INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Common histopathologic features of temporal lobe epilepsy include hippocampal sclerosis with associated Timm staining in the inner molecular layer (IML) of the dentate gyrus (i.e., mossy fiber sprouting), which are found in both humans and animal models of temporal lobe epilepsy (e.g., Nadler et al. 1980; Ben-Ari 1985; Sutula et al. 1989; Houser et al. 1990; Babb et al. 1991; Buckmaster and Dudek 1997a,b). Mossy fiber sprouting in the dentate gyrus (and synaptic reorganization in other temporal lobe circuits) may be a factor in the pathogenesis of temporal lobe epilepsy, and it is hypothesized to be caused by hilar neuron loss and subsequent deafferentation of the inner molecular layer (for review see, Dudek and Spitz 1997; McNamara 1994, 1999; Nadler 1981). Experimental evidence strongly suggests that these aberrant mossy fibers form new recurrent excitatory synapses that may trigger seizures and/or alter the normal gating function of the dentate gyrus (e.g., Lynch and Sutula 2000; Molnar and Nadler 1999; Tauck and Nadler 1985; Wuarin and Dudek 1996, 2001). This new excitatory circuit is normally masked by inhibitory input into the granule cells, but can be revealed under conditions of reduced inhibition and/or high extracellular potassium (Cronin et al. 1992; Hardison et al. 2000; Lynch and Sutula 2000; Patrylo and Dudek 1998; Wuarin and Dudek 1996, 2001). Other evidence suggests that the new mossy fiber axons synapse onto basket cells, and that this leads to a form of synaptic reorganization that may be restorative (Buhl et al. 1996; Sloviter 1992).

Recently, Longo and Mello (1997, 1998) reported that when rats were pre-treated with cyclohexamide (CHX, a potent inhibitor of protein synthesis) prior to receiving pilocarpine [(PILO, a muscarinic agonist that causes status epilepticus and multi focal epilepsy; Turski et al. 1983], Timm staining in the IML was prevented, but the CHX pre-treated rats still developed spontaneous motor seizures. These data have been used as evidence that mossy fiber sprouting was not necessary for the generation of spontaneous chronic motor seizures.

We tested the hypothesis that CHX would prevent hippocampal neuron loss following PILO-induced status epilepticus, and thus prevent mossy fiber sprouting. We further hypothesized that the rats that lack mossy fiber sprouting would be deficient in the electrophysiological events suggestive of new recurrent excitatory synapses in the hippocampal granule cell layer (e.g., see Wuarin and Dudek 1996, 2001). We used a septo-temporal analysis of Timm staining in the IML and a quantitative stereological assessment of the hippocampus to address the first hypothesis. The second hypothesis was examined with in vitro extracellular and whole cell patch-clamp recordings in hippocampal slices. We found, however, that pre-treatment with CHX did not prevent hippocampal neuron death nor did it block mossy fiber sprouting in the IML of the dentate gyrus, even though CHX reduced PILO-induced status epilepticus and may have increased the mortality of PILO-treated rats. Furthermore, when the PILO- and PILO+CHX-treated rats with only a few seizures during treatment (i.e., no status epilepticus) were examined, some of these animals had spontaneous motor seizures, but we did not detect clear and consistent histopathologic changes in the hippocampus. Together, these data suggest that 1) pre-treatment with CHX does not block mossy fiber sprouting, but rather affects the likelihood of convulsive status epilepticus; 2) neuron loss in the hilus, CA3, and CA1 and mossy fiber sprouting (detected by the Timm stain method) are not necessary for the development of chronic motor seizures after PILO treatment; and 3) rats with PILO-induced epilepsy have seizures that possibly arise from other areas of the brain that may have been lesioned, but these were not examined in this study (i.e., amygdala, subiculum, entorhinal cortex, and the thalamus).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Pilocarpine and cyclohexamide treatment

Adult male Sprague-Dawley rats (180-200 g; Harlan) were kept in a standard light/dark cycle (12/12 h) and fed ad libitum. Rats were first injected with CHX (1 mg/kg sc). The CHX was stored at 2-8°C in a desiccator. This product is stable for 5 yr and was manufactured in 1996 and used in 1999. Ten minutes after the CHX injection, the rats were administered methyl-scopolamine (1 mg/kg ip). After 30-45 min, the rats were given PILO (320 mg/kg ip). Many of the rats developed status epilepticus (defined as 10 or more convulsive seizures in a 90-min period) within 20-30 min following injection of PILO. After 90 min, convulsive seizures were stopped with sodium pentobarbital (17.5 mg/kg ip). Positive control rats were injected with the same protocol as above, except CHX was not given. Control animals were also given the same drugs as above, except saline was used instead of PILO. After the rats were treated with pentobarbital, they were allowed to recover and were monitored for motor seizures until they were used for histological and electrophysiological experiments (approximately 60 days later).

Behavioral monitoring

The rats were coded after status epilepticus so that the observers were unaware of the treatment. This procedure was followed for all experiments so that all observations were made blind. Rats were directly monitored 6 h/wk until killed. The animals were also video-monitored for 24-h periods, intermittently throughout the 60 days, until used for experiments. The average amount of time each rat was monitored was 50 h of direct monitoring and 110 h of 24-h video-monitoring. The relative latent period was estimated as the time between the injection date and the first observed motor seizure (i.e., direct or video-monitored). A seizure rate was determined by dividing the total number of observed convulsive seizures by the total number of hours the rats were observed. This included both direct- and video-monitored seizures. The relative latent period and seizure rate for each treatment group were averaged, and the treatments were compared to each other using an ANOVA with a Tukey's multiple-comparison analysis.

Histology

A modified Timm histological procedure was used to label the zinc-containing axons of the granule cells. The hippocampus that was not used for in vitro electrophysiological experiments was dissected from the hemisphere and prefixed in phosphate buffered 0.37% Na₂S to precipitate the zinc in the mossy fibers. The hippocampus was immersion-fixed in phosphate buffered 4% paraformaldehyde (pH = 7.2). This same procedure was used on the recorded slices, and the hemisphere that was used for electrophysiology was assigned randomly. The tissue was saturated with 30% sucrose and sectioned at 35 µm on a sliding microtome. The hippocampus was straightened prior to mounting on the microtome (Buckmaster and Dudek 1997b). Every sixth section was mounted for Timm stain with cresyl violet counter stain, and the adjacent sections were mounted for cresyl violet staining alone. Every section from the recorded slices was mounted and processed for Timm stain analysis. Timm staining was performed in batches that included sections from saline-, PILO-, and PILO+CHX-treated rats, and the rats with only a few motor seizures at treatment. All the sections were coded and the intensity of the Timm stain in the IML was graded with blind procedures according to the rating scale of Tauck and Nadler (1985). The sections were grouped according to their location along the septotemporal axis starting at the septal end (i.e., 25% is the septal end and 100% is the temporal end). The grades of all the grouped sections were summed and then divided by the total number of sections/group to determine an average Timm score for each region. The Timm score for each region was then averaged for each treatment and rounded to the nearest whole number, and a median was determined. The treatments were compared to a hypothetical mean of 0 using the Wilcoxan signed rank test, a non-parametric analysis.

