- Email: [email protected]
A single episode of neonatal seizures alters the cerebellum of immature rats
Epilepsy Research (2011) 93, 17—24
journal homepage: www.elsevier.com/locate/epilepsyres
A single episode of neonatal seizures alters the cerebellum of immature rats Selene Lomoio a, Daniela Necchi a, Vladislav Mares b,c, Elda Scherini a,∗ a
Department of Animal Biology, Laboratory of Cell Biology and Neurobiology, University of Pavia, via Ferrata 1, 27100 Pavia, Italy Institute of Physiology, Academy of Sciences of the Czech Republic v.v.i., Prague, Czech Republic c Faculty of Natural Sciences, University of J.E. Purkinje, Usti n.L., Czech Republic b
Received 20 September 2010; accepted 22 October 2010 Available online 20 November 2010
KEY WORDS Neonatal seizure; Cerebellum; AMPA receptor; Purkinje cell degeneration
Summary Purpose: To test whether a single episode of early-life seizures may interfere with the development of the cerebellum. The cerebellum is particularly vulnerable in infants, since it is characterized by an important postnatal histogenesis that leads to the settling of adult circuitry. Methods: Seizures were induced in 10-day-old Wistar rats with a single convulsive dose (80 g/g b.w., s.c.) of pentylentetrazole (PTZ). Immediately after rats were treated with 3 H-thymidine (3 HTdR, 2.5 Ci/g b.w, s.c.). Rats were killed 4 h later and paraffin sections of the cerebellar vermis were processed for 3 HTdR autoradiography and immunocytochemistry for 2/3 subunits of AMPA glutamate receptor (GluR2/3), glutamate transporter 1 (GLT1) and calbindin. Results: Seizures reduced the proliferation rate of cells in the external germinal layer. Purkinje cells showed increased GluR2/3 immunoreactivity. However, some Purkinje cells were unstained or lost. Increased GLT1 immunoreactivity was present in glial cells surrounding Purkinje cells. Calbindin immunoreaction confirmed that some Purkinje cells were missed. The remaining Purkinje cells showed large spheroids along the course of their axon. Conclusions: Data indicate that seizures lead to a loss and alteration of Purkinje cells in the cerebellum of immature rats. Since at 10 days of life Purkinje cells are no more proliferating, the loss of Purkinje cells should be permanent. © 2010 Elsevier B.V. All rights reserved.
Introduction Neonatal seizures, which affect approximately 3 in 1000 infants, are often associated with an adverse developmen-
∗ Corresponding author. Tel.: +39 0382 986320; fax: +39 0382 986325. E-mail address: [email protected] (E. Scherini).
tal outcome. However, whether seizures are themselves the primary cause of the poor developmental outcome is still a matter of controversy. In fact, other factors, which in turn may have triggered seizures, such as illnesses, brain insult or medication effect may concur to the brain damage. An animal model study, which eliminates other causes of brain damage, demonstrated that multiple episodes of early-life seizures result in late learning impairment correlated with cell loss and synaptic reorganization in
0920-1211/$ — see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.eplepsyres.2010.10.013
18 the hippocampus (Swann, 2004). On the contrary, other studies indicated that immature rats experiencing a single episode of early-life seizures did not suffer late behavioral alteration nor cell loss and synaptic reorganization (Holmes et al., 1988; Haas et al., 2001). Nevertheless, a report describes hippocampal-dependent memory and synaptic plasticity impairment in rats after a single episode of early-life seizures (Lynch et al., 2000) and, recently, neuronal degeneration has been described in the hippocampus following lithium-pilocarpine status epilepticus in 12 days old rats (Druga et al., 2010). So far, the majority of studies have been focused on hippocampus, the key brain region for memory, and little attention has been directed to other brain areas and, in particular, to the cerebellum. The cerebellum may be especially vulnerable in infants, since it is characterized by an important postnatal histogenesis leading to the settling of adult circuitry. In addition, a recent study indicates the cerebellum as a region of seizure focus (Carmody and Brennan, 2010) and cerebellar dysplastic lesions can be epileptogenic (Vander et al., 2004). Several studies indicate a role for ␣-amino-3-hydroxy-5-methyl-D-aspartate glutamate receptors (AMPAR) in long lasting sequelae after early-life seizures (Sanchez et al., 2001; Cornejo et al., 2007; Friedman et al., 2007). Moreover, AMPAR antagonists, but not NMDA receptor antagonists or GABA agonists, administered within the first 48 h after seizures, attenuated seizure-induced neuronal injury in the hippocampus (Koh et al., 2004). AMPARs mediate fast synaptic transmission and in the cerebellum, they are highly expressed at both climbing and parallel fiber synapses on Purkinje cells (Zhao et al., 1998). AMPARs are a tetrameric assembly of GluR1—4 subunits. In the adult cerebellum, the Ca2+ -permeable GluR1 and GluR4 subunits are exclusively expressed in glial cells, while the Ca2+ -impermeable GluR2 and GluR3 subunits are present on Purkinje cell membranes (Douyard et al., 2007). During the cerebellar development, these subunits undergo profound changes in their distribution and recently it has been speculated that AMPAR may be involved in stabilizing outgrowing branches of Purkinje cell dendrites (Douyard et al., 2007). The purpose of the present work was to examine whether a single episode of early-life seizures interferes with the developmental program of the cerebellum and GluR2/3 distribution and expression. These questions were explored by treating 10-day-old rats with a single convulsive dose of pentylentetrazole (PTZ). PTZ is a convulsive drug widely used to induce seizures experimentally. Though the mechanism of action of PTZ is not fully understood, it is generally accepted that part of its action is due to its antagonist binding to the picrotoxin-binding site of the postsynaptic GABAA receptor (Macdonald and Barker, 1977). In the rat, the 10th day of life is a crucial age for the development of the cerebellum. At 10 days, the generation of granule cells reaches a peak (Altman, 1972) and synaptogenesis of parallel fibers starts on Purkinje cell dendrites (Zhao et al., 1998). Here we report damage to Purkinje cells, accompanied by inhibition of DNA synthesis and up-regulation of GluR2/3 subunits in the cerebellum of early postnatal rats after a single episode of PTZ-induced seizures.
S. Lomoio et al.
Material and methods Animals and treatment Male 10-day-old Wistar rats (4 animals, date of birth 0) were treated with a single dose of 80 g/g b.w. of pentylentetrazole (PTZ, Sigma, MO, USA) dissolved in saline, applied subcutaneously. Control animals (4 rats from the same litter) received saline only. Immediately after, the animals were injected subcutaneously with 3 H-thymidine (3 HTdR, 2.5 Ci/g b.w., specific activity 600 GBq/mM, UVVVR, Prague, CZ), followed by the same dose 1 h later. All animals were killed by decapitation 4 h after the last 3 HTdR injection, brains immediately excised, fixed in Carnoy solution and embedded in Paraplast X-tra (Polysciences Inc., Warrington, PA, USA). Ten-m thick midsagittal sections were used for autoradiography and immunocytochemistry. Experiments were approved by the Animal Care and Use Committee of the Institute of Physiology of the Academy of Sciences of the Czech Republic. Animal care and experimental procedures were conducted in accordance with the guidelines of the European Community Council directives 86/609/EEC.
Autoradiography Sections were covered with Kodak AR10 stripping film and exposed for 57 days at 4 ◦ C in the dark. Autoradiographs were counterstained with Mayer hematoxylin.
Immunocytochemistry The following antibodies were used for immunocytochemistry: - 1:30 rabbit polyclonal anti-␣-amino-3-hydroxy-5-methyl-Daspartate (AMPA) glutamate receptor 2/3 antiserum (GluR2/3, Chemicon, Temecula, CA, USA). The antibody recognizes both GluR2 and GluR3 AMPA receptor subunits; - 1:5000 guinea pig polyclonal anti-glial glutamate transporter 1 antiserum (GLT1, Chemicon, Temecula, CA, USA). The antibody recognizes the carboxy terminus of GLT1; - 1:5000 rabbit polyclonal anti-calbindin antiserum (CB38, Swant, Bellinzona, CH). In the cerebellum, this antibody stains uniquely Purkinje cells, even in their thinnest processes, giving Golgi-like pictures. Sections were processed as follows. After treatment with 3% H2 O2 in 10% methanol in PBS and with 10% normal goat serum in PBS, the sections were incubated overnight with the primary antibodies. Afterwards the sections were incubated for 30 min in goat anti-rabbit or goat anti-guinea pig biotinylated IgG, followed by streptavidin-HRP complex (Vectastain Elite kit, Vector, CA, USA). The complex was revealed by 3 ,5 -diaminobenzidine tetrahydrochloride. For the reaction control, the primary antibodies were omitted. Sections were photographed with Olympus Camedia C-5050 mounted on Olympus BX51 microscope.
