respectively, but were not statistically different from the KA+ vehicle group.

A. Pitka¨nen et al. / Epilepsy Research 33 (1999) 67–85

It is well known that the hippocampal damage caused by kainate-induced status epilepticus is more severe in the temporal end than in the septal end of the hippocampus (Schwob et al., 1980; Sperk, 1994). Therefore, averaging cell counts across the entire septotemporal axis of the hippocampus may mask variations in neuronal densities at specific septotemporal locations. Therefore, we analyzed the number of SOM-ir neurons at each of five different septotemporal levels of the hippocampus. Surprisingly, we found that vigabatrin treatment beginning 2 days after the onset of status epilepticus protected SOM-ir neurons in the septal end of the hilus. Previous neuropathological analyses of the hippocampus indicate that hippocampal pyramidal cell damage can be detected 4 h after intraperitoneally administered kainic acid, and that it progresses over a period of several days (Schwob et al., 1980; Filipkowski et al., 1994). It is possible that chronic vigabatrin treatment started 2 days after status epilepticus prevented the delayed neuronal damage to SOM-ir neurons directly in the septal dentate gyrus. Another explanation could be that, during the 2-month period, vigabatrin prevented the spread of epileptic activity via intrahippocampal pathways from the severely damaged temporal end of the hippocampus to the septal end, and therefore halted the progression of epileptic neuronal damage septally. This assumption is supported by previous findings which showed that vigabatrin delays the development of kindling (Shin et al., 1986). Vigabatrin treatment beginning 7 days after the onset of status epilepticus did not alleviate damage to SOM-ir neurons in the septal end of the hilus. Most likely, by this time the ongoing neuronal damage is already extensive in various brain regions, including the hippocampus (Schwob et al., 1980), rendering such late vigabatrin treatment ineffective. Why treatment starting 1 h after the beginning of status epilepticus also lacked neuroprotective efficacy is unclear. Particularly since one might expect that vigabatrin treatment during status epilepticus would suppress seizure activity, and consequently prevent or diminish further damage.

81

On the other hand, this lack of an effect early on could be due to inter-animal variability since the severity and duration of status epilepticus after kainate injection can vary largely between rats (Sperk, 1994). Therefore, an additional group of rats was continuously monitored with a videoEEG recording system to measure electrographic variations in animals, which displayed similar behavioral patterns of kainate-induced status epilepticus. In line with previous studies, there was substantial variability between the animals. For example, the last HAFD appeared anywhere between 3.9 and 16.0 h and 6.1 and 19.9 h after kainate injection in the vehicle and vigabatrin groups, respectively. Inter-animal variability was also evident in histological preparations 2 months after kainate injection. For example, the variation percentage [(standard deviation/mean) × 100%] in hilar cell counts was 13% in the control group, compared to 35% in the KA + vehicle group, and between 16 and 25% in the KA + VGB-treated groups. So the fact that the 1-h VGB group did not show a significant effect on septal SOM-ir neurons in the hilus could be due to the fact that rats are undergoing a high degree of variability in EEG activity and ongoing neuronal damage, particularly at this early time point. Nevertheless, such variability between subjects did not mask a significant effect of vigabatrin in protecting SOMir neurons in the septal end of the hilus 2 days after status epilepticus. Another reason that vigabatrin initiated within 1 h failed to show a neuroprotective effect could be due to the fact that high doses of vigabatrin itself can have proconvulsant/toxic effects. In fact, in the KA+ VGB 1-h group, we initially began vigabatrin treatment with a bolus injection of 1000 mg/kg. In these early attempts, 12 of 19 of these rats died during status epilepticus, one more was dead 2 days later, and three more animals had to be sacrificed 2 weeks later because they were in poor condition. A total of 16 out of 19 animals were lost in this pilot group (data not shown). Thereafter, we reduced the initial vigabatrin dose in this group to 500 mg/kg. With this dose, only one of ten rats died during status epilepticus. We have previously shown that 5 h after a 1000-mg/kg intraperitoneal injection of