Stereology

Neuronal counts of the hilus, CA1, and CA3 were estimated using the optical fractionator probe (West et al. 1991) in the counting software Stereoinvestigator (MicroBrightfield, Colchester, VT). The cresyl violet-stained sections were used to estimate the total number of neurons, and the sections were coded so that the investigators performing the counts were blind to the treatment. Neurons were defined by their large nuclei with dispersed chromatin and prominent nucleoli. Only those nuclei that came into focus while focusing down through the dissector height were counted. The estimated counts for each section were grouped as described in the Timm stain analysis. Each region contained no less than three sections. Neuronal counts for each region were then derived using the formula from West et al. (1991). The estimated counts from each region were then averaged for each treatment group. Each region for the treatment groups was compared to each other using an ANOVA with a Tukey's multiple-comparisons test.

Slice preparation and recording methods

All rats were anesthetized with pentobarbital sodium (50 mg/kg ip), and their brains were dissected and placed for 30-60 s in oxygenated (95% O₂-5% CO₂), ice-cold artificial cerebrospinal fluid (ACSF) containing (in mM) 124 NaCl, 26 NaHCO₃, 3 KCl, 1.3 MgSO₄, 1.4 NaH₂PO₄, 1.3 CaCl₂, and 11 glucose (pH 7.4). The brains were bisected, and one-half was glued on the stage of a vibratome (Campden Instruments, Lafayette, IN), and the other half was saved for histology. Four to six 300- to 400-µm-thick slices, mostly from the middle one-third of the hippocampus, were cut parallel to the base of the brain (i.e., ventral horizontal sections).

Extracellular recordings

Slices were trimmed to isolate the hippocampus and were kept in an interface chamber at 32°C, with humidified 95% O₂-5% CO₂ blown over the slices. Micropipettes for extracellular recordings were pulled from borosilicate glass capillaries (1B100F-4; World Precision Instruments, Sarasota, FL) to an open resistance of 2-30 M and filled with 1 M NaCl. Extracellular recordings were amplified with an AC coupled Axoprobe-1A amplifier at a gain of 100. Recording and analysis of all the electrophysiological data was done without knowledge of the treatment. The recording pipette was placed into the granule cell layer, and the bipolar stimulating electrode (platinum/iridium Teflon-coated) was placed in the hilus to evoke population spikes. Recordings were initially obtained in control solution (ACSF). Three stimulus intensities based on the minimum current needed to evoke a 0.5 mV potential (minimum, 2 times minimum, 4 times minimum) were used to assess population spike responses. Slices that did not produce population spikes of 5 mV or greater with hilar stimulation were not included in the analysis. The hippocampal slices were then bathed in 30 µM bicuculline and the stimulation protocol was repeated, as established in the control solution. While still in 30 µM bicuculline, the extracellular potassium was raised from 3 to 6 mM, and the stimulation protocol was again repeated. The glutamate receptor antagonists, 2-amino-5-phosphonovaleric acid (AP-5; Sigma) at 50 µM and 6,7-dinitroquinoxaline-2,3,(1H,4H)-dione (DNQX; Sigma) at 50 µM, were used to assess the role of glutamate receptors in the generation of population spike bursts. After the completion of the recordings, the slices were individually fixed for Timm stain analysis. A linear regression was performed to compare burst duration and Timm stain grade.

Whole cell patch-clamp recordings

For the whole cell recordings, two to four 300-µm-thick slices were kept for 1-2 h in oxygenated ACSF at 32°C before being transferred to the recording chamber where recordings were obtained at room temperature (21-23°C). Perfusion solution (10 ml) containing caged glutamate [L-glutamic acid, -(-carboxy-2-nitrobenzyl) ester (250 µM); Molecular Probes, Eugene OR] was oxygenated and recirculated at 2 ml/min. Pipettes for whole cell recordings were made from borosilicate glass capillaries (KG-33; Garner Glass, Claremont, CA) with a horizontal pipette puller (P-87 Flaming-Brown pipette puller; Sutter Instruments, Novato, CA) to an open resistance of 2-4 M. Pipette solution contained (in mM) 140 K-gluconate, 10 HEPES, 1 NaCl, 1 CaCl₂, 1 MgCl₂, 5 EGTA, and 4 magnesium ATP (pH 7.2). The data were not used if the resting membrane potential was less than 70 mV when whole cell configuration was obtained. Whole cell currents were amplified with an Axopatch-1D amplifier (Axon Instruments, Foster City, CA), filtered (2 kHz), digitized (44 kHz), and stored on videotapes (Neuro-Corder; Neurodata Instruments, New York, NY). Data analysis was done off-line with sampling rates of 5-10 kHz (pClamp 6; Axon Instruments). The detection and measurement of spontaneous excitatory postsynaptic currents (EPSCs) was done with Mini Analysis (Synaptosoft, Leonia, NJ). Every detected event was examined, and only EPSCs, characterized by a typical fast rising phase and slow decay phase, were included. Recording and analysis of the electrophysiological data was done without knowledge of the treatment. Cumulative amplitude distributions were compared with the Kolmogorov-Smirnov two-sample test. The significance of the maximum difference between distributions was determined using a ² distribution (Siegel 1956). The ² test was also used to test for differences in the effects of the flash photolysis of caged glutamate between the different treatment groups. Data are expressed as mean ± SE.

Flash photolysis of caged glutamate

Uncaging of glutamate (Callaway and Katz 1993) was obtained with a xenon flash lamp (Chadwick-Helmuth, El Monte, CA). The flash of UV light was transmitted through a high-numerical-aperture, oil-immersion objective (×40; Nikon) mounted underneath the bottom of the recording chamber (coverslip) and focused approximately 200 µm into the tissue. An HeNe laser aimed directly through the objective into the tissue was used to determine the location of the photostimulation. A monochrome CCD camera (Cohu, San Diego, CA) and video monitor were used to determine the location of, and the distance between, the recorded cells and the spot of laser light. Flashes with intensities of 50-100 mJ combined to a concentration of caged glutamate of 250 µM and produced a spatial resolution of approximately 100 µm. Consequently, photostimulations were applied in the granule cell layer at sites 150 µm apart.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Morbidity and mortality during SE