Cell counting To evaluate the extent of cell proliferation in the external germinal layer, 10 sections per animal (4 PTZ treated rats and 4 controls), submitted to autoradiography were scored. The sections were photographed and the number of 3 HTdR labeled cells/100 m of cerebellar cortex convolution was counted in the external germinal layer. Since it is known that the rate of cell proliferation varies between surface of folia and depth of fissures and among lobes (Mares and Lodin, 1970), the cell counting was performed in
A single episode of neonatal seizures alters the cerebellum of immature rats
19
Figure 1 Autoradiography for DNA synthesis in the cerebellum of control and PTZ treated rats. (a) Control rats. In the external germinal layer many cells appear to be labeled. (b) PTZ treated rats. Only a few cells are labeled, mainly in the upper part of the layer. EGL = external germinal layer; ML = molecular layer; Pc = Purkinje cells; IGL = granule cell layer. Scale bars: 40 m.
the middle part of the fissura prima. A convolution length of about 2 mm vas scored in every section. To evaluate the extent of Purkinje cell degeneration in PTZ treated and control rats, twelve sections per animal, immunostained for AMPA GluR2/3 (six sections per animal) and for calbindin (six sections per animal), were selected. The sections were photographed at low magnification and the length of the Purkinje cell layer was measured using the NIH Image program for Macintosh. The total number of Purkinje cells per section was counted at the microscope by using differential interference contrast (DIC) optic at 40× magnification, and divided by the length of the Purkinje cell layer, giving a linear cell density. In addition, in AMPA GluR2/3 immunostained sections, the number of unstained cells was determined and their percentage in relation to the total number of cells was calculated. Statistical evaluation of differences was performed by means of the two-level nested ANOVA on ranks test.
Results Animals treated with PTZ developed generalized tonic—clonic jerk, scratching and body shake seizure activity within few min after injection. The ictal bursts lasted less then 10 min, separated by 5—10 min intervals. The seizure activity lasted about 45 min. All animals survived to seizures. Autoradiographs revealed that cell proliferation in the external germinal layer (EGL) was slightly, but significantly (p < 0.05) decreased in animals treated with PTZ. The number of 3 HTdR labeled cells in the external germinal layer (EGL) was reduced by 12% in comparison with control rats (12.24 ± 2.23 versus 13.92 ± 2.57). No difference was found within groups. Though as previously described (Mares and Lodin, 1970) 3 HTdR labeled cells were not uniformly distributed, but occurred in nests, in PTZ treated animals labeled cells appeared less densely packed (Fig. 1). In control rats, immunocytochemistry for AMPA GluR2/3 faintly decorated the somata and growing dendritic trees of Purkinje cells (Fig. 2(a)), though with a certain degree of variability. Some Purkinje cells had nearly negative soma and the immunoreaction stained the principal branches of the dendrite only (Fig. 2(b)). In PTZ treated rats, in most Purkinje cells the staining was more intense, decorating also the thinnest branches of Purkinje cell dendrites (Fig. 2(c)). However, in some tracts of the cerebellar convolutions,
Purkinje cells were not labeled at all or absent (Fig. 2(d and e)). In other tracts of convolutions, in which the somata of Purkinje cells were not visible even by using phase contrast or DIC optics, in the molecular layer (ML) there were well stained dendrite remnants (Fig. 2(f)). Cell counting revealed that the linear cell density of Purkinje cells was reduced in all treated animals in comparison with control rats, as well evidenced by the graphic representation of single measurements (Fig. 3). Statistic evaluation of data revealed that differences between treated and control animals are highly significant (p < 0.0001) and that the 89.81% of the total variability is ascribable to the treatment. The percentage of unlabeled cells was not coherent in all sections and/or animals, ranging from 15.59% to 29.35% (mean percentage 21.84 ± 4.83). Unlabeled or weakly labeled cells often appeared shrunk, with irregular contour (Fig. 4(a)). Routine hematoxylin staining revealed few apoptotic cells in the EGL of both PTZ treated and control rats (Fig. 4(b)). In addition, only in treated rats, in the Purkinje cell layer some large cell remnants were present among Purkinje cells (Fig. 4(c and d)). After immunocytochemistry for GLT1, in both control and PTZ treated animals, staining was present in Golgi epithelial cell processes enwrapping Purkinje cells and in their Bergman fibers in the ML (Fig. 5). In addition, diffuse staining was visible in the ML in between the branches of Purkinje cell dendritic trees. The staining was more intense in treated animals (Fig. 5(b)). Calbindin immunocytochemistry was performed for additional morphological analysis of Purkinje cells. In the cerebellum, calbindin immunocytochemistry stains exclusively Purkinje cells even in their thinnest dendrites and axon, giving a Golgi-like picture (Fig. 6(a)). The immunostaining confirmed that in PTZ treated rats, in some tracts of convolution Purkinje cells were lost and/or only remnants of their dendritic trees were present (Fig. 6(b)). Moreover, in some small areas the ML displayed a diffuse staining, as to dendritic trees of Purkinje cells underwent lysis (Fig. 6(c)). The observation of the remaining Purkinje cells revealed that in some cells (about 3% of the cell population) the axons bore large spheroids along their course in the underlying granule cell layer (IGL, Fig. 6(d and e)). Cell counting confirmed the data obtained in AMPA GluR2/3 immunostained sections (Fig. 7).
20
S. Lomoio et al.
Figure 2 Immunocytochemistry for Glu2/3 AMPA receptors. (a and b) Control rats. The reaction decorates Purkinje cells somata and growing dendrites. In some tracts of convolutions, some Purkinje cells have negative or faintly stained soma (arrows in b). (c—f) PTZ treated rats. Purkinje cells appear more densely stained in comparison with controls (c). However, loss of Purkinje cells occurred in some tracts of convolution (d and e, arrows). In other tracts, only the dendritic trees are preserved (f). ML = molecular layer; Pc = Purkinje cells; IGL = granule cell layer. Scale bars: 40 m.
A single episode of neonatal seizures alters the cerebellum of immature rats
Figure 3 Graphic representation of single measurements of Purkinje cell linear density (number of cell/mm of convolution, ordinate) obtained in 6 sections per animal, immunostained for AMPA GluR2/3. The maximum value obtained in PTZ-treated animals (PTZ1—4) does not overlap with the minimum value for control animals (Ctr1—4).
Discussion Neonatal seizures can be the cause of learning difficulties and other intellectual or motor disabilities, as well as cerebral atrophy and permanent reduction in the brain cell number (Sarkisian et al., 1997; Mizrahi, 1999; Swann, 2004; Nairismagi et al., 2006). Our data indicate that DNA synthesis and cell proliferation are reduced in the rat cerebellum as soon as 4 h after a single episode of seizures, induced by PTZ administration at 10 days of life. Inhibition of DNA synthesis and cell mitotic activity have been already reported in the brain, including the cerebellum, of rats in which a status epilepticus was induced by bicuculline treatment at 4 days of life and killed 1 or 7 days later (Wasterlain, 1976; Suga and Wasterlain, 1980). However, after 26 days the dam-
21
age was partially recovered. Actually, a restoration of the EGL has been observed after DNA synthesis inhibition by cisdichlorodiammineplatinum treatment (Mares et al., 1986). However, we also found a conspicuous degeneration and loss of Purkinje cells, as well evidenced by both AMPA GluR2/3 and calbindin immunostainings. On the other hand, Purkinje cell loss and injury have been described in the cerebellum of epileptic patients (Leifer et al., 1991). In this context, the observation in calbindin immunostained sections of axonal spheroids in Purkinje cells deserves particular attention. Axonal spheroids have been described in Purkinje cells in numerous murine models of pathologies, such as cerebellar ataxias (Sotelo, 1990; Jeong et al., 2000), acrylamide poisoning (Lehning et al., 2002), Down syndrome (Necchi et al., 2008; Lomoio et al., 2009) and after axotomy (Dusart et al., 1999). The presence of this type of axonal abnormality has been interpreted as a unique Purkinje cell reaction to axon injury, either by chemical insult or physical damage (axotomy) or degeneration and loss of target cells in the cerebellar nuclei. In our model of neonatal seizures, a direct damage to the Purkinje cell axon is difficult to explain. Indeed, observation of cerebellar nuclei in sections taken at the level of the intermediate cortex failed to reveal any sign of damaged cells or cell loss (not shown). Though we cannot exclude a direct toxic action of PTZ, the damage to the axon must be secondary to a seizure insult to the somata and dendritic trees of Purkinje cells. As a consequence, several Purkinje cells underwent degeneration and were quickly removed by the surrounding glial cells. This interpretation, that does not contradict previous data indicating that Purkinje cells are extremely resistant to axonal damage and can survive to axotomy up to 18 months (Dusart et al., 1999; Rossi et al., 2006), indicates that the development of axonal spheroids is not an exclusive feature of axonal injury, but rather reflects a more generalized damage or altered physiology of the cell.