82

A. Pitka¨nen et al. / Epilepsy Research 33 (1999) 67–85

vigabatrin, the drug level in the rat cisternal CSF is about 90 mM (Halonen et al., 1990). It is interesting that 100 mM of vigabatrin was recently reported to actually decrease GABAergic inhibition in a hippocampal slice preparation (Jackson et al., 1994) and to decrease the uptake of chloride ions within membrane vesicles in the cerebral cortex (Suzuki et al., 1991). Therefore, such acutely elevated vigabatrin levels may have effects that are unrelated to GABA transaminase inhibition and GABA elevation. Especially in the KA +VGB 1-h group, in which the permeability of the blood–brain barrier may be increased due to status epilepticus, thus allowing the suddenly high levels of vigabatrin, produced by the bolus intraperitoneal injection, to more readily penetrate the brain. In these rats, increased vigabatrin levels may have led to unfavorable neurotoxic effects that preceded and overcame the seizuresuppressing effects of vigabatrin-induced increased GABA release. Overall, our evidence suggests that vigabatrin treatment may have a mild neuroprotective effect on hilar SOM-ir neurons in the septal end of the hippocampus when administered within 2 days, soon after status epilepticus has subsided. This effect is similar to that found in rats whose fimbria-fornix was transected 2 days after kainateinduced status epilepticus to reduce disinhibition caused by the septohippocampal GABAergic pathway (Jolkkonen et al., 1997). The absence of a neuroprotective effect if vigabatrin is started either during or 1 week after status epilepticus indicates that the timing of drug administration after brain insults may be critical in preventing further damage associated with epileptogenesis.

4.2. Mossy fiber sprouting may also be modified by 6igabatrin treatment Several studies have shown that mossy fiber sprouting is a relatively slow process occurring over the weeks or months after a seizure episode and that it is permanent (for review, see Represa et al., 1994). Very few previous studies have, however, investigated the effects of anticonvulsant medication on axonal sprouting. Sutula et al. (1992) found that phenobarbital treatment started

immediately after an intracerebroventricular injection of kainate and continued for 2 weeks thereafter reduced mossy fiber sprouting in a dose-dependent manner. They also found that even if the treatment was started 1 day after the induction of status epilepticus and continued only for 5 days, mossy fiber sprouting could be reduced. No effect was found if the treatment was started 9 days after status epilepticus onset. More recently, Moraes and Cavalheiro (1995) reported that diazepam or phenobarbital started immediately after pilocarpine-induced status epilepticus in rats and continued for 1 month reduced mossy fiber sprouting. Our data show that chronic vigabatrin treatment started 2 or 7 days after status epilepticus did not prevent mossy fiber sprouting that was assessed 2 months after kainate treatment. If chronic vigabatrin treatment was started during status epilepticus (i.e. 1 h following onset), sprouting appeared even more dense than in vehicletreated animals. This suggests that chronic elevation of GABA in the brain does not prevent or reduce mossy fiber sprouting after status epilepticus. Little is known about the factors that initiate and maintain sprouting, determine its magnitude, and guide the mossy fibers to new synaptic targets in the adult epileptic hippocampus. One hypothesis states that the deafferentation of mossy fibers from their natural target neurons in the hilus due to seizure-induced neuronal loss may be involved in the initiation of sprouting (Sutula et al., 1994). Accordingly, in the present study we found a correlation between the severity of neuronal damage in the hilus and the density of mossy fiber sprouting. Alternatively, axonal reorganization can also be triggered by excessive synchronous activity (Sutula et al., 1994; Stringer et al., 1997). As mentioned above, acute vigabatrin application to rat hippocampal slice preparations may actually reduce GABA-mediated inhibition (Jackson et al., 1994). Whether the occasional appearance of high-frequency discharges above the suppressed background activity in the KA+ VGB 1-h group is related to the direct effect of vigabatrin on neuronal activity, and whether it contributes to excessive sprouting of mossy fibers, will be determined in future studies.

A. Pitka¨nen et al. / Epilepsy Research 33 (1999) 67–85

It was recently shown that mossy fiber sprouting may facilitate the development of excitatory circuitry within the dentate gyrus (Wuarin and Dudek, 1996). Even though we did not systematically record behavioral or electroencephalographic seizures of rats that were treated with vigabatrin chronically, five of ten animals in the KA +VGB 1-h group had spontaneous epileptic seizures during handling over the 2-month treatment period (data not shown).