The morbidity and mortality rates during status epilepticus were analyzed to provide an external physiologic measure of the systemic effect of CHX. The saline-injected animals were not included in this analysis. The morbidity rate (i.e., rats having at least 1 seizure by approximately 30 min after PILO administration; Fig. 1A) was significantly lower in the PILO+CHX rats (57%; a total of 35 animals were injected: 17 rats had status epilepticus, 3 rats 5 seizures, and 15 rats had no motor seizures) compared to PILO treatment alone (90%; a total of 29 rats were injected: 20 rats had status epilepticus, 6 rats 5 seizures, and 3 rats had no motor seizures; P = 0.01; ²). In those animals that did have motor seizures during treatment, both groups behaved in a similar manner during status epilepticus, but the mortality rate (Fig. 1B) was three times higher in the PILO+CHX-treated rats (35%, 7 of 20) compared to the PILO-treated animals (11%, 3 of 26); however, this difference was not significant (P = 0.06; Fisher exact test). If these ratios existed for a larger number of replications (i.e., 30 per group), the mortality rate would have been significantly different between these two groups. The rats that did not have seizures (n = 18) or had five or fewer seizures during treatment (n = 9) were kept for behavioral observation to assess the occurrence of spontaneous motor seizures. All animals in these groups survived treatment. Data from the rats that did not have any seizures during treatment are not included in any part of this study. These data suggest that CHX acted systemically to depress the occurrence of PILO-induced status epilepticus in a significant proportion of animals.

View larger version (27K): [in this window] [in a new window]   Fig. 1. Morbidity and mortality rate during pilocarpine treatment. A: morbidity rate was defined as the number of rats that responded to treatment with at least 1 motor seizure. Morbidity rate was significantly lower if the rats were pre-treated with cycloheximide [CHX; 57% for pilocarpine (PILO)+CHX and 90% for PILO, P = 0.01, 2]. B: mortality rate is defined as the total number of rats with acute motor seizures that died within a 24-h period after administration of PILO. Mortality rate was 35% in PILO+CHX rats (n = 20) and 17% in rats treated with only PILO (n = 26). Although the mortality rate was 3-fold higher for the CHX pre-treated rats, the difference was not significant between these 2 groups (P = 0.06, 2). Rats that did not respond with status epilepticus, but experienced a few seizures during treatment, showed no mortality (data not shown).

Estimated latent period and seizure rate

To determine if the CHX treatment had an effect on PILO-induced epileptogenesis, the relative latent period (i.e., the interval between PILO treatment and the observation of the 1st motor seizure) and the seizure rate were determined by both direct observation and 24-h video-monitoring. The average duration of observation of each animal was 160 h (50 h of direct observation and 110 h of video-monitoring). The saline-treated animals were observed with the other two groups but not included in this analysis because they were never seen to have seizures. The relative latent period is considered to be an estimate, since we did not video-monitor for 24 h/day, 7 days/wk. The relative latent period (Fig. 2A) was not significantly different between the PILO+CHX rats (31.5 ± 8 days), PILO-treated animals (49 ± 6 days), and the rats that had only a few seizures at treatment (48 ± 14 days, P > 0.05; ANOVA). The seizure rate (i.e., the number of observed motor seizures per total hours of observation; Fig. 2B) was not significantly different between PILO+CHX (0.074 ± 0.02 status epilepticus; n = 10) and PILO-treated rats (0.027 ± 0.007 status epilepticus; n = 10; P > 0.05; ANOVA; Tukey). The seizure rate for the PILO-treated animals was also not significantly different from the animals that had only a few seizures during treatment (0.0064 ± 0.002 status epilepticus; n = 4 of 9; P > 0.05; ANOVA; Tukey). However, the PILO+CHX-treated rats had a significantly higher seizure rate than the animals with only a few seizures at treatment (P < 0.05; ANOVA; Tukey). Only the data from the four animals that had 5 seizures at treatment (i.e., no status epilepticus) and later had spontaneous seizures were used in the rest of this study. These results support the hypothesis that CHX pre-treatment does not block or otherwise depress the course of epileptogenesis after PILO-induced status epilepticus.

View larger version (25K): [in this window] [in a new window]   Fig. 2. Relative latent period and seizure rate. A: relative latent period was based on an average of 160 h of both direct observation and 24-h video-monitoring. Average time to 1st observed seizure was not significantly different between the PILO+CHX-, PILO-treated, and rats with only a few seizures during treatment (P > 0.05, ANOVA). B: spontaneous chronic motor seizure rate for each animal was calculated using the same data as were used for calculating the relative latent period. Seizure rate was defined as the total number of class III, IV, and V seizures (Racine scale, 1972) observed during the total hours of observation. PILO+CHX rats had a significantly higher seizure rate than rats that had only a few seizures during treatment (P  0.05, ANOVA, Tukey). There was no significant difference between the PILO+CHX- and PILO-treated rats nor was there a statistical difference between the PILO-treated animals and rats with only a few seizures at treatment.

Timm stain in the IML

Timm stain in the IML of the dentate gyrus was analyzed throughout the extent of the septotemporal axis using the scale from Tauck and Nadler (1985) to assess the hypothesis that treatment with CHX prior to PILO-induced status epilepticus prevents the development of mossy fiber sprouting. Timm stain in the IML was not significantly different anywhere along the septotemporal axis between the CHX-pre-treated rats and the PILO-only animals (Figs. 3 and 4), and both groups were significantly different from a score of 0 (P < 0.05; Wilcoxan signed-rank test). No differences in Timm staining were seen between the recorded hippocampal slices and the intact hippocampus for individual rats. Both the saline-treated rats and the rats with only a few seizures at treatment had Timm scores that were not significantly different from a score of 0 (P > 0.05; Wilcoxan signed-rank test). However, the range for the rats with only a few seizures at treatment increased from 0 to 1 in the regions up to and including the 75% septotemporal distance to 0-2 in the 100% region. These data indicate that CHX had no detectable effect on mossy fiber sprouting in the IML. However, the data from those animals that had only a few seizures during PILO treatment suggested that substantial amounts of mossy fiber sprouting in the IML, as detected by the Timm stain method, were not necessary for the generation of spontaneous motor seizures.

View larger version (115K): [in this window] [in a new window]   Fig. 3. Timm stain in the inner molecular layer (IML). A: Timm stain in the IML of the dentate gyrus from a saline-treated rat (scale bars = 250 µm), and this was scored as a 0 on the Tauck and Nadler scale (1985). B: Timm stain from a PILO-treated animal. C: a PILO+CHX-treated rat. D: a rat that had only a few seizures during treatment. All sections are from approximately the same level of the septotemporal axis. For both the PILO- and PILO+CHX-treated animals, this amount of Timm stain in the IML was scored as a 3 using the Tauck and Nadler (1985) scale.