Figure 4 Cell degeneration in the cerebellum of PTZ treated rats. (a) Immunocytochemistry for Glu2/3 AMPA receptors. An example of an unstained Purkinje cell (arrow) near a well stained cell (arrowhead). The contour of the unstained cell appears irregular. Differential interference contrast optic. (b) Mayer hematoxylin staining. In the external germinal layer, some apoptotic cells are present (arrows). (c and d) Mayer hematoxylin staining. Large cell remnants (arrows) can be observed in the Purkinje cell layer. Scale bars: 10 m
22
S. Lomoio et al.
Figure 5 Immunocytochemistry for GLT1 transporter. (a) Control rats. Staining is present in Golgi epithelial cell processes enwrapping Purkinje cell somata and in their Bergman fibers (arrows) in the molecular layer. (b) PTZ treated rats. The staining appears more intense. EGL = external germinal layer; ML = molecular layer; Pc = Purkinje cells. Scale bars: 40 m.
The immunocytochemical data for Glu2/3 AMPA receptors fit in well with this interpretation. AMPA glutamate receptors are a tetrameric assembly of subunits GluR1—4, which are developmentally regulated (Douyard et al., 2007). In the adult rat cerebellum, the postsynaptic densities of climbing and parallel fibers, which represent the main excitatory glutamatergic pathways, on Purkinje cell dendrites contain only the Ca2+ -impermeable receptor complex GluR2/3 (Zhao et al., 1998). This is already present at 10 days of life, when the expression of GluR2/3 subunits is abolished in other cerebellar neurons and becomes an exclusive feature of Purkinje cells (Douyard et al., 2007). After
PTZ administration, the immunocytochemical expression of GluR2/3 was increased in Purkinje cells. So far, inconsistent data are present in literature concerning the changes in GluR2/3 expression following seizures, some papers reporting an increase (Friedman et al., 2007), a decrease (Jiang et al., 2008), or no effect (Cornejo et al., 2007). These differences may be due to the type and timing of the epileptic insult, the brain area considered, or the different models utilized in the experiments (rats or cell cultures). As to the cerebellum, as far as we know, no studies have been addressed to the changes in Glu2/3 AMPA receptors after seizures. Once again, we cannot exclude that the increase
Figure 6 Immunocytochemistry for calbindin. (a) Control rats. Purkinje cell somata, dendrites and axons are intensely stained. (b—e) PTZ treated rats. In some tracts of convolution, Purkinje cells are missed, though some remnants of branches of their dendritic trees are still present (arrow in b). In other tracts of convolutions Purkinje cells are unstained or only faintly stained and diffuse reaction product is present in the molecular layer occupied by Purkinje cell dendritic trees (c). Some axons of Purkinje cells bear large spheroids (arrows in d and e). Scale bars: a—d: 40 m; e: 10 m
A single episode of neonatal seizures alters the cerebellum of immature rats
Figure 7 Graphic representation of single measurements of Purkinje cell linear density (number of cell/mm of convolution, ordinate) obtained in 6 sections per animal, immunostained for calbindin. The maximum value obtained in PTZ-treated animals (PTZ1—4) does not overlap with the minimum value for control animals (Ctr1—4).
in the immunocytochemical detection of AMPA GluR2/3 is due to a direct toxic effect of PTZ. However, studying the metabolic glutamate turnover in the cerebellum of PTZ treated young-adult rats. Eloqayli et al. (2003, 2004) found that after a subconvulsive dose (20 mg/kg) the cerebellum was only slightly affected, whereas after the convulsive dose of 70 mg/kg the cerebellum was the most susceptible area of the CNS. According to the authors the data indicate that the changes observed in the cerebellum are caused by seizures and not PTZ. Moreover, the cerebellum has been shown to be a target for PTZ-induced seizure. In fact, subconvulsive administration of PTZ is able to induce long lasting paroxysmal EEG activity not only in the cortex and midbrain, but also in the cerebellum (Sierra-Paredes et al., 1989). On the other hand, changes in GluR2/3 immunocytochemical detection or expression have been already described in the cerebellum in other pathologies. In fact, a decrease in the total expression of GluR2/3, accompanied by increased immunoreactivity in the surviving Purkinje cells, has been described in the cerebellum of Creutzfeldt—Jakob disease patients (Ferrer and Puig, 2003). Similarly, in rats in which hepatic encephalopathy was induced by portacaval shunt, the number of GluR2/3-positive Purkinje cells was decreased after 1 month, but increased after 6 months (Suarez et al., 1997). Again, in an experimental model of hypobaric hypoxia, both the levels of GluR2 and GluR3 mRNAs and the immunocytochemical detection of GluR2/3 were enhanced in Purkinje cells (Kaur et al., 2005). Finally, a single fear-inducing stimulus induced the appearance of GluR2-containing AMPA receptors in parallel fiber/stellate interneuron synapses in mice, a localization where this AMPA receptor subunit is normally unexpressed (Liu et al., 2010). The high density of GluR2/3 labeling on Purkinje cells observed by us and reported in the above papers may represent a functional reorganization or intracellular trafficking of AMPA GluR2/3 aimed at reducing the Ca2+ influx caused by an excess in the extracellular glutamate levels. In fact, it is known that in several pathological states, such as ischemia, hypoxia, hypoglycemia, epilepsy and some neurodegenerative diseases, extracellular glutamate levels are increased (Druga et al., 2005; Caudle and Zhang, 2009; Wang and Qin,
23
2010). This increase would result in excessive accumulation of Ca2+ in a cell and excitotoxicity (Olney, 1978; Gonsette, 2008). In this view, we should hypothesize that as soon as a neuron succeeds in expressing more Ca2+ -impermeable AMPA glutamate receptors, the probability to survive to the insult would be enhanced. Consequently, the GluR2/3 immunonegative cells, which contours often appear irregular, would be destined to death. The increase in GLT1 immunoreactivity of Golgi epithelial cells is probably aimed at a faster removing of extracellular glutamate. In conclusion, we have here reported for the first time that a single episode of seizures is able to damage the cerebellum of 10-day-old rats, leading to loss of Purkinje cells as soon as 4 h after the insult. Since at 10 days of life Purkinje cells are no more proliferating, the loss of Purkinje cells should be permanent. Further studies are needed to verify whether further cell loss and circuit remodeling occur at later stages and whether the morphological and AMPA receptor changes are recovered.
Acknowledgements We wish to thank Prof. L.A. Zonta, Department of Genetics and Microbiology, University of Pavia, and Prof. G. Bogliani, Department of Animal Biology, University of Pavia for the precious advices on statistical analysis. This work was supported in part by Project AV0Z50110509 to V.M.