4.3. Final comment Various cerebral insults such as status epilepticus (DeGiorgio et al., 1992; Fujikawa and Itabashi, 1994; Nohria et al., 1994), head trauma (Lowenstein et al., 1992), prolonged febrile seizures (Kuks et al., 1993), or even a low number of generalized seizures (Cavazos et al., 1991, 1994) may cause neuronal damage with or without mossy fiber sprouting that may lead to a lowered seizure threshold and epileptogenesis in the hippocampus (Sloviter, 1994). Prevention of these structural changes after such insults may provide us with tools that can be used to prevent the appearance of spontaneous seizures and the development of epilepsy later in life. Our study shows that chronic vigabatrin treatment did not substantially alleviate the neuropathological alterations caused by kainate-induced status epilepticus. More recently, we also found that vigabatrin treatment did not prevent the development of epilepsy after status epilepticus induced by stimulation of the amygdala (Halonen et al., 1997). Stimulating the GABAergic system may not be an ideal approach in the prevention of epilepsy in cases where epileptogenesis is triggered by injury.

Acknowledgements We thank Raija Pitka¨nen and Merja Lukkari for their expert technical help. We acknowledge and thank Hoechst Marion Roussel for providing vigabatrin. We greatly appreciate the comments of Dr John Mumford on the manuscript. This study was supported by The Academy of Finland and The Vaajasalo Foundation (AP).

83

References Babb, T.L., Pretorius, J.K., 1993. Pathologic substrates of epilepsy. In: Wylie E., (Ed.), The Treatment of Epilepsy: Principles and Practice. Lea and Febiger, Philadelphia, PA, pp. 55 – 70. Buckmaster, P.S., Wenzel, H.J., Kunkel, D.D., Schwartzkroin, P.A., 1996. Axon arbors and synaptic connections of hippocampal mossy cells in the rat in vivo. J. Comp. Neurol. 366, 270 – 272. Cavazos, J., Golarai, G., Sutula, T., 1991. Mossy fiber synaptic reorganization induced by kindling: time course of development, progression and permanence. J. Neurosci. 11, 2795 – 2803. Cavazos, J.E., Das, I., Sutula, T.P., 1994. Neuronal loss induced in limbic pathways by kindling: evidence for induction of hippocampal sclerosis by repeated brief seizures. J. Neurosci. 14, 3106 – 3121. DeGiorgio, C.M., Tomiyasu, U., Gott, P.S., Treiman, D.M., 1992. Hippocampal pyramidal cell loss in human status epilepticus. Epilepsia 33, 23 – 27. Filipkowski, R.K., Hetman, M., Kaminska, B., Kaczmarek, L., 1994. DNA fragmentation in rat brain after intraperitoneal administration of kainate. Neuroreport 5, 1538 – 1540. Fujikawa, D.G., Itabashi, H.H., 1994. Status epilepticus-induced brain damage in humans. Epilepsia 35 (Suppl. 8), 9. Gibson, J.P., Yarrington, J.T., Loudy, D.E., Gerbig, C.G., Hurst, G.H., Newberne, J.W., 1990. Chronic toxicity studies with vigabatrin, a GABA-transaminase inhibitor. Toxicol. Pathol. 18, 225 – 238. Grant, S.M., Heel, R.C., 1991. Vigabatrin. A review of its pharmacodynamic and pharmacokinetic properties, and therapeutic potential in epilepsy and disorders of motor control. Drugs 41, 889 – 926. Halonen, T., Pitka¨nen, A., Riekkinen, P.J., 1990. Administration of vigabatrin (Vinylaminobutyric acid) affects the levels of both inhibitory and excitatory amino acids in rat cerebrospinal fluid. J. Neurochem. 55, 1870 – 1874. Halonen, T., Miettinen, R., Toppinen, A., Tuunanen, J., Kotti, T., Riekkinen, P.J. Sr., 1995. Vigabatrin protects against kainic acid-induced neuronal damage in the rat hippocampus. Neurosci. Lett. 195, 13 – 16. Halonen, T., Nissinen, J., Pitka¨nen, A., 1997. Potentiation of GABAergic system after status epilepticus does not prevent the development of epilepsy in rats. Epilepsia 38 (Suppl. 8), 111. Hernandez, T.D., 1997. Preventing post-traumatic epilepsy after brain injury: weighting the costs and benefits of anticonvulsant prophylaxis. Trends Pharmacol. Sci. 18, 59 – 62. Houser, C.R., Miyashiro, J.E., Swartz, B.E., Walsh, G.O., Rich, J.R., Delgado-Esqueta, A.V., 1990. Altered patterns of dynorphin immunoreactivity suggest mossy fiber reorganization in human hippocampal epilepsy. J. Neurosci. 10, 267 – 282.