View larger version (13K): [in this window] [in a new window]   Fig. 4. Timm stain in the IML along the septotemporal axis. The septal end is defined as 25% and the temporal end as 100%. Every 6th section was sampled along the septotemporal axis. At no point throughout the extent of the hippocampus was there a significant difference in the average Timm score between the PILO+CHX rats and the PILO-treated animals, although both groups had significantly elevated Timm scores compared to the saline-treated rats at all points along the septotemporal axis. Rats that had only a few seizures during treatment did not have significant Timm stain in the IML compared to the saline-treated animals. Each point represents an average score and status epilepticus for the region from each treatment group. The top line represents the median scores for both the PILO+CHX and PILO animals that were identical to each other, and the bottom line represents the median scores for both the saline rats and the 4 rats with only a few seizures at treatment that had spontaneous motor seizures, which were also identical to each other. The range of scores for all regions in the PILO+CHX- and PILO-treated rats is 1-3. The range of scores for all regions in the saline and up to 75% septotemporal distance in the rats with few seizures at treatment is 0-1. The range in the 100% septotemporal region for the rats with few seizures at treatment is 0-2. Each point and the error bars represent an average and status epilepticus; this was done to compare these data to previous work by the authors.

Sterologic neuron counts

To address the hypothesis that pre-treatment with CHX prevents neuron loss during PILO-induced status epilepticus, we performed a stereologic estimate of neuronal populations in the hilus, CA1, and CA3 along the septotemporal axis. Hilar neuron loss was not significantly different between the PILO+CHX- and the PILO-treated animals. However, both the PILO+CHX- and PILO-treated rats had significant neuronal loss at the temporal end of the hippocampus (100%) compared to the saline-treated animals (Figs. 5 and 6A; P < 0.05; Tukey). The PILO+CHX-treated animals did not have significantly different neuron counts at 75% along the septotemporal axis compared to the saline-treated and PILO-treated rats, but the PILO-treated animals did have significant neuron loss in this region compared to the saline-treated animals. The PILO- and PILO+CHX-treated animals had significant neuronal loss in CA3 in the temporal region when compared to the saline-treated rats (Figs. 5 and 6C, P < 0.05; Tukey). The CA1 pyramidal cell layer showed significant neuron loss in both the PILO+CHX- and PILO-treated animals throughout the extent of the septotemporal axis compared to the saline-treated rats, and no significant difference was detected between the PILO+CHX and PILO treatments (Figs. 5 and 6B;  = 0.05; Tukey). The rats with few seizures at treatment were not significantly different from the saline-treated rats in any of the areas examined (i.e., hilus, CA3, and CA1). These data show that CHX had little or no detectable effect in preventing neuronal loss in the hilus (i.e., the 75% region of the hilus) and was not effective in areas CA3 and CA1 of the hippocampus. These data also suggest that detectable neuron loss (by the optical fractionator method) in the hilus, CA3, and CA1 is not required for the appearance of spontaneous motor seizures.

View larger version (131K): [in this window] [in a new window]   Fig. 5. Cresyl violet stain and neuronal loss. A-D: cresyl violet staining from a saline-treated rat (both scale bars = 300 µm; A, E, and I are low magnification, all others are high magnification). E-H: sections from a rat treated with PILO. I-L: a PILO+CHX-treated animal. F and J: neuronal loss in CA1 for both the PILO- and PILO+CHX-treated rats compared to the saline-treated rats in B. G and K: slight neuronal loss in CA3 in both the PILO and PILO+CHX treatments when compared to CA3 from the saline-treated rat in C. Hilar neuronal losses in both the PILO- and PILO+CHX-treated rats are shown in H and L, respectively, and the hilus from the saline-treated animal is represented in D. CHX pre-treatment did not appear to prevent neuronal loss in the hilus, CA1, or CA3.

View larger version (19K): [in this window] [in a new window]   Fig. 6. Neuron counts along the septotemporal axis. The septal-most end was defined as 25% and the temporal pole as 100%. Graphs of neuronal counts revealed no significant difference at any point along the septotemporal axis between the PILO+CHX-treated and the PILO-treated animals in the hilus (A), CA1 (B), and CA3 (C). Both the PILO+CHX- and PILO-treated animals had significant neuronal loss in the temporal area of the hilus and CA3 compared to the saline-treated rats (*significant difference). In the 75% region of the hilus (**), the PILO+CHX-treated rats were not significantly different from either the PILO- or saline-treated animals, but the rats treated with PILO were significantly different from the saline-treated rats. CA1 for both the PILO+CHX and the PILO treatments had significant neuronal loss compared to the saline treatment throughout the extent of the septotemporal axis. Rats that had only a few seizures during treatment did not have significant neuron loss compared to the saline-treated animals in any of the hippocampal areas examined.

Field-potential recordings from the dentate gyrus

The purpose of these studies was to test the hypothesis that new recurrent excitatory circuits would not be formed after CHX pre-treatment prevented mossy fiber sprouting in the IML. Therefore the dentate gyrus would hypothetically not show epileptiform activity under conditions of reduced inhibition and increased extracellular potassium with hilar (i.e., antidromic) stimulation. Abnormal network activity was defined as the presence of repetitive firing of population spikes, with a distinct threshold for burst discharge (i.e., all-or-none; Fig. 7). Often the initial burst was followed, after a variable latent period, with additional bursts (i.e., afterdischarges). Control responses consisted of a single population spike or multiple population spikes with a graded response (i.e., the number of population spikes increased with higher stimulus intensity, and no threshold could be determined; Fig. 8, A-E). In control solution, only a single population spike could be elicited from all the treatment groups. In 30 µM bicuculline, all of the hippocampal slices from saline-treated animals and from rats that had only a few seizures during treatment responded with single population spikes. However, four of nine PILO+CHX-treated rats and five of nine PILO-treated animals had hippocampal slices that showed bursts of population spikes when bathed in 30 µM bicuculline. When the extracellular potassium was raised from 3 to 6 mM, while still in 30 µM bicuculline, the hippocampal slices from saline-treated animals still did not demonstrate bursts of population spikes to hilar stimulation, but slices from 2 of 10 saline-treated rats did show a graded response to increasing stimulus intensity. In two of four of the rats that had only a few seizures during treatment, some of the hippocampal slices responded with a mixture of a graded response and small afterdischarges (Fig. 8F). In this same bathing solution, eight of nine of the PILO+CHX- (Fig. 9D) and seven of nine of the PILO-treated rats had hippocampal slices with all-or-none bursts of population spikes. When bursts of population spikes were found, 50 µM AP-5 and 50 µM DNQX were added to the bath to block N-methyl-D-aspartate (NMDA) and -amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors, respectively. In all cases (n = 11) where AP-5 and DNQX were added to the bath, population spike bursts were blocked, and the responses recovered when AP-5 and DNQX were washed from the slices (Fig. 9 E, F, and G). The hippocampal slices were processed for Timm stain after the recording, and the average Timm score was obtained in the same fashion as described in METHODS (n for recorded slices = 23 for PILO+CHX; n = 27 for PILO; n = 29 for saline). The duration of the population spike bursts was determined from responses recorded in 30 µM bicuculline and 6 mM extracellular potassium by measuring from the stimulus artifact to the last 0.5-mV population spike (at 4 times minimum stimulus), and the burst duration from each slice was averaged for each animal. All graded responses were eliminated from this analysis. The average burst duration was transformed into the natural log to reduce variance, and a correlation of the average burst duration and the median IML Timm-stain grade for each animal was performed. Burst duration correlated with Timm stain in the IML (r² = 0.6820; P = 0.0006; Fig. 10). Timm-stained sections from the recorded hippocampi for both the PILO+CHX- and the PILO-treated rats were significantly different from a hypothetical mean of 0 (P < 0.004, Wilcoxan signed-rank test), while recorded sections from the control rats were not significantly different from 0. These data suggest that CHX did not prevent the formation of new recurrent excitatory circuits in the dentate gyrus, and that as the Timm stain increased in the IML, so did the propensity to generate an epileptiform burst.

