Structural and functional MRI following 4-aminopyridine-induced seizures: A comparative imaging and anatomical study

www.elsevier.com/locate/ynbdi Neurobiology of Disease 21 (2006) 80 – 89

Structural and functional MRI following 4-aminopyridine-induced seizures: A comparative imaging and anatomical study P.F. Fabene, a,* R. Weiczner, b P. Marzola, c E. Nicolato, a L. Calderan, a A. Andrioli, a E. Farkas,b Z. Su¨le,b A. Mihaly,b and A. Sbarbati a a

Section of Anatomy and Histology, Department of Morphological and Biomedical Sciences, Faculty of Medicine, Strada Le Grazie 8, 37134 Verona, Italy Department of Anatomy, Faculty of Medicine, Albert Szent-Gyo¨rgyi Health Science Center, University of Szeged, Hungary c Center of Experimental Magnetic Resonance Imaging, University of Verona, Verona, Italy b

Received 28 April 2005; revised 11 June 2005; accepted 16 June 2005 Available online 9 August 2005 Structural and functional MRI was used in conjunction with computerized electron microscopy morphometry to study changes 2 h, 24 h and 3 days after 4-aminopyridine-induced seizures lasting 2 h in rats. T2 (relaxation time) values showed changes throughout the cerebral cortex, hippocampus, amygdala and medial thalamus, with a different temporal progression, showing a complete recovery only after 3 days. Two hours after seizures, the apparent diffusion coefficient was decreased throughout the brain compared to control animals, and a further decrease was evident 24 h after seizures. This was followed by a complete recovery at 3 days post-seizures. Functional MRI was performed using regional cerebral blood volume (rCBV) maps. The rCBV was increased shortly after convulsions (2 h) in all structures investigated, with a significant return to baseline values in the parietal cortex and hippocampus, but not in the medial thalamic nuclei, 24 h after seizure onset. No rCBV alterations were detected 3 days after seizures. Electron microscopy of tissue samples of parietal neocortex and hippocampus revealed prominent astrocytic swelling 2 h post-convulsions which decreased thereafter gradually. In conclusion, this experiment reports for the first time structural and functional brain alterations, lasting several hours, in 4aminopyridine-treated rats after seizure onset. MRI approach combined with histological and ultrastructural analysis provided a clarification of the mechanisms involved in the brain acute response to ictal activity. D 2005 Elsevier Inc. All rights reserved. Keywords: Convulsions; Epilepsy; Edema; Cerebral cortex; Temporal lobe; Parietal cortex; Hippocampus; Medial thalamus; Histology; Electron microscopy

Introduction Recent advances in magnetic resonance imaging (MRI) have had considerable impact on clinical and research imaging in epilepsy (Riederer et al., 1995; Fabene and Sbarbati, 2004). In the

* Corresponding author. Fax: +39 045 8027 163. E-mail address: [email protected] (P.F. Fabene). Available online on ScienceDirect (www.sciencedirect.com). 0969-9961/$ - see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.nbd.2005.06.013

present study, MRI analysis was performed using T2-weighted (T2W) images and diffusion-weighted imaging (DWI) in order to study the temporal evolution of damage after 4-aminopyridine (4AP)-induced seizures. Previous MRI studies based on the rat models of temporal lobe epilepsy (TLE) showed alterations in T2W images after intrahippocampal injection of kainate (Bouilleret et al., 2000), in the lithium-pilocarpine model of epilepsy (Roch et al., 2002) and after pilocarpine-induced epilepsy (Fabene et al., 2003). Previous studies on experimental TLE induced by pilocarpine or kainic acid emphasized the sensitivity of DWI for the detection of the consequent early brain damage (Nakasu et al., 1995; Wang et al., 1996; Wall et al., 2000). We therefore utilized DWI and T2W RARE (Rapid Acquisition Relaxation Enhancement) imaging for the study of acute brain damage. Functional analysis was performed by the analysis of the regional cerebral blood volume (rCBV) (Mandeville et al., 1998; Hamberg et al., 1996) combining the structural data with functional approaches for the study of rCBV changes. In this study, an ultrasmall superparamagnetic iron oxide contrast agent (USPIO), characterized by an extended plasma half-time (Mandeville et al., 1998; Sbarbati et al., 2000), was used. The advantage of this technique is that it can be used to create parametric maps that monitor the perfusion state in different brain regions. We used these techniques to study the structural and functional changes induced by prolonged seizures. A single systemic injection of the K+ channel antagonist 4-aminopyridine (4-AP) produces brief behavioral seizures in experimental animals (Fragoso-Veloz et al., 1990; Miha´ly et al., 1990, 2005). The effects of 4-AP include delayed presynaptic repolarization and increased transmitter release (Thesleff, 1980) and result in neuronal hyperactivity and paroxysmal depolarization shifts in cerebrocortical neurons (Perrault and Avoli, 1991). In the hippocampus, 4-AP increases the release of glutamate (GLU) and GABA in a dose-dependent manner and causes typical convulsive electroencephalographic (EEG) changes (Pen˜a and Tapia, 2000). It has been reported that 4-AP-induced seizures increase the regional cerebral blood flow (rCBF) measured by 99mTc-HMPAO autoradiography (Miha´ly et al., 2000).