References Altman, J., 1972. Postnatal development of the cerebellar cortex in the rat. 3. Maturation of the components of the granular layer. J. Comp. Neurol. 145, 465—513. Carmody, S., Brennan, L., 2010. Effects of pentylentetrazoleinduced seizures on metabolomic profiles of rat brain. Neurochem. Int. 56, 340—344. Caudle, W., Zhang, J., 2009. Glutamate, excitotoxicity, and programmed cell death in Parkinson disease. Exp. Neurol. 220, 230—233. Cornejo, B., Mesches, M., Coultrap, S., Browning, M., Benke, T., 2007. A single episode of neonatal seizures permanently alters glutamatergic synapses. Ann. Neurol. 61, 411—426. Douyard, J., Shen, L., Huganir, R., Rubio, M., 2007. Differential neuronal and glial expression of GluR1 AMPA receptor subunit and the scaffolding proteins SAP97 and 4.1N during rat cerebellar development. J. Comp. Neurol. 502, 141—156. Druga, R., Mares, P., Otahal, J., Kubova, H., 2005. Degenerative neuronal changes in the rat thalamus induced by status epilepticus at different developmental stages. Epilepsy Res. 63, 43—65. Druga, R., Mares, P., Kubova, H., 2010. Time course of neuronal damage in the hippocampus following lithium-pilocarpine status epilepticus in 12-day-old rats. Brain Res. 1355, 174—179. Dusart, I., Morel, M., Wehrlè, R., Sotelo, C., 1999. Late axonal sprouting of injured Purkinje cells and its temporal correlation with permissive changes in the glial scar. J. Comp. Neurol. 408, 399—418. Eloqayli, H., Dahl, C.D., Gotestam, K.G., Unsgard, G., Hadidi, H., Sonnewald, U., 2003. Pentylentetrazole decreases metabolic glutamate turnover in rat brain. J. Neurochem. 85, 1200—1207. Eloqayli, H., Dahl, C.D., Gotestam, K.G., Unsgard, G., Sonnewald, U., 2004. Changes of glial—neuronal interaction and metabolism after a subconvulsive dose of pentylentetrazole. Neurochem. Int. 45, 739—745.
24 Ferrer, I., Puig, B., 2003. GluR2/3, NMDA1 and GABAA receptors in Creutzfeldt—Jakob disease. Acta Neuropathol. 106, 311—318. Friedman, L., Avallone, J., Magrys, B., 2007. Maturational effects of single and multiple early-life seizures on AMPA receptors in prepubescent hippocampus. Dev. Neurosci. 29, 427—437. Gonsette, R., 2008. Neurodegeneration in multiple sclerosis: the role of oxidative stress and excitotoxicity. J. Neurol. Sci. 274, 48—53. Haas, K., Sperber, E., Opanashuk, L., Stanton, P., Moshè, S., 2001. Resistance of immature hippocampus to morphologic and physiologic alterations following status epilepticus or kindling. Hippocampus 11, 615—625. Holmes, G., Thompson, J., Marchi, T., Feldman, D., 1988. Behavioral effects of kainic acid administration on the immature brain. Epilepsia 29, 721—730. Jeong, Y., Hyun, B., Hawkes, R., 2000. Abnormalities in cerebellar Purkinje cells in the novel ataxic mutant mouse, pogo. Brain Res. Dev. Brain Res. 125, 61—67. Jiang, Q., Wang, J., Wu, Y., Wu, X., Qin, J., Jiang, Y., 2008. Early-life epileptiform discharges exert both rapid and long-lasting effects on AMPAR subunit composition and distribution in developing neurons. Neurosci. Lett. 444, 31—35. Kaur, C., Sivakumar, V., Singh, G., Singh, J., Ling, E., 2005. Response of Purkinje neurons to hypobaric hypoxic exposure as shown by alteration in expression of glutamate receptors, nitric oxide synthases and calcium binding proteins. Neuroscience 135, 1217—1229. Koh, S., Tibayan, F., Simpson, J., Jensen, F., 2004. NBQX or topiramate treatment after perinatal hypoxia-induced seizures prevents later increases in seizure-induced neuronal injury. Epilepsia 45, 569—575. Lehning, E., Balaban, C., Ross, J., Reid, M., LoPachin, R., 2002. Acrylamide neuropathy. I. Spatiotemporal characteristics of nerve cell damage in rat cerebellum. Neurotoxicology 23, 397—414. Leifer, D., Cole, D., Kowall, N., 1991. Neuropathologic asymmetries in the brain of a patient with a unilateral status epilepticus. J. Neurol. Sci. 103, 127—135. Liu, Y., Formisano, L., Savtchouk, I., Takayasu, Y., Szabo, G., Zukin, R., Liu, S., 2010. A single fear-inducing stimulus induces a transcription-dependent switch in synaptic AMPAR phenotype. Nat. Neurosci. 13, 223—231. Lomoio, S., Scherini, E., Necchi, D., 2009. Beta-amyloid overload does not directly correlate with SAPK/JNK activation and tau protein phosphorylation in the cerebellar cortex of Ts65Dn mice. Brain Res. 1297, 198—206. Lynch, M., Sayin, U., Bownds, J., Janumpalli, S., Sutula, T., 2000. Long-term consequences of early postnatal seizures on hippocampal learning and plasticity. Eur. J. Neurosci. 12, 2252—2264. Macdonald, R., Barker, J., 1977. Pentylenetetrazol and penicillin are selective antagonists of GABA-mediated post-synaptic inhibition in cultured mammalian neurones. Nature 267, 720—721. Mares, V., Lodin, Z., 1970. The cellular kinetics of the developing mouse cerebellum. II. The function of the external granular layer in the process of gyrification. Brain Res. 23, 343—352.
S. Lomoio et al. Mares, V., Scherini, E., Biggiogera, M., Bernocchi, G., 1986. Influence of cis-dichlorodiammineplatinum on the structure of the immature rat cerebellum. Exp. Neurol. 91, 246—258. Mizrahi, E., 1999. Acute and chronic effects of seizures in the developing brain: lessons from clinical experience. Epilepsia 40 (Suppl. 1), S42—50. Nairismagi, J., Pitkanen, A., Kettunen, M., Kauppinen, R., Kubova, H., 2006. Status epilepticus in 12-day-old rats leads to temporal lobe neurodegeneration and volume reduction: a histologic and MRI study. Epilepsia 47, 479—488. Necchi, D., Lomoio, S., Scherini, E., 2008. Axonal abnormalities in cerebellar Purkinje cells of the Ts65Dn mouse. Brain Res. 1238, 181—188. Olney, J., 1978. Neurotoxicity of amino acids. In: McGeer, E.G., Olney, J.W., McGeer, P.L. (Eds.), Kainic Acid as a Tool in Neurobiology. Raven Press, New York, pp. 95—121. Rossi, F., Gianola, S., Corvetti, L., 2006. The strange case of Purkinje axon regeneration and plasticity. Cerebellum 5, 174—182. Sanchez, R., Koh, S., Rio, C., Wang, C., Lamperti, E., Sharma, D., Corfas, G., Jensen, F., 2001. Decreased glutamate receptor 2 expression and enhanced epileptogenesis in immature rat hippocampus after perinatal hypoxia-induced seizures. J. Neurosci. 21, 8154—8163. Sarkisian, M., Tandon, P., Liu, Z., Yang, Y., Hori, A., Holmes, G., Stafstrom, C., 1997. Multiple kainic acid seizures in the immature and adult brain: ictal manifestations and long-term effects on learning and memory. Epilepsia 38, 1157—1166. Sierra-Paredes, G., Soto-Otero, R., Mendez-Alvarez, E., GalanValiente, J., Aguilar-Veiga, E., Sierra-Marcuno, G., 1989. Experimental spike-and-wave discharges induced by pentylentetrazole and tolerance to repeated injections: an electrophysiological and biochemical study. Epilepsy Res. 4, 139—146. Sotelo, C., 1990. Axonal abnormalities in cerebellar Purkinje cells of the ‘hyperspiny Purkinje cell’ mutant mouse. J. Neurocytol. 19, 737—755. Suarez, I., Bodega, G., Fernandez, B., 1997. Modulation of AMPA receptor subunits GluR1 and GluR2/3 in the rat cerebellum in an experimental hepatic encephalopathy model. Brain Res. 778, 346—353. Suga, S., Wasterlain, C., 1980. Effects of neonatal seizures or anoxia on cerebellar mitotic activity in the rat. Exp. Neurol. 67, 573—580. Swann, J., 2004. The effects of seizures on the connectivity and circuitry of the developing brain. Ment. Retard. Dev. Disabil. Res. Rev. 10, 96—100. Vander, T., Medvedovsky, M., Herishanu, Y., 2004. Epilepsia partialis continua possibly caused by cerebellar lesion. Cerebellum 3, 126—128. Wang, Y., Qin, Z., 2010. Molecular and cellular mechanisms of excitotoxic neuronal death. Apoptosis, doi:10.1007/s10495-0100481-0. Wasterlain, C., 1976. Effects of neonatal status epilepticus on rat brain development. Neurology 26, 975—986. Zhao, H., Wenthold, R., Petralia, R., 1998. Glutamate receptor targeting to synaptic populations on Purkinje cells is developmentally regulated. J. Neurosci. 18, 5517—5528.