View larger version (26K): [in this window] [in a new window]   Fig. 7. Burst discharges in the dentate gyrus. Recordings in hippocampal slices from the granule cell layer of a PILO-treated rat with antidromic stimulation at 40 µA. Stimulation frequency was 0.2 Hz. The slice was bathed in 30 µM bicuculline and 6 mM [K+]o. A: initial burst during prolonged afterdischarges had a variable onset latency. B: expanded traces demonstrate the variable latency of burst onset (B1 and B2), plus occasional failure of the bursts even though the initial population spike was unchanged (B3). Identical responses were recorded in slices from PILO+CHX rats.

View larger version (17K): [in this window] [in a new window]   Fig. 8. Example of a graded response and a response with small afterdischarges from a rat with only a few seizures at treatment that subsequently had spontaneous motor seizures. A-E: graded response in 30 µM bicuculline and 6 mM [K+]o. Note that as the stimulus amplitude increased, so did the number of population spikes. This response was seen in control animals and in the rats with only a few seizures at treatment. Slices from saline-treated animals with this response typically only had 5-6 population spikes at the highest stimulus intensity. F: response from an animal with only a few seizures at treatment that demonstrated small afterdischarges occurring after the initial burst of population spikes, which were graded with stimulus intensity.

View larger version (18K): [in this window] [in a new window]   Fig. 9. Glutamate-mediated network bursts in a hippocampal slice from a PILO+CHX-treated rat. All of the traces shown are extracellular recordings from the granule cell layer with hilar stimulation at the maximum stimulus intensity (i.e., 400 µA). A and B: average of 5 responses from saline- and PILO+CHX-treated rats, respectively, in control solution. C: average of 5 responses from a saline-treated rat bathed in 30 µM bicuculline and 6 mM [K+]o. D: single response from a PILO+CHX-treated rat in the same solution. E: expanded version of D, F and G are from the same slice as E. F: blockade of the bursts with 50 µM AP-5 and 50 µM DNQX, and washout of the block is shown in G.

View larger version (11K): [in this window] [in a new window]   Fig. 10. Linear regression plot of burst duration vs. Timm stain in the IML. The burst duration (the time from the stimulus artifact to the last 0.5-mV population spike) was plotted against the median Timm score from the recorded slices. This plot includes data from all 3 groups (i.e., saline-, PILO+CHX-, and PILO-treated rats, but not rats that only had a few seizures at treatment). All responses were to hilar stimulation at the maximum stimulus. Hippocampal slices that produced a graded response (i.e., the number of population spikes increased as the stimulus intensity was increased, and multiple population spikes were not evoked in an all-or-none manner) to hilar stimulation were not included in the analysis. The right side of the graph shows the transformed numbers in the original format. Linear regression analysis shows r2 = 0.6820 and a P = 0.0006. Each point is the median Timm score of the hippocampal slices recorded from 1 rat and the average burst duration evoked from those slices. The line represents the linear regression. This graph indicates that as the average Timm score increased, the duration of the network bursts also increased.

Changes in spontaneous EPSC amplitude and frequency

The following electrophysiological experiments were aimed at testing two hypotheses: 1) PILO treatment results in an increase of the amplitude and the frequency of spontaneous EPSCs (sEPSCs) in granule cells, and 2) pretreatment with CHX blocks the PILO-induced increase in amplitude and frequency of sEPSCs. A whole cell patch-clamp analysis of the granule cells from rats with only a few seizures at treatment was not performed. Whole cell recordings were obtained from granule cells at their resting membrane potential, and the amplitude and frequency of sEPSCs was measured during one 120-s period for each cell (for representative examples of these types of recordings see Wuarin and Dudek 2001). The results from PILO-treated rats (n = 15 granule cells, 8 rats) were compared to the results obtained from PILO+CHX-treated animals (n = 10 granule cells, 5 rats) and to the results from saline-injected animals (n = 16 granule cells, 12 rats).

Cumulative amplitude distributions were constructed for each group and compared using the Kolmogorov-Smirnov (K-S) test. When compared with controls, the cumulative amplitude distribution for spontaneous EPSCs in granule cells from PILO-treated animals revealed a highly significant increase in the EPSC amplitudes (P < 0.001; 1-tailed; Fig. 11). Similarly, comparison between PILO+CHX-treated rats and controls showed significantly larger amplitudes in the PILO+CHX group (P < 0.001; 1-tailed; Fig. 11). However the K-S test showed no significant difference between the PILO group and the PILO+CHX group (P = 0.19; 2-tailed, Fig. 11). This result suggests that EPSC amplitudes were larger in the PILO-treated animals than in saline-injected controls, and that this increase was not blocked by pre-treatment with CHX.

View larger version (14K): [in this window] [in a new window]   Fig. 11. Cumulative probability plots of spontaneous excitatory postsynaptic currents (EPSCs) in granule cells. Pre-treatment with CHX did not prevent the increase in amplitude of spontaneous EPSCs after PILO treatment. Spontaneous EPSCs (n = 507) were pooled from 5 saline-treated rats, 5 PILO-treated rats (n = 714), and 5 PILO+CHX-treated rats (n = 1453). An example of these EPSCs in control and hippocampi from rats with kainate-induced epilepsy can be seen in Wuarin and Dudek (2001). The Kolmogorov-Smirnov test revealed no significant difference between the PILO and PILO+CHX group, but both groups were different from controls (see text for statistical analysis).

The mean EPSC frequency for each treatment group was saline = 0.26 ± 0.02, PILO = 0.43 ± 0.23, and PILO+CHX = 1.21 ± 3.70 (EPSC/s). The trend shown here was that the PILO and PILO+CHX groups had a higher frequency of sEPSCs compared to the saline-treated animals, but the three groups were not significantly different from each other (P = 0.06; ANOVA).