P.F. Fabene et al. / Neurobiology of Disease 21 (2006) 80 – 89

81

Because a comparison between MRI findings and histological results are lacking, we investigated brain alterations following 4AP-induced convulsions at different time points after seizures (2 h, 24 h and 72 h) by in vivo MRI analysis and compared the MRI results to ultrastructural alterations of the cerebral cortex. The comparison between ultrastructural and in vivo functional MRI alterations in an experimental acute seizure model in which repeated seizures are not followed by chronic epilepsy is important because several idiopathic epilepsies begin with seizures followed by spontaneous epilepsy (Shinnar and Hauser, 2002). The 4-AP model also offers the possibility of comparison with previously investigated alterations induced by status epilepticus (SE) in animal models of chronic epilepsy (Bouilleret et al., 2000; Roch et al., 2002; Fabene et al., 2003).

T2W images were acquired using a RARE sequence with the following parameters: repetition time (TR) = 5117 ms; echo time (TE) = 65 ms, RARE factor = 8; field of view (FOV) = 4
4 cm2, matrix size 256
256 corresponding to an in-plane resolution of 156
156 Am2. In all animals, one slice was imaged using a multiecho spin-echo sequence with 8 echoes, TR = 2000 ms and TE ranging from 25 to 200 ms for quantitative T2 mapping. The selected slice was also imaged using a diffusion-weighted spin-echo sequence, with TR = 1500 ms, TE = 60 ms, two b-values (6 and 1034 s/mm2) and matrix size 128
64 zero-filled at 128
128. Images were analyzed using the software ParaVisionR 2.1.1 (Bruker, Germany). For DWI, the apparent diffusion coefficient (ADC) maps in the brain were obtained pixel by pixel by two point linear fit of the logarithm of signal intensity (M) versus the bfactors according to the expression:

Materials and methods

M ðbÞ ¼ M0 Texpð ADCTbÞ

Animals and seizure induction

where M 0 represents the signal intensity in the absence of diffusive gradients. Analogously, T2 maps were obtained by a least squares non-linear exponential fitting of the signal intensity to the expression:

Male adult Wistar rats (80 – 90 days of age) were kept under controlled environmental parameters and veterinarian control. The animals were habituated to handling for at least 2 weeks prior to the procedures employed in the present study. The experiments received authorization from the Italian Ministry of Health and conformed with the principles of the NIH Guide for the Use and Care of Laboratory Animals and the European Community Council (86/609/EEC) directive. All efforts were made to minimize the number of animals used and avoid their suffering. Rats were randomly divided into two groups: in thirty rats, seizures were induced by 4-AP as described in previous studies (Miha´ly et al., 2001). Seizures were elicited with a single intraperitoneal (i.p.) bolus of 4-AP (5 mg/kg 4-AP; Sigma Chemical Co., St. Louis, MO), dissolved in physiological saline (1 mg/ml). Control animals (n = 30) received the same volume of physiological saline i.p. MRI analysis was performed 2 h, 24 h and 3 days after seizure arrest. Twelve animals for both conditions were analyzed with structural (T2W and DWI) and functional (rCBV) MRI and then processed by histological analysis (Fig. 1). Eighteen animals per condition were analyzed directly by ultrastructural or histological procedures (Fig. 1). Magnetic resonance imaging Structural MRI MRI experiments were performed using a Bruker Biospec Tomograph (Bruker Medical, Ettlingen, Germany) equipped with an Oxford, 33 cm bore, horizontal magnet operating at 4.7 T and a Bruker gradient insert (maximum intensity 20 G/cm). The rats were anesthetized by inhalation of a mixture of oxygen and air (1 l/min) containing 1% of halothane (initial dose 4% halothane) 2, 24 and 72 h after seizure onset; rectal temperature and heart beat rate were monitored by Biotrig physiological monitor (Bruker, Germany) and were similar in control and experimental animals. The rats were placed into a 7.2 cm i.d. bird cage transmitter coil. The signal was received through a 2 cm surface coil, actively decoupled from the transmitter coil and placed directly on the animal’s head. Three mutually perpendicular slices were acquired through the brain as scout images. Five contiguous, transversal, T2w, 2-mm-thick slices were imaged starting 1 mm posterior to the olfactory bulbs.

M ðtÞ ¼ M0 Texpð  t=T2Þ where M 0 represents the initial signal intensity. ADC and T2 values were extracted from the relative maps using the region-of-interest (ROI) method: 5 square ROIs of 2
2 pixels were defined in every selected brain region for each side. In each selected brain region, the variation of ADC value between the control and experimental animals was evaluated as follows: DADC% ¼