84

A. Pitka¨nen et al. / Epilepsy Research 33 (1999) 67–85

Jackson, M.F., Dennis, T., Esplin, B., Capek, R., 1994. Acute effects of vinyl GABA (vigabatrin) on hippocampal GABAergic inhibition in vitro. Brain Res. 651, 85– 91. Jolkkonen, J., Jolkkonen, E., Pitka¨nen, A., 1997. Seizure-induced damage to somatostatin-immunoreactive neurons in the rat hippocampus is regulated by fimbria–fornix transection. Exp. Neurol. 145, 141–153. Kuks, J.B.M., Cook, M.J., Fish, D.R., Stevens, J.M., Shorvon, S.D., 1993. Hippocampal sclerosis in epilepsy and childhood febrile seizures. Lancet 342, 1391–1394. Lowenstein, D.H., Thomas, M.J., Smith, D.H., McIntosh, T.K., 1992. Selective vulnerability of dentate hilar neurons following traumatic brain injury: a potential mechanistic link between head trauma and disorders of the hippocampus. J. Neurosci. 12, 4846–4853. Moraes, W.A.S., Cavalheiro, E.A., 1995. Antiepileptic treatment reduces status epilepticus induced supragranular sprouting. Epilepsia 36 (Suppl. 3), 47. Moser, E., Moser, M.-B., Andersen, P., 1993. Spatial learning impairment parallel the magnitude of dorsal hippocampal lesions, but is hardly present following ventral lesions. J. Neurosci. 13, 3916–3925. Nohria, V., Lee, N., Tien, R.D., Heinz, E.R., Smith, J.S., DeLong, G.R., Skeen, M.B., Resnick, T.J., Crain, B., Lewis, D.W., 1994. Magnetic resonance imaging evidence of hippocampal sclerosis in progression: a case report. Epilepsia 35, 1332 – 1336. Okazaki, M.M., Evenson, D.A., Nadler, J.V., 1995. Hippocampal mossy fiber sprouting and synapse formation after status epilepticus in rats: visualization after retrograde transport of biocytin. J. Comp. Neurol. 352, 515 – 534. Pitka¨nen, A., Tuunanen, J., Halonen, T., 1996. Vigabatrin and carbamazepine have different efficacies in the prevention of status epilepticus induced neuronal damage in the hippocampus and amygdala. Epilepsy Res. 24, 29–45. Pollard, H., Cantagrel, S., Charriaut-Marlangue, C., Moreau, J., Ben-Ari, Y., 1994a. Apoptosis associated DNA fragmentation in epileptic brain damage. Neuroreport 5, 10531055. Pollard, H., Charriaut-Marlangue, C., Cantagrel, S., Represa, A., Robain, O., Moreau, J., Ben-Ari, Y., 1994b. Kainate-induced apoptotic cell death in hippocampal neurons. Neuroscience 63 (1), 7–18. Represa, A., Jorquera, I., Le Gal La Salle, G., Ben-Ari, Y., 1993. Epilepsy induced collateral sprouting of hippocampal mossy fibers: does it induce the development of ectopic synapses with granule cell dendrites? Hippocampus 3, 257 – 268. Represa, A., Niquet, J., Pollard, H., Khrestchatisky, M., Ben-Ari, Y., 1994. From seizures to neo-synaptogenesis: intrinsic and extrinsic determinants of mossy fiber sprouting in the adult hippocampus. Hippocampus 4, 270–274. Schwob, J.E., Fuller, T., Price, J.L., Olney, J.W., 1980. Widespread patterns of neuronal damage following sys-