Flash photolysis

Local stimulation with application of glutamate microdrops and photoactivation of caged glutamate has been shown to evoke repetitive excitatory postsynaptic potentials (EPSPs) and EPSCs in hippocampal slices from kainate-treated rats with Timm stain in the IML (Lynch and Sutula 2000; Molnar and Nadler 1999; Wuarin and Dudek 1995, 2001). With flash photolysis of caged glutamate, we tested the hypotheses that 1) PILO treatment increases the number of granule cells showing repetitive EPSCs, when compared to controls, and 2) pretreatment with CHX prevents this increase.

Photostimulations were first applied directly on the recorded granule cell to assure that the flash of UV light was uncaging glutamate and producing action potentials. The flash was then moved away from the recorded cells in 150-µm steps, first on one side of the recorded cell and then on the other side, until reaching the tip of the granule cell layer. Each location was stimulated several times (>3 stimulations per location), and the presence or absence of repetitive EPSCs was assessed with whole cell recordings at resting membrane potential. The response to photostimulation was consistent for a given location (i.e., photostimulation evoked repetitive EPSCs every time or it never evoked repetitive EPSCs). Thus for each stimulation location, the response was positive every time or negative every time (Fig. 12; for a control response see Fig. 8A in Wuarin and Dudek 2001). Two of 21 granule cells tested in slices from saline-injected animals (9.5%) showed repetitive EPSCs in response to photostimulation. In contrast, photostimulation evoked repetitive EPSCs in a significantly larger proportion of granule cells tested from PILO- (13/16 granule cells, 81%; P < 0.002; ANOVA) and PILO+CHX-treated rats (7/10 granule cells, 70%; P < 0.002; ANOVA). There was no significant difference in the proportion of the granule cells responding to photostimulation with repetitive EPSCs between PILO- and PILO+CHX-treated animals (P = 0.6; ANOVA).

View larger version (27K): [in this window] [in a new window]   Fig. 12. Effect of flash photolysis of caged glutamate applied in the granule cell layer. A: examples are shown of photostimulation-evoked repetitive EPSCs in a granule cell from a PILO-treated rat and in a granule cell from a PILO+CHX-treated rat. Both traces show whole cell patch-clamp recordings at resting membrane potential (PILO = 72 mV; PILO+CHX = 69 mV, corrected for the liquid junction potential). The granule cell from the PILO-treated animal was located at the tip of the outer blade, and photostimulation was applied in the outer blade at a distance of approximately 300 µm. The granule cell from the PILO+CHX-treated animal was located in the apex, and the flash was applied approximately 450 µm away in the outer blade. Arrows show the artifact produced by the flash. B: plot of the percentage of granule cells showing repetitive EPSCs in response to photostimulation for each treatment group.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study provides evidence that in PILO-treated rats, a single pre-treatment injection with CHX does not 1) inhibit mossy fiber sprouting; 2) prevent neuronal loss in the hilus, CA1, or CA3; or 3) alter the abnormal electrophysiological responses that are commonly observed months after status epilepticus and associated with mossy fiber sprouting. However, the data also suggest that CHX interferes with the development of status epilepticus after PILO treatment, and that CHX may increase the mortality rate associated with PILO treatment. This study provides evidence that, independent of any effects of CHX, mossy fiber sprouting (detected by the Timm stain method) and measurable neuron loss in the hilus, CA3, and CA1 (assessed with the optical fractionator) are not necessary for the generation of spontaneous motor seizures after PILO-induced seizures.

Morbidity and mortality rate

One of the difficult aspects of this study was that a significant number of rats, when pre-treated with CHX, appeared to be resistant to the convulsant effect of PILO. Similarly, Williams and Jope (1994) reported that both CHX and anisomycin attenuate seizure induction in lithium/pilocarpine-induced status epilepticus. Their paper reported that the induction of the first spike train and status epilepticus are delayed by pre-treatment with protein synthesis inhibitors, and that in 20% of the rats pre-treated with CHX, neither spike trains nor status epilepticus were seen. When this effect was coupled with a threefold increase in the mortality rate for the PILO+CHX rats, it made the establishment of a suitable number of replications difficult. The studies by Longo and Mello (1997, 1998) do not report a physiologic effect of CHX on seizures or mortality rate, which may mean that their compound was not biologically active or was contaminated so that the effect reported by them was not due to inhibition of protein synthesis, as proposed by the authors. One possible explanation for the resistance to PILO treatment may be found in the effect that CHX has on the inositol triphosphate (IP3) signaling pathway, which appears to be the main route of seizure induction for PILO (for a review on this topic, see Savolainen and Hirvonen 1992). The rate-limiting steps in the formation of IP3 appear to occur during the phosphorylation of inositol monophosphate via the two enzymes phosphotidylinositol kinase (PI kinase) and phosphotidylinositol phosphate kinase (PIP kinase); the latter of these two enzymes is the most limiting in the process (Bazenet et al. 1990; Lundberg et al. 1985; Majerus et al. 1984; Singhal et al. 1994). A study by Weber et al. (1996) showed that the activity of PI kinase and PIP kinase was reduced to 15% and 12% of normal, respectively, in rat bone marrow 30 min after the animals were injected with CHX, and also that the IP3 levels were reduced to 50% of normal at this time period. The PI kinase and PIP kinase activity remained at or below these levels for at least 6 h, and the IP3 levels also remained at 50% of normal for at least 6 h. In this model, PILO was given approximately 40 min after CHX injection; thus, in these pretreated rats, the IP kinase and PIP kinase activity and IP3 levels were probably reduced prior to PILO injection, and threshold levels of these substances may be needed for PILO induction of seizures. The data presented here suggest this possibility, but further studies would be necessary for a rigorous analysis of this question. These data do indicate that the CHX did have a systemic effect on the animals, and that a reduction of protein synthesis (specifically PI kinase and PIP kinase) probably mediated the effect. Both of these enzymes have a high turnover rate, which would account for their reduced activity after CHX treatment (Weber et al. 1996). Cycloheximide, or other protein synthesis inhibitors, should probably be avoided in the PILO model because of the high potential for interfering with the IP3 signaling pathway. The overall effect of a threefold increase in mortality and a significant decrease in the induction of status epilepticus in the animals pre-treated with CHX may mean that these two epileptic groups were not equivalent from the outset in these experiments.