ADCseizure ADCcontrol % ADCseizure

Regional CBV Contrast medium. SineremR, an USPIO particle (kindly supplied by Guerbet, Aulnay-Sous-Bois, France), was used as contrast agent. SineremR is constituted by an iron-oxide core of about 6 nm diameter coated by dextran (coated particle dimensions of about 20 nm) and is characterized by a blood-half time longer than 2 h in rats (Dousset et al., 1999). SineremR was dissolved in saline and injected at a 6 mg/kg dose (mg is referred to iron). Animals were anesthetized as indicated above, and the tail vein was catheterized for contrast medium administration. Regional blood volume mapping. A RARE sagittal acquisition was performed in order to localize the olfactory bulb. Starting 1 mm posterior to it, 5 slices, 2-mm-thick, were acquired axially, using a RARE sequence with the following parameters: TR/TE, 5117/65 ms; FOV, 4
4 cm2; matrix 256
128 zero-filled at 256
256; RARE factor, 8. Three slices were chosen for rCBV mapping. After localized shimming over these slices, a gradient-echo sequence was acquired before contrast medium injection and, 2 min later, with the following acquisition parameters: TR/TE, 200/15 ms; FOV, 4
4 cm2; matrix 256
128 zero-filled at 256
256; slice thickness, 2 mm. Due to the presence of air, rCBV maps could not be acquired in the brain areas located in the plane of the internal acoustic meatus, such as the amygdala and piriform cortex. Image analysis. After injection of a superparamagnetic contrast agent, the T2* relaxation rate in the brain decreases (compared to

82

P.F. Fabene et al. / Neurobiology of Disease 21 (2006) 80 – 89

Fig. 1. Experimental design and animals used in each experimental groups.

the pre-injection baseline) proportionally to the local CBV times a certain function of the plasma concentration f( P) of the contrast agent (Mandeville et al., 1998):

signal intensity (SI) before and after contrast medium ln(SI(pre) / SI(post) to CBV: InðSIðpreÞ=SIðpostÞÞ ¼ ðTEð DR2TÞÞ ¼ KCBV

DR2T ¼ kf ð PÞCBV ¼ KVCBV where the last equation holds when the agent reaches a plateau concentration in the blood, as in the case of SineremR. For a gradient-echo sequence with echo time TE, the following equation correlates the natural logarithm (ln) of the ratio between

This equation shows that alterations in the local CBV can be probed by MRI acquisitions. In this work, we have normalized the rCBV values obtained in brain by the value obtained in muscle. A similar approach has been recently used (Bremer et al., 2003) in a study aimed at investigating blood volume in an experimental tumor model.

P.F. Fabene et al. / Neurobiology of Disease 21 (2006) 80 – 89

83

Maps of rCBV were obtained using the algebra tool of ParaVisionR. Briefly, after absolute reconstruction of images, background noise below threshold value (fixed by the ParaVisionR software) was cut off, as routinely done. The natural logarithm of the ratio between the SI before and after contrast medium administration (R value) was then calculated pixel by pixel and displayed. Values of rCBV were extracted using ROIs. Pseudocolor images were obtained using Matlab software (Mathworks Inc., Natick, MA).

contacting the basal lamina was measured on digitized images (Soft Imaging Systems, Mu¨nster, Germany). The increase of the perivascular astrocyte area was expressed as percentage of the control values. The lumen diameter of the microvessels was measured with the same software. The number of endothelial mitochondria per capillary profile was counted directly on the electron microscopic screen.

Histology and ultrastructural analysis

In the present study, a mixed experimental design has been used. All animals examined by MRI were used for histological or ultrastructural analysis (within design). Furthermore, in order to increase the n value of the groups, further animals have been sacrificed (with no previous MRI analysis) and the brains processed for histological or ultrastructural analysis (between design). However, it should be emphasized that no significant differences were noticed between MRI-analyzed and non-MRIanalyzed animals of the same group. MRI and electron microscopy data were analyzed for the statistical evaluation by the mean of the software SPSS. For MRI data, difference between DWI, T2W and rCBV values obtained in control vs. 4-AP-treated rats was evaluated with one-way analysis of variance (ANOVA) for repeated measures followed by the LSD post-hoc test, setting the significance at P < 0.05. The same test was used for histological data.