journal homepage: www.elsevier.com/locate/epilepsyres
A single episode of neonatal seizures alters the cerebellum of immature rats Selene Lomoio a, Daniela Necchi a, Vladislav Mares b,c, Elda Scherini a,∗ a
Department of Animal Biology, Laboratory of Cell Biology and Neurobiology, University of Pavia, via Ferrata 1, 27100 Pavia, Italy Institute of Physiology, Academy of Sciences of the Czech Republic v.v.i., Prague, Czech Republic c Faculty of Natural Sciences, University of J.E. Purkinje, Usti n.L., Czech Republic b
Received 20 September 2010; accepted 22 October 2010 Available online 20 November 2010
KEY WORDS Neonatal seizure; Cerebellum; AMPA receptor; Purkinje cell degeneration
Summary Purpose: To test whether a single episode of early-life seizures may interfere with the development of the cerebellum. The cerebellum is particularly vulnerable in infants, since it is characterized by an important postnatal histogenesis that leads to the settling of adult circuitry. Methods: Seizures were induced in 10-day-old Wistar rats with a single convulsive dose (80 g/g b.w., s.c.) of pentylentetrazole (PTZ). Immediately after rats were treated with 3 H-thymidine (3 HTdR, 2.5 Ci/g b.w, s.c.). Rats were killed 4 h later and paraffin sections of the cerebellar vermis were processed for 3 HTdR autoradiography and immunocytochemistry for 2/3 subunits of AMPA glutamate receptor (GluR2/3), glutamate transporter 1 (GLT1) and calbindin. Results: Seizures reduced the proliferation rate of cells in the external germinal layer. Purkinje cells showed increased GluR2/3 immunoreactivity. However, some Purkinje cells were unstained or lost. Increased GLT1 immunoreactivity was present in glial cells surrounding Purkinje cells. Calbindin immunoreaction confirmed that some Purkinje cells were missed. The remaining Purkinje cells showed large spheroids along the course of their axon. Conclusions: Data indicate that seizures lead to a loss and alteration of Purkinje cells in the cerebellum of immature rats. Since at 10 days of life Purkinje cells are no more proliferating, the loss of Purkinje cells should be permanent. © 2010 Elsevier B.V. All rights reserved.
Introduction Neonatal seizures, which affect approximately 3 in 1000 infants, are often associated with an adverse developmen-
∗ Corresponding author. Tel.: +39 0382 986320; fax: +39 0382 986325. E-mail address: [email protected] (E. Scherini).
tal outcome. However, whether seizures are themselves the primary cause of the poor developmental outcome is still a matter of controversy. In fact, other factors, which in turn may have triggered seizures, such as illnesses, brain insult or medication effect may concur to the brain damage. An animal model study, which eliminates other causes of brain damage, demonstrated that multiple episodes of early-life seizures result in late learning impairment correlated with cell loss and synaptic reorganization in
0920-1211/$ — see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.eplepsyres.2010.10.013
18 the hippocampus (Swann, 2004). On the contrary, other studies indicated that immature rats experiencing a single episode of early-life seizures did not suffer late behavioral alteration nor cell loss and synaptic reorganization (Holmes et al., 1988; Haas et al., 2001). Nevertheless, a report describes hippocampal-dependent memory and synaptic plasticity impairment in rats after a single episode of early-life seizures (Lynch et al., 2000) and, recently, neuronal degeneration has been described in the hippocampus following lithium-pilocarpine status epilepticus in 12 days old rats (Druga et al., 2010). So far, the majority of studies have been focused on hippocampus, the key brain region for memory, and little attention has been directed to other brain areas and, in particular, to the cerebellum. The cerebellum may be especially vulnerable in infants, since it is characterized by an important postnatal histogenesis leading to the settling of adult circuitry. In addition, a recent study indicates the cerebellum as a region of seizure focus (Carmody and Brennan, 2010) and cerebellar dysplastic lesions can be epileptogenic (Vander et al., 2004). Several studies indicate a role for ␣-amino-3-hydroxy-5-methyl-D-aspartate glutamate receptors (AMPAR) in long lasting sequelae after early-life seizures (Sanchez et al., 2001; Cornejo et al., 2007; Friedman et al., 2007). Moreover, AMPAR antagonists, but not NMDA receptor antagonists or GABA agonists, administered within the first 48 h after seizures, attenuated seizure-induced neuronal injury in the hippocampus (Koh et al., 2004). AMPARs mediate fast synaptic transmission and in the cerebellum, they are highly expressed at both climbing and parallel fiber synapses on Purkinje cells (Zhao et al., 1998). AMPARs are a tetrameric assembly of GluR1—4 subunits. In the adult cerebellum, the Ca2+ -permeable GluR1 and GluR4 subunits are exclusively expressed in glial cells, while the Ca2+ -impermeable GluR2 and GluR3 subunits are present on Purkinje cell membranes (Douyard et al., 2007). During the cerebellar development, these subunits undergo profound changes in their distribution and recently it has been speculated that AMPAR may be involved in stabilizing outgrowing branches of Purkinje cell dendrites (Douyard et al., 2007). The purpose of the present work was to examine whether a single episode of early-life seizures interferes with the developmental program of the cerebellum and GluR2/3 distribution and expression. These questions were explored by treating 10-day-old rats with a single convulsive dose of pentylentetrazole (PTZ). PTZ is a convulsive drug widely used to induce seizures experimentally. Though the mechanism of action of PTZ is not fully understood, it is generally accepted that part of its action is due to its antagonist binding to the picrotoxin-binding site of the postsynaptic GABAA receptor (Macdonald and Barker, 1977). In the rat, the 10th day of life is a crucial age for the development of the cerebellum. At 10 days, the generation of granule cells reaches a peak (Altman, 1972) and synaptogenesis of parallel fibers starts on Purkinje cell dendrites (Zhao et al., 1998). Here we report damage to Purkinje cells, accompanied by inhibition of DNA synthesis and up-regulation of GluR2/3 subunits in the cerebellum of early postnatal rats after a single episode of PTZ-induced seizures.
S. Lomoio et al.
Material and methods Animals and treatment Male 10-day-old Wistar rats (4 animals, date of birth 0) were treated with a single dose of 80 g/g b.w. of pentylentetrazole (PTZ, Sigma, MO, USA) dissolved in saline, applied subcutaneously. Control animals (4 rats from the same litter) received saline only. Immediately after, the animals were injected subcutaneously with 3 H-thymidine (3 HTdR, 2.5 Ci/g b.w., specific activity 600 GBq/mM, UVVVR, Prague, CZ), followed by the same dose 1 h later. All animals were killed by decapitation 4 h after the last 3 HTdR injection, brains immediately excised, fixed in Carnoy solution and embedded in Paraplast X-tra (Polysciences Inc., Warrington, PA, USA). Ten-m thick midsagittal sections were used for autoradiography and immunocytochemistry. Experiments were approved by the Animal Care and Use Committee of the Institute of Physiology of the Academy of Sciences of the Czech Republic. Animal care and experimental procedures were conducted in accordance with the guidelines of the European Community Council directives 86/609/EEC.
Autoradiography Sections were covered with Kodak AR10 stripping film and exposed for 57 days at 4 ◦ C in the dark. Autoradiographs were counterstained with Mayer hematoxylin.
Immunocytochemistry The following antibodies were used for immunocytochemistry: - 1:30 rabbit polyclonal anti-␣-amino-3-hydroxy-5-methyl-Daspartate (AMPA) glutamate receptor 2/3 antiserum (GluR2/3, Chemicon, Temecula, CA, USA). The antibody recognizes both GluR2 and GluR3 AMPA receptor subunits; - 1:5000 guinea pig polyclonal anti-glial glutamate transporter 1 antiserum (GLT1, Chemicon, Temecula, CA, USA). The antibody recognizes the carboxy terminus of GLT1; - 1:5000 rabbit polyclonal anti-calbindin antiserum (CB38, Swant, Bellinzona, CH). In the cerebellum, this antibody stains uniquely Purkinje cells, even in their thinnest processes, giving Golgi-like pictures. Sections were processed as follows. After treatment with 3% H2 O2 in 10% methanol in PBS and with 10% normal goat serum in PBS, the sections were incubated overnight with the primary antibodies. Afterwards the sections were incubated for 30 min in goat anti-rabbit or goat anti-guinea pig biotinylated IgG, followed by streptavidin-HRP complex (Vectastain Elite kit, Vector, CA, USA). The complex was revealed by 3 ,5 -diaminobenzidine tetrahydrochloride. For the reaction control, the primary antibodies were omitted. Sections were photographed with Olympus Camedia C-5050 mounted on Olympus BX51 microscope.