Latent period and seizure rate

The observation that the relative latent period and seizure rate were not significantly different for the PILO+CHX and the PILO treatment groups supports the hypothesis that CHX has no effect on epileptogenesis. A sub-population of rats responded to PILO treatment with only a few motor seizures (i.e., 1-2), regardless of whether they were pre-treated with CHX. One of the more interesting findings in this study was that when this sub-population of rats was subsequently monitored, some of them had spontaneous motor seizures several weeks after treatment. The seizure rate in these rats was low, but not significantly different from the PILO-treated rats, which was probably due to the small number of animals in the group that had only a few seizures during treatment but later also had spontaneous motor seizures (n = 4). Another factor was that the PILO-treated rats had not reached their maximum seizure rate by the end of the 2-mo observation period. Kainate-treated rats, for example, generally require several months to attain a maximum seizure rate (Hellier et al. 1998). Therefore if the rats had been used at 120 days and not 60 days after treatment, a high and stable seizure rate may have been reached, and the seizure rates in the PILO-treated rats that experienced status epilepticus versus those that had a few seizures at treatment may have been different.

Mossy fiber sprouting

Our finding that Timm stain is present in the IML of rats pre-treated with CHX is in contrast to what was found by Longo and Mello (1997). In their study, the majority of rats pre-treated with CHX did not demonstrate Timm stain in the IML. At least three potentially significant factors are relevant to the differences between these two studies. 1) We assessed Timm staining in every sixth section (i.e., at 210-µm intervals with 35-µm thickness for each section). The variation across sections was noted to be as great as one full grade in the Timm's scale of Tauck and Nadler (1985). The effect of this variation can be reduced by increasing the number of sections sampled. The level of tissue sampling in the Longo and Mello (1997) study was not reported. 2) We used a different strain of rat (i.e., Sprague-Dawley as opposed to Wistar) to compare our electrophysiological data from the PILO model to results with the kainate model. Strain differences exist in the magnitude of IP metabolism during PILO administration (for review, see Savolainen and Hirvonen 1992). The seizure response appears to be identical in the two strains, but the accumulation of inositol-1-phosphate (1 metabolic measure of the signaling pathway) can be five times higher in Sprague-Dawley rats compared to Wistar rats. Two strains of mice can have different responses in the amount of Timm staining in the IML after kainate treatment (Schauwecker et al. 2000), with one strain apparently showing no Timm stain in the IML following kainate-induced status epilepticus. Thus, strain differences can affect the response to status epilepticus and subsequent injury. Both our experiments and the study of Long and Mello (1997) did demonstrate Timm stain in the IML of the PILO-treated rat; therefore any possible strain differences in the response to PILO treatment are insufficient to offset the end result of mossy fiber sprouting and the chronic signs after PILO-induced status epilepticus. It remains possible that the differences in Timm stain in the IML found in the two studies could be because Sprague-Dawley rats may not respond to CHX pre-treatment in the same fashion as Wistar rats. 3) In our study, a saline control group was included to provide a baseline for Timm stain in the IML, which meant that a more rigorous multiple-comparisons test was the appropriate statistical analysis. However, when the saline group was not included, and a t-test was performed on the data, the CHX pre-treated rats were still not significantly different from the PILO-treated rats. One technical aspect of Timm staining that could also account for some of the differences seen between these two studies is the possibility of false negatives with Timm stain in the IML.

For CHX to block mossy fiber sprouting, two assumptions must be made. The first assumption is that the signals for mossy fiber sprouting occur in the first 24 h after SE. GAP-43 is a potential signal for mossy fiber sprouting that increases after PILO-induced status epilepticus, and CHX can prevent the rise of GAP-43 mRNA, which occurs after kainate treatment, as shown by Bendotti et al. (1996). However, it should not be assumed that the effect of a single dose of CHX on protein synthesis lasts any longer than 8 h. Jonec and Wasterlain (1979) showed that protein synthesis was inhibited by 75% about 2 h after injection of CHX at 1.5 mg/kg. Another study reported that 15 mg/kg of CHX (a 10 times higher dose) reduced the expression of c-fos and c-jun protein after kainate injection up to 4 h after treatment, but that protein expression of both c-jun and c-fos were both increased above control levels by 8 h, although not to the levels of the kainate-treated animals (Won et al. 1997). This short-term suppression of protein synthesis on signaling pathways and regulation of mRNA levels can be longer lasting (Bendotti et al. 1996), because 2 mg/kg of CHX prior to kainate treatment can reduce the surge of GAP-43 mRNA for 24 h. Also, c-fos mRNA levels were reduced for 16 h in kainate-treated animals if they were pre-treated with 2 mg/kg of CHX (Schreiber et al. 1993). In both of these studies, the actual GAP-43 and c-fos protein levels at these time points was not reported. In the studies presented above, none of the data suggest that the direct effect of CHX lasts any longer than 12 h, and the indirect effects last no longer than 24 h. These reports also indicate that even at higher doses of CHX, protein synthesis is not completely blocked. The second assumption that is made is that CHX will block synthesis of a specific signaling protein. In in vitro synaptoneuronal preparations, low levels of CHX could reduce total protein synthesis, but synthesis of CamII was actually increased at the same time (Scheetz et al. 2000). Thus it is not always the case that CHX blocks increases in the concentration of all proteins. Therefore, based on this earlier research, this protocol would not be expected to block protein synthesis completely, and the blockade would be short-lived. Furthermore, it may not specifically block the increase in a signaling protein; therefore the experimental and conceptual basis for proposing that CHX would block mossy fiber sprouting is only practical if a critical signal for sprouting occurs only in the 24-h period after status epilepticus.

A preliminary finding in this study was that the rats that had subsequent spontaneous motor seizures, but only a few seizures at treatment did not have increased amounts of Timm stain in the IML compared to the saline controls. However, the temporal region of the hippocampus did have a greater variability in the amount of Timm stain seen in the IML of these animals. Thus if a larger number of animals had been studied, it is possible that the amount of Timm stain seen in the IML of the rats with only a few seizures at treatment may have been significantly increased above the saline-treated rats.

Neuronal loss

The finding that pre-treatment with CHX did not prevent neuronal loss in the hilus (although it may have had a minimal protective effect in the hilus at 75% along the septotemporal axis), CA1, or CA3 was similar but not identical to what was found by Longo and Mello (1997). They reported no significant difference between the PILO+CHX and PILO treatments in hilar and CA1 loss, but CA3 was significantly preserved in the PILO+CHX rats. This is in contrast to our finding, in which CA3 loss was not significantly different between the two treatment groups. There are two potential reasons for this difference: 1) the method of neuronal counting and the sampling frequency could have been different, and 2) this study includes baseline data from saline-injected animals. It is unknown what method was used by Longo and Mello (1997) to estimate neuronal counts, nor is it clear how frequently the septotemporal axis was sampled. The present study used the optical fractionator method to derive neuronal counts (West 1999; West et al. 1991), which was not utilized by other studies that have reported CHX prior to kainate treatment may prevent neuronal loss (Schreiber et al. 1993). We also sampled more extensively than Buckmaster and Dudek (1997b) to reduce the inter-sectional variance for more accurate counts. The use of saline controls was not reported by Longo and Mello (1997), and this lack of baseline data to compare neuronal counts could influence the overall statistical analysis. An example of this can be seen in our data. When the saline data are removed and a t-test is utilized to compare the PILO+CHX and the PILO data, they are significantly different in the septal 50% of the hilus and septal 25% in CA3 (P = 0.05 and 0.04, respectively). When the multiple comparison test was used, which incorporates the baseline data from the saline rats, the differences for these two regions are not significant. This indicates that if saline controls would have been used in the Longo and Mello (1997) study, then their results might not have been significant.