Histology At the end of MRI analysis, anesthesia was implemented with barbiturates (Tionembutal, 50 mg/kg, i.p.), and three rats per group were perfused via the ascending aorta with phosphatebuffered saline (PBS) followed by 4% paraformaldehyde in 0.1 M phosphate buffer, pH 7.4 (PB). The brains were then dissected, embedded in paraffin and 7-Am-thin coronal sections were cut. The sections were stained with hematoxylin – eosin for routine histological examination. Astrocytic swelling and neuronal damage were observed and evaluated according to Evans et al. (1983), with a Nikon Eclipse 600 microscope equipped with a SPOT RT Slider digital camera. The occurrence of dark neurons was evaluated in the area of interest (AOI; 1 mm2) of the camera, as follows: 0 – 5 dark cells (+); 6 – 15 dark cells (++); 16 – 30 dark cells (+++). The astrocytic swelling was judged on the basis of the vacuolation of the white matter: slight vacuolation (+); medium vacuolation (++); severe vacuolation (+++). Astrocytic swelling of the gray matter was described as perivascular vacuolation in every case. Electron microscopy Samples of the left parietal cortex and hippocampus (at the level corresponding to AP 3.60 mm from bregma in the rat atlas of Paxinos and Watson, 1986) were prepared for electron microscopy (Farkas et al., 2003). Following transcardial perfusion fixation (2% paraformaldehyde, 2.5% glutaraldehyde in 0.1 M phosphate buffer) in 3 animals per experimental group, tissue blocks were incubated in an aqueous solution of 1% OsO4 and 5% K2Cr4O7 (1:1) after thorough rinsing. The samples were dehydrated, incubated in 1% uranyl acetate and embedded in Durcupan epoxy resin (Fluka, Buchs, SG, Switzerland). Semithin sections were cut on an ultramicrotome (Ultracut E, ReichertJung, Vienna, Austria) and stained on object glasses with a 1:1 mixture of 1% methylene blue and 1% Azure II blue. The samples were then coverslipped with DPX and analyzed under a light microscope (Nikon E600, Nikon Co., Tokyo, Japan). Ultrathin sections were cut of the same blocks and collected on 200 mesh copper grids. The preparations were then contrasted with 5% uranyl – acetate and Reynolds lead – citrate solution. Finally, the samples were analyzed by a Philips TM10 transmission electron microscope (Eindhoven, Netherlands). Photographs were taken with a computer assisted digital camera (MegaView II, Soft Imaging Systems, Mu¨nster, Germany). Quantitative analysis was conducted on samples from the parietal cortex. Approximately 900 Am2 tissue surface was scanned systematically through all neocortical layers, and 15 T 4 capillary cross sections were examined in each case. Capillary density was calculated for a standard surface area with the help of the sample grid. The area of the perivascular astrocytic endfeet

Statistical evaluation

Results 4-AP-induced seizures Rats treated with 4-AP exhibited the typical behavioral changes described in previous reports (Miha´ly et al., 1990; Szaka´cs et al., 2003). The i.p. administration of 4-AP caused characteristic behavioral symptoms within 15 min: tremor of the vibrissa and masticator muscles followed by generalized tremor of the body musculature, detectable as continuous fasciculation of the muscles, and generalized tonic – clonic seizures (GTCS). The symptoms were always sudden and clear-cut. The main behavioral signs disappeared after 90 – 120 min, and the animals then displayed mild tremor or brief myoclonic episodes for up to 3 h. All animals exhibit the same behavioral pattern and severity of the seizures. During seizure induction, 10 animals died and were excluded from the study. Hyperventilation was occasionally observed during seizures but ended after their arrest, and all animals exhibited normal spontaneous breathing at the time of MRI analysis. Structural MRI T2 maps Quantitative T2 maps calculated from multiecho spin-echo acquisitions (Figs. 2E – H) clearly showed the overall pattern of changes in control and epileptic rats (both at 2 h and 24 h), highlighting the signal increase throughout the cerebral cortex. The T2 value was also slightly increased in the amygdala. In the diencephalon, selective increase in the T2 signal was documented in medial thalamic regions. T2 values exhibited a 56% increase (P < 0.001) in the parietal cortex at 2 h vs. control; this increase was still present (P < 0.01) at

84

P.F. Fabene et al. / Neurobiology of Disease 21 (2006) 80 – 89

Fig. 2. Parametric maps showing diffusion-weighted images in control (A) and experimental (2 h, 24 h and 3 days after seizures, B – D, respectively) conditions. T2 parametric maps are represented in E (control), F (2 h), G (24 h) and H (3 days after seizure onset). Regional cerebral blood volume maps are shown in the three different experimental conditions (I = control, J = 2 h, K = 24 h and L = 3 days after seizure). (M – O) Graphic representations of the abovementioned parameters in different brain structures. Boxes in B, E, I represent the different ROIs used in the data analysis; squares are drawn only in one side, while the data acquisition was performed on both sides. Blue bars represent the medial thalamus, green bars represent the hippocampus, brown bars represent temporal cortex, while red bars represent parietal cortex in panels M – O.

24 h (+29%). A dramatic increase in the T2 values was found also in the hippocampus after 2 h (+72%, P < 0.001) and after 24 h (25%, P < 0.01). The temporal cortex showed the same pattern of alteration (+97%, P < 0.0001 at 2 h; + 39%, P < 0.001 at 24 h). A highly significant (P < 0.001) increase in T2 (52% at 2 h and 33% at 24 h) was also found in the medial thalamus. In all of these structures, the T2 values observed 24 h after seizures were significantly higher than those observed in the controls (P < 0.001). No evident differences were detected 3 days after seizure induction (Figs. 2H, N).

Diffusion-weighted imaging Diffusion-weighted images, based on the sensitization of the MRI signal to the Brownian motion of water molecules, reveal maps of brain tissue water motion, resulting in hyperintense alterations. ADC maps of brain tissue water, calculated from DWI, showed consistent changes in ADC in the cerebral cortex (parietal cortex), hippocampus and amygdala of the epileptic animals at 2 h and 24 h (Figs. 2B– C) compared with control ones (Fig. 2A). An evident reduction of the average ADC value was documented in the parietal (44%) and temporal (34%)