temic or intracerebral injections of kainic acid: a histological study. Neuroscience 5, 991 – 1014. Shin, C., Rigsbee, L.C., McNamara, J.O., 1986. Anti-seizure and anti-epileptogenic effect of g-vinyl-g– aminobutyric acid in amygdaloid kindling. Brain Res. 398, 370 – 374. Sloviter, R.S., 1982. A simplified Timm stain procedure compatible with formaldehyde fixation and routine paraffin embedding of the rat brain. Brain Res. Bull. 8, 771 – 774. Sloviter, R.S., 1992. Possible functional consequences of synaptic reorganization in the dentate gyrus of kainatetreated rats. Neurosci. Lett. 137, 91 – 96. Sloviter, R.S., 1994. The functional organization of the hippocampal dentate gyrus and its relevance to the pathogenesis of temporal lobe epilepsy. Ann. Neurol. 35, 640 – 654. Sperk, G., 1994. Kainic acid seizures in rat. Prog. Neurobiol. 42, 1 – 32. Stringer, J.L., Agarwal, K.S., Dure, L.S., 1997. Is cell death necessary for hippocampal mossy fiber sprouting? Epilepsy Res. 27, 67 – 76. Sutula, T., Cascino, G., Cavazos, J., Parada, I., Ramirez, L., 1989. Mossy fiber synaptic reorganization in the epileptic human temporal lobe. Ann. Neurol. 26, 321 – 330. Sutula, T., Cavazos, J., Golarai, G., 1992. Alteration of long-lasting structural and functional effects of kainic acid in the hippocampus by brief treatment with phenobarbital. J. Neurosci. 12, 4173 – 4187. Sutula, T.P., Cavazos, J.E., Woodard, A.R., 1994. Longterm structural and functional alterations induced in the hippocampus by kindling: implication for memory dysfunction and the development of epilepsy. Hippocampus 4, 254 – 258. Suzuki, Y., Mimaki, T., Arai, H., Okada, S., Kuriyama, K., 1991. Effect of vinyl aminobutyric acid on the aminobutyric acid receptor-coupled chloride ion channel in vesicles from the brain of the rat. Neuropharmacology 30, 423 – 427. Tauck, D.L., Nadler, J.V., 1985. Evidence of functional mossy fiber sprouting in hippocampal formation of kainic acid-treated rats. J. Neurosci. 5, 1016 – 1022. Valtonen, P., Haapalinna, A., Riekkinen, P. Sr., Halonen, T., 1995. Effect of alpha 2-adrenergic drug dexmedetomidine and atipamezole on extracellular amino acid levels in vivo. Eur. J. Pharmacol. 285, 239 – 246. West, M.J., Slomianka, L., Gundersen, H.J.G., 1991. Unbiased stereological estimation of the total number of neurons in the subdivisions of the rat hippocampus using optical fractionator. Anat. Rec. 231, 482 – 497. Wuarin, J.-P., Dudek, F.E., 1996. Electrographic seizures and new recurrent excitatory circuits in the dentate gyrus of hippocampal slices from kainate-treated epileptic rats. J. Neurosci. 16, 4438 – 4448. Ylinen, A., Miettinen, R., Pitka¨nen, A., Gulyas, A.I., Freund, T.F., Riekkinen, P.J., 1991. Enhanced GABAergic inhibition preserves hippocampal structure and function in a model of epilepsy. Proc. Natl. Acad. Sci. USA 88, 7650 – 7653.

A. Pitka¨nen et al. / Epilepsy Research 33 (1999) 67–85 Zhang, N., Houser, C.R., 1995. Dynorphin immunoreactivity in the dentate gyrus in human temporal lobe epilepsy: an electron microscopic study of reorganized mossy fiber synapses. Soc. Neurosci. Abstr. 21, 1962. Zola-Morgan, S., Squire, L.R., Amaral, D.G., 1986. Human amnesia and the medial temporal region: enduring memory

.

85

impairment following a bilateral lesion limited to field CA1 of the hippocampus. J. Neurosci. 6, 2950 – 2967. Zola-Morgan, S., Squire, L.R., Rempel, N.L., Clower, R.P., Amaral, D.G., 1992. Enduring memory impairment in monkeys after ischemic damage to the hippocampus. J. Neurosci. 12, 2582 – 2595.