An interesting, but preliminary, finding of this study is that neuronal loss was not significantly different from the saline controls in any of the areas examined in those animals that had only a few seizures during treatment and that subsequently developed spontaneous motor seizures. This implies that animals that experience a few motor seizures during PILO treatment, and that do not have obvious hippocampal injury and subsequent mossy fiber sprouting as detected by the Timm stain method, may still potentially become epileptic. The optical fractionator method seeks to reduce the inherent error in neuronal counting by utilizing unbiased techniques (for a review, see West 1999). However, it is possible that this technique is not sensitive enough, because it was applied to the relatively small sample of rats used in this part of the study, to detect small differences in neuronal populations that may be present in the rats pre-treated with CHX or in the rats with only a few seizures at treatment. These findings also do not preclude the possibility that certain subpopulations of neurons in the hilus (i.e., mossy cells) or in CA3 and CA1 were lost but not detected by the quantitative methods employed (i.e., we did not perform immunocytochemistry to detect these subpopulations). It is also possible that basilar dendrites of granule cells have formed and are contributing to synaptic reorganization (Buckmaster and Dudek 1999; Ribak et al. 2001; Spigelman et al. 1998), but we would not be able to detect these by the methods employed in this work.

This study utilized systemic PILO injection that likely damaged other areas of the brain (multifocal epilepsy; amygdala, subiculum, entorhinal cortex, and the thalamus), but we did not observe other possible lesions because we only examined the hippocampus. Since this study did not use a focal model of hippocampal injury, we cannot specifically address the effects a protein synthesis inhibitor may have on mossy fiber sprouting during focal injury and reorganization. However, we propose that synaptic reorganization occurs in other areas of the brain following neuronal loss after a variety of etiologies (e.g., see Smith and Dudek 2001 for discussion in regard to the CA1 area). If this mechanism is occurring in other areas of the brain following status epilepticus induced by PILO, then CHX should also hypothetically block synaptic reorganization in these areas. Therefore this approach would possibly address the general question of the importance of synaptic reorganization in motor seizure generation, regardless of whether the lesion/epileptogenic zone was focal or multifocal.

Electrophysiology

The premise behind these experiments was that pre-treatment with CHX would prevent mossy fiber sprouting, and therefore, the granule cell layer under conditions of reduced inhibition and elevated extracellular potassium would have responses to antidromic stimulation that were similar to saline controls. This would have supported the hypothesis that the all-or-none bursting found under these conditions in the PILO- and kainate-treated rats (Patrylo and Dudek 1998; Wuarin and Dudek 1996) was due to the mossy fiber sprouting. However, CHX did not prevent mossy fiber sprouting, and we found that the PILO+CHX-treated rats had responses that could not be distinguished from the PILO-treated rats. These results in the PILO model of temporal lobe epilepsy thus support the previous findings from our laboratory in the kainate model. Furthermore, all-or-none burst duration was strongly correlated with Timm stain in the IML, which has not been shown before. One of the possible reasons for this finding is due to our using the rats at a fairly early time point after PILO-induced status epilepticus. Wuarin and Dudek (2001) found that the temporal progression of Timm stain in the IML, which appears to reflect axon growth in the dentate gyrus, was associated with new recurrent excitatory synapses beginning within the second week after kainate-induced status epilepticus, and that this progression proceeded for several months after kainate treatment. Thus we may have been sampling epileptic rats that were at different levels of synaptic reorganization, which implies that as granule cell interconnections increase, so does the ability of the network to sustain an all-or-none burst under conditions of reduced inhibition and elevated potassium.

In the hippocampal slices from two of the four rats that had only a few seizures during treatment, but had spontaneous seizures later, electrophysiological recordings revealed mildly hyperexcitable slices. This might have been due to a small amount of mossy fiber sprouting that was associated with enough new excitatory synapses to cause hilar-evoked EPSPs and action potentials. Other mechanisms, such as formation of basilar dendrites or alterations in glutamate receptors, could also potentially explain these findings.

The whole cell patch-clamp recordings demonstrated that slices from both the PILO- and PILO+CHX-treated rats had increased EPSC amplitudes compared to the slices from the saline-treated animals. This finding is consistent with what has been found by Simmons et al. (1997) and Wuarin and Dudek (2001), where the amplitude and frequency of the EPSCs were found to increase with the density of Timm stain in the IML. This study did not find a significant increase in the frequency of EPSCs in granule cells from either PILO- or PILO+CHX-treated rats, but there was a general trend towards a higher frequency in these two groups. This general trend may have been significant if more neurons were sampled for this study. These data suggest that excitatory drive onto the granule cells in the PILO- and PILO+CHX-treated animals was increased. This increased excitatory drive could be from new recurrent excitatory circuits or from other neurons in the slice. Postsynaptic changes in glutamate receptor type and density could also account for these changes in the EPSC amplitude.

Focal flash photolysis of caged glutamate has been used to map local neuronal circuitry (Callaway and Katz 1993; Dalva and Katz 1994; Katz and Dalva 1994). It has also been shown that glutamate microdrops and photostimulation of caged glutamate in the granule cell layer could evoke EPSPs and EPSCs in recorded granule cells in tissue from kainate- and PILO-treated rats (Lynch and Sutula 2000; Molnar and Nadler 1999; Wuarin and Dudek 1996, 2001). The results reported here from a blind study support these earlier findings, and further suggest the formation of new recurrent excitatory circuits between granule cells.

The main conclusions of this study are as follows: 1) CHX pre-treatment has considerable complicating effects that appear to invalidate an analysis of synaptic reorganization and hippocampal histopathology and their relation to chronic epileptogenesis; 2) CHX did not block mossy fiber sprouting, and it had virtually no effect on hippocampal histopathology; and 3) although a preliminary result, it appears that chronic motor seizures can occur, albeit infrequently, after treatment with PILO (with or without CHX) independent of any obvious alterations in hippocampal histopathology. It is conceivable that significant changes did occur in the hippocampus, but they were not detected by cell counts or the Timm stain method, although the preliminary electrophysiological results suggested subtle changes in the granule cell network.