P.F. Fabene et al. / Neurobiology of Disease 21 (2006) 80 – 89

85

Fig. 3. Light microscopic images (hematoyxlin – eosin staining) of control (A, C, E) and convulsing animal neocortex (B), hippocampus (D) and fornix (F) 2 h after 4-AP administration. (A) Control parietal neocortex. Scale bar: 100 Am. Insert: semithin section of the control parietal neocortex. Original magnification: 40
. Scale bar: 100 Am. (B) Parietal neocortex, 2 h after 4-AP injection. Note the dark cell somata (arrows). Scale bar: 100 Am. Insert: semithin section of the parietal neocortex, 2 h after 4-AP administration. Note the perivascular edema (arrowheads). Original magnification: 40
. Scale bar: 100 Am. (C) Control hippocampus with the hilus and lower blade of the dentate gyrus. Scale bar: 100 Am. (D) Hippocampus, 2 h after 4-AP injection. Note the edematous areas (arrows). Scale bar: 100 Am. (E) Control image showing the dorsal fornix (dF). Scale bar: 100 Am. (F) Dorsal fornix, 2 h after the 4-AP injection. Note the edematous changes (arrows). Scale bar: 100 Am.

86

P.F. Fabene et al. / Neurobiology of Disease 21 (2006) 80 – 89

Table 1 Structural and functional alterations of brain areas 2 h following 4-AP seizures

M). ADC values returned to basal levels 3 days after seizure onset (Figs. 2D, M).

Brain structure

Edema (vacuolation)

Neuron damage (dark cells)

rCBV changes

ADC values

Functional MRI

Parietal neocortex Hippocampus Fimbria – fornix

++ ++ +++

+ ++

Subcortical white matter Medial thalamus Amygdala and temporal cortex

++

+32% +16% Not measured Not measured +44% None

44% 23% Not measured Not measured 11% 34

+ +

None +

The structural alterations were evaluated on 7-Am-thin hematoxylin – eosinstained sections (see Results section).

cortices, hippocampus (23%) and medial thalamus (11%) 2 h after 4-AP-induced seizures (Figs. 2B, M). A further drop of ADC value was documented 24 h after 4-AP injections in the hippocampus (46%), medial thalamus (21%) and in the cerebral cortex (parietal: 57%; temporal: 64%) (Figs. 2C,

rCBV maps Quantitative parametric maps reconstructed by images acquired with gradient-echo sequence before and after USPIO administration in control and epileptic animals at 2 h and 24 h (Figs. 2I – K) showed an increase of average blood volume value in the early stages of the seizures, with a progressive normalization of the values at 24 h (Fig. 2K) in the parietal cortex and hippocampus. In the medial thalamus, the value failed to return to baseline 24 h post-convulsions ( P < 0.001). Two hours after seizure onset, the values were higher in the parietal cortex (+32%, P < 0.0001) and in the hippocampus (+16%, p<0.01) versus control. At subcortical levels, rCBV increased in the medial thalamus (+44%, P < 0.0001). Twentyfour hours after seizures, rCBV showed marginal alteration in the parietal cortex (+4%) and hippocampus (+4%) but was still highly significantly ( P < 0.0001) increased (+21%) in the medial thalamus. No significant alterations were found 3 days after seizure onset (Figs. 2L, O).

Fig. 4. Ultrastructure of control (A) and experimental rat neocortex (B, C) and hippocampus (D), 2 h after 4-AP administration. (A) Control parietal neocortex, in which a pyramidal neuron (N) soma, two capillaries (Cap) and a pericyte (P) can be seen. Scale bar: 5 Am. (B) Parietal neocortex, 2 h after 4-AP injection. Note largely swollen pericapillary astrocyte endfeet (asterisks). E: endothelial cell. Scale bar: 5 Am. (C) Slightly shrunken pyramidal neuron (N) surrounded by swollen astrocytes (asterisks) in the parietal neocortex. Scale bar: 5 Am. (D) Molecular layer of CA3 in the hippocampus, 2 h after 4-AP treatment. Note slightly swollen small dendrite (D), containing normal mitochondrion (M). Asterisks: swollen astrocyte processes; A: axon terminal; 1,2: dendritic spines. The extracellular space is slightly dilated. Scale bar: 1 Am.

P.F. Fabene et al. / Neurobiology of Disease 21 (2006) 80 – 89

Histology Light microscopic investigation (Fig. 3) of the hematoxylin – eosin-stained sections 2 h after the 4-AP convulsions revealed slight perivascular swelling in the parietal and temporal neocortex and scattered (few) dark nerve cell bodies, which appeared as deeply stained, shrunken neurons (not shown). The subcortical white matter displayed medium vacuolization. In the hippocampus, mainly the dentate gyrus was affected: scattered dark neurons in the hilus and in the granule cell layer, and vacuolization of the neuropil was observed. The fornix and fimbria were severely vacuolized, indicating edematous alterations. The amygdala and medial thalamus displayed mild alterations, including perivascular edema and occasional dark nerve cells (Table 1). No obvious pathology was apparent 24 h and 3 days after seizures. Ultrastructural alterations Two hours after seizure induction, prominent astrocytic swelling was apparent throughout the layers of the parietal cortex and in the hippocampus. Astrocyte cell bodies and processes were edematous, which was particularly prominent around capillaries (Fig. 4) and neuronal cell bodies. Scattered neuronal cell bodies were moderately shrunken with slightly increased electron density (Fig. 4). In these ‘‘dark’’ neurons, slightly dilated endoplasmic reticulum cisternae were seen. No other signs of cellular damage were detected. Small dendritic branches in the molecular layer of CA3 and layers I and II of the neocortex were slightly swollen, with decreased numbers of dendritic microtubules (Fig. 4). Slightly dilated extracellular space was also seen in the vicinity of small dendrites. The swelling of pericapillary astrocytic endfeet (i.e. the increase of the area occupied by astrocytic glia limitans) was significant compared to controls (Fig. 5). Consequently, moderate but significant decrease of capillary lumen diameter was observed (Fig. 5). No other ultrastructural alterations were apparent in the endothelial cells: the number of endothelial mitochondria was not obviously different compared to controls. The astrocytic swelling decreased significantly at 24 h and 3 days after seizures at this