Cell counting To evaluate the extent of cell proliferation in the external germinal layer, 10 sections per animal (4 PTZ treated rats and 4 controls), submitted to autoradiography were scored. The sections were photographed and the number of 3 HTdR labeled cells/100 m of cerebellar cortex convolution was counted in the external germinal layer. Since it is known that the rate of cell proliferation varies between surface of folia and depth of fissures and among lobes (Mares and Lodin, 1970), the cell counting was performed in
A single episode of neonatal seizures alters the cerebellum of immature rats
19
Figure 1 Autoradiography for DNA synthesis in the cerebellum of control and PTZ treated rats. (a) Control rats. In the external germinal layer many cells appear to be labeled. (b) PTZ treated rats. Only a few cells are labeled, mainly in the upper part of the layer. EGL = external germinal layer; ML = molecular layer; Pc = Purkinje cells; IGL = granule cell layer. Scale bars: 40 m.
the middle part of the fissura prima. A convolution length of about 2 mm vas scored in every section. To evaluate the extent of Purkinje cell degeneration in PTZ treated and control rats, twelve sections per animal, immunostained for AMPA GluR2/3 (six sections per animal) and for calbindin (six sections per animal), were selected. The sections were photographed at low magnification and the length of the Purkinje cell layer was measured using the NIH Image program for Macintosh. The total number of Purkinje cells per section was counted at the microscope by using differential interference contrast (DIC) optic at 40× magnification, and divided by the length of the Purkinje cell layer, giving a linear cell density. In addition, in AMPA GluR2/3 immunostained sections, the number of unstained cells was determined and their percentage in relation to the total number of cells was calculated. Statistical evaluation of differences was performed by means of the two-level nested ANOVA on ranks test.
Results Animals treated with PTZ developed generalized tonic—clonic jerk, scratching and body shake seizure activity within few min after injection. The ictal bursts lasted less then 10 min, separated by 5—10 min intervals. The seizure activity lasted about 45 min. All animals survived to seizures. Autoradiographs revealed that cell proliferation in the external germinal layer (EGL) was slightly, but significantly (p < 0.05) decreased in animals treated with PTZ. The number of 3 HTdR labeled cells in the external germinal layer (EGL) was reduced by 12% in comparison with control rats (12.24 ± 2.23 versus 13.92 ± 2.57). No difference was found within groups. Though as previously described (Mares and Lodin, 1970) 3 HTdR labeled cells were not uniformly distributed, but occurred in nests, in PTZ treated animals labeled cells appeared less densely packed (Fig. 1). In control rats, immunocytochemistry for AMPA GluR2/3 faintly decorated the somata and growing dendritic trees of Purkinje cells (Fig. 2(a)), though with a certain degree of variability. Some Purkinje cells had nearly negative soma and the immunoreaction stained the principal branches of the dendrite only (Fig. 2(b)). In PTZ treated rats, in most Purkinje cells the staining was more intense, decorating also the thinnest branches of Purkinje cell dendrites (Fig. 2(c)). However, in some tracts of the cerebellar convolutions,
Purkinje cells were not labeled at all or absent (Fig. 2(d and e)). In other tracts of convolutions, in which the somata of Purkinje cells were not visible even by using phase contrast or DIC optics, in the molecular layer (ML) there were well stained dendrite remnants (Fig. 2(f)). Cell counting revealed that the linear cell density of Purkinje cells was reduced in all treated animals in comparison with control rats, as well evidenced by the graphic representation of single measurements (Fig. 3). Statistic evaluation of data revealed that differences between treated and control animals are highly significant (p < 0.0001) and that the 89.81% of the total variability is ascribable to the treatment. The percentage of unlabeled cells was not coherent in all sections and/or animals, ranging from 15.59% to 29.35% (mean percentage 21.84 ± 4.83). Unlabeled or weakly labeled cells often appeared shrunk, with irregular contour (Fig. 4(a)). Routine hematoxylin staining revealed few apoptotic cells in the EGL of both PTZ treated and control rats (Fig. 4(b)). In addition, only in treated rats, in the Purkinje cell layer some large cell remnants were present among Purkinje cells (Fig. 4(c and d)). After immunocytochemistry for GLT1, in both control and PTZ treated animals, staining was present in Golgi epithelial cell processes enwrapping Purkinje cells and in their Bergman fibers in the ML (Fig. 5). In addition, diffuse staining was visible in the ML in between the branches of Purkinje cell dendritic trees. The staining was more intense in treated animals (Fig. 5(b)). Calbindin immunocytochemistry was performed for additional morphological analysis of Purkinje cells. In the cerebellum, calbindin immunocytochemistry stains exclusively Purkinje cells even in their thinnest dendrites and axon, giving a Golgi-like picture (Fig. 6(a)). The immunostaining confirmed that in PTZ treated rats, in some tracts of convolution Purkinje cells were lost and/or only remnants of their dendritic trees were present (Fig. 6(b)). Moreover, in some small areas the ML displayed a diffuse staining, as to dendritic trees of Purkinje cells underwent lysis (Fig. 6(c)). The observation of the remaining Purkinje cells revealed that in some cells (about 3% of the cell population) the axons bore large spheroids along their course in the underlying granule cell layer (IGL, Fig. 6(d and e)). Cell counting confirmed the data obtained in AMPA GluR2/3 immunostained sections (Fig. 7).
20
S. Lomoio et al.
Figure 2 Immunocytochemistry for Glu2/3 AMPA receptors. (a and b) Control rats. The reaction decorates Purkinje cells somata and growing dendrites. In some tracts of convolutions, some Purkinje cells have negative or faintly stained soma (arrows in b). (c—f) PTZ treated rats. Purkinje cells appear more densely stained in comparison with controls (c). However, loss of Purkinje cells occurred in some tracts of convolution (d and e, arrows). In other tracts, only the dendritic trees are preserved (f). ML = molecular layer; Pc = Purkinje cells; IGL = granule cell layer. Scale bars: 40 m.
A single episode of neonatal seizures alters the cerebellum of immature rats
Figure 3 Graphic representation of single measurements of Purkinje cell linear density (number of cell/mm of convolution, ordinate) obtained in 6 sections per animal, immunostained for AMPA GluR2/3. The maximum value obtained in PTZ-treated animals (PTZ1—4) does not overlap with the minimum value for control animals (Ctr1—4).
Discussion Neonatal seizures can be the cause of learning difficulties and other intellectual or motor disabilities, as well as cerebral atrophy and permanent reduction in the brain cell number (Sarkisian et al., 1997; Mizrahi, 1999; Swann, 2004; Nairismagi et al., 2006). Our data indicate that DNA synthesis and cell proliferation are reduced in the rat cerebellum as soon as 4 h after a single episode of seizures, induced by PTZ administration at 10 days of life. Inhibition of DNA synthesis and cell mitotic activity have been already reported in the brain, including the cerebellum, of rats in which a status epilepticus was induced by bicuculline treatment at 4 days of life and killed 1 or 7 days later (Wasterlain, 1976; Suga and Wasterlain, 1980). However, after 26 days the dam-
21
age was partially recovered. Actually, a restoration of the EGL has been observed after DNA synthesis inhibition by cisdichlorodiammineplatinum treatment (Mares et al., 1986). However, we also found a conspicuous degeneration and loss of Purkinje cells, as well evidenced by both AMPA GluR2/3 and calbindin immunostainings. On the other hand, Purkinje cell loss and injury have been described in the cerebellum of epileptic patients (Leifer et al., 1991). In this context, the observation in calbindin immunostained sections of axonal spheroids in Purkinje cells deserves particular attention. Axonal spheroids have been described in Purkinje cells in numerous murine models of pathologies, such as cerebellar ataxias (Sotelo, 1990; Jeong et al., 2000), acrylamide poisoning (Lehning et al., 2002), Down syndrome (Necchi et al., 2008; Lomoio et al., 2009) and after axotomy (Dusart et al., 1999). The presence of this type of axonal abnormality has been interpreted as a unique Purkinje cell reaction to axon injury, either by chemical insult or physical damage (axotomy) or degeneration and loss of target cells in the cerebellar nuclei. In our model of neonatal seizures, a direct damage to the Purkinje cell axon is difficult to explain. Indeed, observation of cerebellar nuclei in sections taken at the level of the intermediate cortex failed to reveal any sign of damaged cells or cell loss (not shown). Though we cannot exclude a direct toxic action of PTZ, the damage to the axon must be secondary to a seizure insult to the somata and dendritic trees of Purkinje cells. As a consequence, several Purkinje cells underwent degeneration and were quickly removed by the surrounding glial cells. This interpretation, that does not contradict previous data indicating that Purkinje cells are extremely resistant to axonal damage and can survive to axotomy up to 18 months (Dusart et al., 1999; Rossi et al., 2006), indicates that the development of axonal spheroids is not an exclusive feature of axonal injury, but rather reflects a more generalized damage or altered physiology of the cell.