87

survival periods, although some slightly swollen glial processes were still detected (Fig. 5). However, the total area of the glial endfeet was not obviously different from controls (Fig. 5). No pathological changes were observed in neurons and dendrites 24 h and 3 days.

Discussion In the present study, 4-AP-induced seizures were investigated by direct correlation between structural and functional MRI and histological and ultrastructural analysis. Functional alterations, detected by rCBV images, were correlated to structural alterations investigated both with MRI and with light and ultrastructural microscopy. Structural alterations were studied by T2W RARE images that were already reported to be sensitive in detecting epileptic alterations (Fabene et al., 2003). T2W maps indicate that the peak of T2 alterations occurs 2 h after seizures, whereas 24 h after seizures, these values are decreased close to baseline levels. The long-lasting modification of these values has been reported after SE in animals that subsequently become epileptic, such as kainic acid and pilocarpine seizures (Hasegawa et al., 2003; Fabene et al., 2003, respectively). In these chronic epilepsy models, the alterations of the T2W maps are more evident at 24 h after seizures, compared to our present findings in the 4-AP model. This fast recovery in T2W images would be consistent with the absence of an epileptogenic phase and reflected a reduced severity of the edema following brief, acute seizures. The ADC alterations reflect pathological conditions in brain tissue that are only partially understood and involve changes in the diffusion characteristics of intra- and extracellular water compartments, water exchange across permeable boundaries (Grass et al., 2001) and changes in volume transmission (Sykova´, 2004). In general, reduction of the ADC has been associated with acute cytotoxic edema (Fabene et al., 2003). Previous studies on sustained experimental seizures induced by pilocarpine (Wall et al., 2000), bicuculline (Zhong et al., 1993), systemic administration (Nakasu et al., 1995; Wang et al., 1996) or intrahippocampal injection of kainic acid (Tokumitsu et al., 1997) reported similar

Fig. 5. The results of electron microscopic morphometry. (A) Internal (luminal) diameter of capillaries in the parietal neocortex: 2 h after 4-AP administration, capillary diameters decreased significantly, while 24 h and 3 days after seizure onset, this parameter did not return to basal levels, although the differences are not significant. (B) The swelling of the pericapillary astrocyte endfeet in the parietal neocortex: the area of the glia limitans increases at 2 h, and the increase is strongly significant in convulsing animals. The astrocytic swelling decreases at 24 h and 3 days significantly but does not reach the control levels.

88

P.F. Fabene et al. / Neurobiology of Disease 21 (2006) 80 – 89

changes in ADC values in various brain areas, especially in limbic structures, and DWI was reported to provide sensitive indications on acute alterations in these paradigms (Nakasu et al., 1995; Wang et al., 1996; Wall et al., 2000). The present findings of decreased ADC values in the cerebral cortex can be explained with cytotoxic edema (Lassmann et al., 1984). The swelling of the astrocytes, and the edema of the perivascular glia limitans, suggests the presence of excess amounts of excitatory transmitters (glutamate—see Kova´cs et al., 2003), metabolites (CO2) and K+ in the extracellular space. On the basis of the morphological results, we conclude that the brief, acute seizures caused cellular edema of the astrocytes mainly (Kimelberg, 2004). The swelling of the astrocyte obliterated brain extracellular spaces, inhibiting the clearance of transmitters, ions and HCO3—from the extracellular space, contributing to and enhancing the cellular damage (Van Gelder, 1983). However, the astrocyte that has taken up glutamate from the extracellular space may release it again through connexon hemichannels (Simard and Nedergaard, 2004). Therefore, the astrocyte may sustain a long-lasting decrease in volume transmission (Sykova´, 2004). These changes may explain the long-lasting decrease of the ADC values in our experiments. We measured the prolonged decrease of the capillary lumen diameter, which may well contribute to these changes: the decrease of the capillary diameter impairs the local microcirculation (Farkas et al., 2003) and could contribute to the decrease of ADC values. The partial mismatch between the altered DWI values and the minor astrocytic swelling detected by histological analysis at 24 h may be partially explained by a dehydration process that affects the edema evaluation, allowing only the detection of the more obvious alterations. The increase of rCBV at the same time may reflect the compensatory effects of the local changes at the level of the arterioles and larger vessels (Farkas et al., 2003). In order to better understand CBV data, it should be considered that edema results in the decrease of the diameter of microvessels in the affected regions, thus inducing a relative ischemia (Lassmann et al., 1984), and a compensatory hyperperfusion in adjacent areas may be hypothesized. The decrease of the luminal diameter of neocortical capillaries was measured and demonstrated in our experiments. Taken together, our data show that brain damage following SE involves several presumably pathological processes operant in both limbic and extra-limbic regions, suggesting that sustained seizure activity elicits a complex rearrangement of cortical and subcortical neural networks. The ultrastructural changes indicate different processes controlling diffusion properties of the extracellular spaces. That is, tissue layers containing neuronal cell bodies display decreased extracellular space which is apparently neurotoxic. Conversely, neuropil regions rich in synapses, small dendrites and astrocytic processes have an apparently greater capacity for the clearance of the extracellular space, which may contribute to the maintenance of the seizure process. The present study demonstrates that different MRI strategies can sensitively detect seizure-induced changes in vivo at high resolution.