Figure 4 Cell degeneration in the cerebellum of PTZ treated rats. (a) Immunocytochemistry for Glu2/3 AMPA receptors. An example of an unstained Purkinje cell (arrow) near a well stained cell (arrowhead). The contour of the unstained cell appears irregular. Differential interference contrast optic. (b) Mayer hematoxylin staining. In the external germinal layer, some apoptotic cells are present (arrows). (c and d) Mayer hematoxylin staining. Large cell remnants (arrows) can be observed in the Purkinje cell layer. Scale bars: 10 m
22
S. Lomoio et al.
Figure 5 Immunocytochemistry for GLT1 transporter. (a) Control rats. Staining is present in Golgi epithelial cell processes enwrapping Purkinje cell somata and in their Bergman fibers (arrows) in the molecular layer. (b) PTZ treated rats. The staining appears more intense. EGL = external germinal layer; ML = molecular layer; Pc = Purkinje cells. Scale bars: 40 m.
The immunocytochemical data for Glu2/3 AMPA receptors fit in well with this interpretation. AMPA glutamate receptors are a tetrameric assembly of subunits GluR1—4, which are developmentally regulated (Douyard et al., 2007). In the adult rat cerebellum, the postsynaptic densities of climbing and parallel fibers, which represent the main excitatory glutamatergic pathways, on Purkinje cell dendrites contain only the Ca2+ -impermeable receptor complex GluR2/3 (Zhao et al., 1998). This is already present at 10 days of life, when the expression of GluR2/3 subunits is abolished in other cerebellar neurons and becomes an exclusive feature of Purkinje cells (Douyard et al., 2007). After
PTZ administration, the immunocytochemical expression of GluR2/3 was increased in Purkinje cells. So far, inconsistent data are present in literature concerning the changes in GluR2/3 expression following seizures, some papers reporting an increase (Friedman et al., 2007), a decrease (Jiang et al., 2008), or no effect (Cornejo et al., 2007). These differences may be due to the type and timing of the epileptic insult, the brain area considered, or the different models utilized in the experiments (rats or cell cultures). As to the cerebellum, as far as we know, no studies have been addressed to the changes in Glu2/3 AMPA receptors after seizures. Once again, we cannot exclude that the increase
Figure 6 Immunocytochemistry for calbindin. (a) Control rats. Purkinje cell somata, dendrites and axons are intensely stained. (b—e) PTZ treated rats. In some tracts of convolution, Purkinje cells are missed, though some remnants of branches of their dendritic trees are still present (arrow in b). In other tracts of convolutions Purkinje cells are unstained or only faintly stained and diffuse reaction product is present in the molecular layer occupied by Purkinje cell dendritic trees (c). Some axons of Purkinje cells bear large spheroids (arrows in d and e). Scale bars: a—d: 40 m; e: 10 m
A single episode of neonatal seizures alters the cerebellum of immature rats
Figure 7 Graphic representation of single measurements of Purkinje cell linear density (number of cell/mm of convolution, ordinate) obtained in 6 sections per animal, immunostained for calbindin. The maximum value obtained in PTZ-treated animals (PTZ1—4) does not overlap with the minimum value for control animals (Ctr1—4).
in the immunocytochemical detection of AMPA GluR2/3 is due to a direct toxic effect of PTZ. However, studying the metabolic glutamate turnover in the cerebellum of PTZ treated young-adult rats. Eloqayli et al. (2003, 2004) found that after a subconvulsive dose (20 mg/kg) the cerebellum was only slightly affected, whereas after the convulsive dose of 70 mg/kg the cerebellum was the most susceptible area of the CNS. According to the authors the data indicate that the changes observed in the cerebellum are caused by seizures and not PTZ. Moreover, the cerebellum has been shown to be a target for PTZ-induced seizure. In fact, subconvulsive administration of PTZ is able to induce long lasting paroxysmal EEG activity not only in the cortex and midbrain, but also in the cerebellum (Sierra-Paredes et al., 1989). On the other hand, changes in GluR2/3 immunocytochemical detection or expression have been already described in the cerebellum in other pathologies. In fact, a decrease in the total expression of GluR2/3, accompanied by increased immunoreactivity in the surviving Purkinje cells, has been described in the cerebellum of Creutzfeldt—Jakob disease patients (Ferrer and Puig, 2003). Similarly, in rats in which hepatic encephalopathy was induced by portacaval shunt, the number of GluR2/3-positive Purkinje cells was decreased after 1 month, but increased after 6 months (Suarez et al., 1997). Again, in an experimental model of hypobaric hypoxia, both the levels of GluR2 and GluR3 mRNAs and the immunocytochemical detection of GluR2/3 were enhanced in Purkinje cells (Kaur et al., 2005). Finally, a single fear-inducing stimulus induced the appearance of GluR2-containing AMPA receptors in parallel fiber/stellate interneuron synapses in mice, a localization where this AMPA receptor subunit is normally unexpressed (Liu et al., 2010). The high density of GluR2/3 labeling on Purkinje cells observed by us and reported in the above papers may represent a functional reorganization or intracellular trafficking of AMPA GluR2/3 aimed at reducing the Ca2+ influx caused by an excess in the extracellular glutamate levels. In fact, it is known that in several pathological states, such as ischemia, hypoxia, hypoglycemia, epilepsy and some neurodegenerative diseases, extracellular glutamate levels are increased (Druga et al., 2005; Caudle and Zhang, 2009; Wang and Qin,
23
2010). This increase would result in excessive accumulation of Ca2+ in a cell and excitotoxicity (Olney, 1978; Gonsette, 2008). In this view, we should hypothesize that as soon as a neuron succeeds in expressing more Ca2+ -impermeable AMPA glutamate receptors, the probability to survive to the insult would be enhanced. Consequently, the GluR2/3 immunonegative cells, which contours often appear irregular, would be destined to death. The increase in GLT1 immunoreactivity of Golgi epithelial cells is probably aimed at a faster removing of extracellular glutamate. In conclusion, we have here reported for the first time that a single episode of seizures is able to damage the cerebellum of 10-day-old rats, leading to loss of Purkinje cells as soon as 4 h after the insult. Since at 10 days of life Purkinje cells are no more proliferating, the loss of Purkinje cells should be permanent. Further studies are needed to verify whether further cell loss and circuit remodeling occur at later stages and whether the morphological and AMPA receptor changes are recovered.
Acknowledgements We wish to thank Prof. L.A. Zonta, Department of Genetics and Microbiology, University of Pavia, and Prof. G. Bogliani, Department of Animal Biology, University of Pavia for the precious advices on statistical analysis. This work was supported in part by Project AV0Z50110509 to V.M.
References Altman, J., 1972. Postnatal development of the cerebellar cortex in the rat. 3. Maturation of the components of the granular layer. J. Comp. Neurol. 145, 465—513. Carmody, S., Brennan, L., 2010. Effects of pentylentetrazoleinduced seizures on metabolomic profiles of rat brain. Neurochem. Int. 56, 340—344. Caudle, W., Zhang, J., 2009. Glutamate, excitotoxicity, and programmed cell death in Parkinson disease. Exp. Neurol. 220, 230—233. Cornejo, B., Mesches, M., Coultrap, S., Browning, M., Benke, T., 2007. A single episode of neonatal seizures permanently alters glutamatergic synapses. Ann. Neurol. 61, 411—426. Douyard, J., Shen, L., Huganir, R., Rubio, M., 2007. Differential neuronal and glial expression of GluR1 AMPA receptor subunit and the scaffolding proteins SAP97 and 4.1N during rat cerebellar development. J. Comp. Neurol. 502, 141—156. Druga, R., Mares, P., Otahal, J., Kubova, H., 2005. Degenerative neuronal changes in the rat thalamus induced by status epilepticus at different developmental stages. Epilepsy Res. 63, 43—65. Druga, R., Mares, P., Kubova, H., 2010. Time course of neuronal damage in the hippocampus following lithium-pilocarpine status epilepticus in 12-day-old rats. Brain Res. 1355, 174—179. Dusart, I., Morel, M., Wehrlè, R., Sotelo, C., 1999. Late axonal sprouting of injured Purkinje cells and its temporal correlation with permissive changes in the glial scar. J. Comp. Neurol. 408, 399—418. Eloqayli, H., Dahl, C.D., Gotestam, K.G., Unsgard, G., Hadidi, H., Sonnewald, U., 2003. Pentylentetrazole decreases metabolic glutamate turnover in rat brain. J. Neurochem. 85, 1200—1207. Eloqayli, H., Dahl, C.D., Gotestam, K.G., Unsgard, G., Sonnewald, U., 2004. Changes of glial—neuronal interaction and metabolism after a subconvulsive dose of pentylentetrazole. Neurochem. Int. 45, 739—745.
24 Ferrer, I., Puig, B., 2003. GluR2/3, NMDA1 and GABAA receptors in Creutzfeldt—Jakob disease. Acta Neuropathol. 106, 311—318. Friedman, L., Avallone, J., Magrys, B., 2007. Maturational effects of single and multiple early-life seizures on AMPA receptors in prepubescent hippocampus. Dev. Neurosci. 29, 427—437. Gonsette, R., 2008. Neurodegeneration in multiple sclerosis: the role of oxidative stress and excitotoxicity. J. Neurol. Sci. 274, 48—53. Haas, K., Sperber, E., Opanashuk, L., Stanton, P., Moshè, S., 2001. Resistance of immature hippocampus to morphologic and physiologic alterations following status epilepticus or kindling. Hippocampus 11, 615—625. Holmes, G., Thompson, J., Marchi, T., Feldman, D., 1988. Behavioral effects of kainic acid administration on the immature brain. Epilepsia 29, 721—730. Jeong, Y., Hyun, B., Hawkes, R., 2000. Abnormalities in cerebellar Purkinje cells in the novel ataxic mutant mouse, pogo. Brain Res. Dev. Brain Res. 125, 61—67. Jiang, Q., Wang, J., Wu, Y., Wu, X., Qin, J., Jiang, Y., 2008. Early-life epileptiform discharges exert both rapid and long-lasting effects on AMPAR subunit composition and distribution in developing neurons. Neurosci. Lett. 444, 31—35. Kaur, C., Sivakumar, V., Singh, G., Singh, J., Ling, E., 2005. Response of Purkinje neurons to hypobaric hypoxic exposure as shown by alteration in expression of glutamate receptors, nitric oxide synthases and calcium binding proteins. Neuroscience 135, 1217—1229. Koh, S., Tibayan, F., Simpson, J., Jensen, F., 2004. NBQX or topiramate treatment after perinatal hypoxia-induced seizures prevents later increases in seizure-induced neuronal injury. Epilepsia 45, 569—575. Lehning, E., Balaban, C., Ross, J., Reid, M., LoPachin, R., 2002. Acrylamide neuropathy. I. Spatiotemporal characteristics of nerve cell damage in rat cerebellum. Neurotoxicology 23, 397—414. Leifer, D., Cole, D., Kowall, N., 1991. Neuropathologic asymmetries in the brain of a patient with a unilateral status epilepticus. J. Neurol. Sci. 103, 127—135. Liu, Y., Formisano, L., Savtchouk, I., Takayasu, Y., Szabo, G., Zukin, R., Liu, S., 2010. A single fear-inducing stimulus induces a transcription-dependent switch in synaptic AMPAR phenotype. Nat. Neurosci. 13, 223—231. Lomoio, S., Scherini, E., Necchi, D., 2009. Beta-amyloid overload does not directly correlate with SAPK/JNK activation and tau protein phosphorylation in the cerebellar cortex of Ts65Dn mice. Brain Res. 1297, 198—206. Lynch, M., Sayin, U., Bownds, J., Janumpalli, S., Sutula, T., 2000. Long-term consequences of early postnatal seizures on hippocampal learning and plasticity. Eur. J. Neurosci. 12, 2252—2264. Macdonald, R., Barker, J., 1977. Pentylenetetrazol and penicillin are selective antagonists of GABA-mediated post-synaptic inhibition in cultured mammalian neurones. Nature 267, 720—721. Mares, V., Lodin, Z., 1970. The cellular kinetics of the developing mouse cerebellum. II. The function of the external granular layer in the process of gyrification. Brain Res. 23, 343—352.
S. Lomoio et al. Mares, V., Scherini, E., Biggiogera, M., Bernocchi, G., 1986. Influence of cis-dichlorodiammineplatinum on the structure of the immature rat cerebellum. Exp. Neurol. 91, 246—258. Mizrahi, E., 1999. Acute and chronic effects of seizures in the developing brain: lessons from clinical experience. Epilepsia 40 (Suppl. 1), S42—50. Nairismagi, J., Pitkanen, A., Kettunen, M., Kauppinen, R., Kubova, H., 2006. Status epilepticus in 12-day-old rats leads to temporal lobe neurodegeneration and volume reduction: a histologic and MRI study. Epilepsia 47, 479—488. Necchi, D., Lomoio, S., Scherini, E., 2008. Axonal abnormalities in cerebellar Purkinje cells of the Ts65Dn mouse. Brain Res. 1238, 181—188. Olney, J., 1978. Neurotoxicity of amino acids. In: McGeer, E.G., Olney, J.W., McGeer, P.L. (Eds.), Kainic Acid as a Tool in Neurobiology. Raven Press, New York, pp. 95—121. Rossi, F., Gianola, S., Corvetti, L., 2006. The strange case of Purkinje axon regeneration and plasticity. Cerebellum 5, 174—182. Sanchez, R., Koh, S., Rio, C., Wang, C., Lamperti, E., Sharma, D., Corfas, G., Jensen, F., 2001. Decreased glutamate receptor 2 expression and enhanced epileptogenesis in immature rat hippocampus after perinatal hypoxia-induced seizures. J. Neurosci. 21, 8154—8163. Sarkisian, M., Tandon, P., Liu, Z., Yang, Y., Hori, A., Holmes, G., Stafstrom, C., 1997. Multiple kainic acid seizures in the immature and adult brain: ictal manifestations and long-term effects on learning and memory. Epilepsia 38, 1157—1166. Sierra-Paredes, G., Soto-Otero, R., Mendez-Alvarez, E., GalanValiente, J., Aguilar-Veiga, E., Sierra-Marcuno, G., 1989. Experimental spike-and-wave discharges induced by pentylentetrazole and tolerance to repeated injections: an electrophysiological and biochemical study. Epilepsy Res. 4, 139—146. Sotelo, C., 1990. Axonal abnormalities in cerebellar Purkinje cells of the ‘hyperspiny Purkinje cell’ mutant mouse. J. Neurocytol. 19, 737—755. Suarez, I., Bodega, G., Fernandez, B., 1997. Modulation of AMPA receptor subunits GluR1 and GluR2/3 in the rat cerebellum in an experimental hepatic encephalopathy model. Brain Res. 778, 346—353. Suga, S., Wasterlain, C., 1980. Effects of neonatal seizures or anoxia on cerebellar mitotic activity in the rat. Exp. Neurol. 67, 573—580. Swann, J., 2004. The effects of seizures on the connectivity and circuitry of the developing brain. Ment. Retard. Dev. Disabil. Res. Rev. 10, 96—100. Vander, T., Medvedovsky, M., Herishanu, Y., 2004. Epilepsia partialis continua possibly caused by cerebellar lesion. Cerebellum 3, 126—128. Wang, Y., Qin, Z., 2010. Molecular and cellular mechanisms of excitotoxic neuronal death. Apoptosis, doi:10.1007/s10495-0100481-0. Wasterlain, C., 1976. Effects of neonatal status epilepticus on rat brain development. Neurology 26, 975—986. Zhao, H., Wenthold, R., Petralia, R., 1998. Glutamate receptor targeting to synaptic populations on Purkinje cells is developmentally regulated. J. Neurosci. 18, 5517—5528.