Acknowledgments The authors thanks Guerbet for kindly providing SineremR. This work was supported by grant T32566 from OTKA (Hungarian National Research Fund).

References Bouilleret, V., Nehlig, A., Marescaux, C., Namer, I.J., 2000. Magnetic resonance imaging follow-up of progressive hippocampal changes in a mouse model of mesial temporal lobe epilepsy. Epilepsia 41, 642 – 650. Bremer, C., Mustafa, M., Bogdanov Jr., A., Ntziachristos, V., Petrovsky, A., Weisseleder, R., 2003. Steady-state blood volume measurements in experimental tumors with different angiogenic burdens a study in mice. Radiology 226, 214 – 220. Dousset, V., Gomez, C., Petry, K.G., Delalande, C., Caille, J.M., 1999. Dose and scanning delay using USPIO for central nervous system macrophage imaging. MAGMA 8, 185 – 189. Evans, M., Griffiths, T., Meldrum, B., 1983. Early changes in the rat hippocampus following seizures induced by bicuculline or l-allylglycine: a light and electron microscope study. Neuropathol. Appl. Neurobiol. 9, 39 – 52. Fabene, P.F., Sbarbati, A., 2004. In vivo MRI in different models of experimental epilepsy. Current Drug Targets 57, 629 – 636. Fabene, P.F., Marzola, P., Sbarbati, A., Bentivoglio, M., 2003. Magnetic resonance imaging of changes elicited by status epilepticus in the rat brain: diffusion-weighted and T2-weighted images, regional blood volume maps, and direct correlation with tissue and cell damage. Neuroimage 18, 375 – 389. Farkas, I.G., Czigner, A., Farkas, E., Dobo, E., Soos, K., Penke, B., Endresz, V., Mihaly, A., 2003. Beta-amyloid peptide-induced blood – brain barrier disruption facilitates T-cell entry into the rat brain. Acta Histochem. 105, 115 – 125. Fragoso-Veloz, J., Massieu, L., Alvarado, R., Tapia, R., 1990. Seizures and wet-dog shakes induced by 4-aminopyridine, and their potentiation by nifedipine. Eur. J. Pharmacol. 178, 275 – 284. Grass, A., Niendorf, T., Hirsch, J.G., 2001. Acute and chronic changes of the apparent diffusion coefficient in neurological disorders—Biophysical mechanisms and possible underlying histopathology. J. Neurol. Sci. 186, S15 – S23. Hamberg, L.M., Boccalini, P., Stranjalis, G., Hunter, G.J., Huang, Z., Halpern, E., Weisskoff, R.M., Moskowitz, M.A., Rosen, B.R., 1996. Continuous assessment of relative cerebral blood volume in transient ischemia using steady state susceptibility-contrast MRI. Magn. Reson. Med. 35, 168 – 173. Hasegawa, D., Orima, H., Fujita, M., Nakamura, S., Takahashi, K., Ohkubo, S., Igarashi, H., Hashizume, K., 2003. Diffusion-weighted imaging in kainic acid-induced complex partial status epilepticus in dogs. Brain Res. 9831-2, 115 – 127. Kimelberg, H.K., 2004. Water homeostasis in the brain: basic concepts. Neuroscience 129, 851 – 860. ´ ., Gyenge´si, E., Szente, M., Kova´cs, A., Miha´ly, A., Koma´romi, A Weiczner, R., Krisztin-Pe´va, B., Szabo´, Gy., Telegdy, Gy., 2003. Seizure, neurotransmitter release, and gene expression are closely related in the striatum of 4-aminopyridine-treated rats. Epilepsy Res. 55, 117 – 129. Lassmann, H., Petsche, U., Kitz, K., Baran, H., Sperk, G., Seitelberger, F., Hornykiewicz, O., 1984. The role of brain edema in epileptic brain damage induced by systemic kainic acid injection. Neuroscience 13, 691 – 704. Mandeville, J.B., Marota, J.J., Kosofsky, B.E., Keltner, J.R., Weissleder, R., Rosen, B.R., Weisskoff, R.M., 1998. Dynamic functional imaging of relative cerebral blood volume during rat forepaw stimulation. Magn. Reson. Med. 39, 615 – 624. Miha´ly, A., Bencsik, K., Solymosi, T., 1990. Naltraxone potentiates 4aminopyridine seizures in the rat. J. Neural Transm.: Gen. Sect. 79, 59 – 67. Miha´ly, A., Shihab-Eldeen, A., Owunwanne, A., Gopinath, S., Ayesha, A., Mathew, M., 2000. Acute 4-aminopyridine seizures increase the regional cerebral blood flow in the thalamus and neocortex, but not in the entire allocortex of the mouse brain. Acta Physiol. Hung. 87, 43 – 52. Miha´ly, A., Szaka´cs, R., Bohata, Cs., Dobo´, E., Krisztin-Pe´va, B., 2001.

P.F. Fabene et al. / Neurobiology of Disease 21 (2006) 80 – 89 Time-dependent distribution and neuronal localization of c-fos protein in the rat hippocampus following 4-aminopyridine seizures. Epilepsy Res. 44, 97 – 108. Miha´ly, A., Borbe´ly, S., Vila´gi, I., De´ta´ri, L., Weiczner, R., Za´dor, Z., Krisztin-Pe´va, B., Bagosi, A., Kopniczky, Z., Za´dor, E., 2005. Neocortical c-fos mRNA transcription in repeated, brief, acute seizures: is c-fos a coincidence detector? Int. J. Mol. Med. 15, 481 – 486. Nakasu, Y., Nakasu, S., Morikawa, S., Uemura, S., Inubushi, T., Handa, J., 1995. Diffusion-weighted MR in experimental sustained seizures elicited with kainic acid. Am. J. Neuroradiol. 16, 1185 – 1192. Paxinos, G., Watson, C., 1986. The Rat Brain In Stereotaxic Coordinates, 2nd edR Academic Press, San Diego, CA. Pen˜a, F., Tapia, R., 2000. Seizures and neurodegeneration induced by 4aminopyridine in rat hippocampus in vivo: role of glutamate- and GABA-mediated neurotransmission and of ion channels. Neuroscience 101, 547 – 561. Perrault, P., Avoli, M., 1991. Physiology and pharmacology of epileptiform activity induced by 4-aminopyridine in rat hippocampal slices. J. Neurophysiol. 65, 771 – 779. Riederer, S.J., Jack, C.R., Grimm, R.C., Rydberg, J.N., Slavin, G.S., 1995. New technical developments in magnetic resonance imaging of epilepsy. Magn. Reson. Imaging 13, 1095 – 1098. Roch, C., Leroy, C., Nehlig, A., Namer, I.J., 2002. Magnetic resonance imaging in the study of the lithium – pilocarpine model of temporal lobe epilepsy in adult rats. Epilepsia 43, 325 – 335. Sbarbati, A., Reggiani, A., Lunati, E., Arban, R., Nicolato, E., Marzola, P., Asperio, R.M., Bernardi, P., Osculati, F., 2000. Regional cerebral blood volume mapping after ischemic lesions. NeuroImage 12, 418 – 424.

89

Shinnar, S., Hauser, W.A., 2002. Do occasional brief seizures cause detectable clinical consequences? Prog. Brain Res. 135, 221 – 235. Simard, M., Nedergaard, M., 2004. The neurobiology of glia in the context of water and ion homeostasis. Neuroscience 129, 877 – 896. Sykova´, E., 2004. Extrasynaptic volume transmission and diffusion parameters of the extracellular space. Neuroscience 129, 861 – 876. Szaka´cs, R., Weiczner, R., Miha´ly, A., Krisztin-Pe´va, B., Za´dor, Zs., Za´dor, E., 2003. Non-competitive NMDA receptor antagonists moderate seizure-induced c-fos expression in the rat cerebral cortex. Brain Res. Bull. 59, 485 – 493. Thesleff, S., 1980. Aminopyridines and synaptic transmission. Neuroscience 5, 1413 – 1419. Tokumitsu, T., Mancuso, A., Weinstein, P.R., Weiner, M.W., Naruse, S., Maudsley, A.A., 1997. Metabolic and pathological effects of temporal lobe epilepsy in rat brain detected by proton spectroscopy and imaging. Brain Res. 744, 57 – 67. Van Gelder, N.M., 1983. Metabolic interactions between neurons and astroglia: glutamine synthetase, carbonic anhydrase and water balance. In: Jasper, H.H., Van Gelder, N.M. (Eds.), Basic Mechanisms of Neuronal Hyperexcitability. Alan R. Liss, New York, pp. 5 – 29. Wall, C.J., Kendall, E.J., Obenaus, A., 2000. Rapid alterations in diffusionweighted images with anatomic correlates in a rodent model of status epilepticus. Am. J. Neuroradiol. 21, 1841 – 1852. Wang, Y., Majors, A., Naim, I., Xue, M., Comair, Y., Modic, M., Ng, T.C., 1996. Postictal alteration of sodium content and apparent diffusion coefficient in epileptic rat brain induced by kainic acid. Epilepsia 37, 1000 – 1006. Zhong, J., Petroff, O.A.C., Prichard, J.W., Gore, J.C., 1993. Changes in water diffusion and relaxation properties of rat cerebrum during status epilepticus. Magn. Reson. Med. 30, 241 – 246.