The circadian dynamics of the hippocampal transcriptome and proteome is altered in experimental temporal lobe epilepsy

Gene and protein expressions display circadian oscillations, which can be disrupted in diseases in most body organs. Whether these oscillations occur in the healthy hippocampus and whether they are altered in epilepsy are not known. We identified more than 1200 daily oscillating transcripts in the hippocampus of control mice and 1600 in experimental epilepsy, with only one-fourth oscillating in both conditions. Comparison of gene oscillations in control and epilepsy predicted time-dependent alterations in energy metabolism, which were verified experimentally. Although aerobic glycolysis remained constant from morning to afternoon in controls, it increased in epilepsy. In contrast, oxidative phosphorylation increased in control and decreased in epilepsy. Thus, the control hippocampus shows circadian molecular remapping, which is altered in epilepsy. We suggest that the hippocampus operates in a different functioning mode in epilepsy. These alterations need to be considered when studying epilepsy mechanisms, designing drug treatments, and timing their delivery.

Is mossy fiber sprouting present at the time of the first spontaneous seizures in rat experimental temporal lobe epilepsy?

The contribution of mossy fiber sprouting to the generation of spontaneous seizures in the epileptic brain is under dispute. The present study addressed this question by examining whether sprouting of mossy fibers is present at the time of appearance of the first spontaneous seizures in rats, and whether all animals with increased sprouting have spontaneous seizures. Epileptogenesis was induced in 16 rats by electrically stimulating the lateral nucleus of the amygdala for 20-30 min until the rats developed self-sustained status epilepticus (SSSE). During and after SSSE, rats were monitored in long-term by continuous video-electroencephalography until they developed a second spontaneous seizure (8-54 days). Thereafter, monitoring was continued for 11 days to follow seizure frequency. The density of mossy fiber sprouting was analyzed from Timm-stained preparations. The density of hilar neurons was assessed from thionin-stained sections. Of 16 rats, 14 developed epilepsy. In epileptic rats, the density of mossy fiber sprouting did not correlate with the severity or duration (115-620 min) of SSSE, delay from SSSE to occurrence of first (8-51 days) or second (8-54 days) spontaneous seizure, or time from SSSE to perfusion (20-63 days). In the temporal end of the hippocampus, the sprouting correlated with the severity of neuronal damage (ipsilateral: r = -0.852, P < 0.01 contralateral: r = -0.748, P < 0.01). The two animals without spontaneous seizures also had sprouting. Increased density of sprouting in animals without seizures, and its association with the severity of neuronal loss was confirmed in another series of 30 stimulated rats that were followed-up with video-EEG monitoring for 60 d. Our data indicate that although mossy fiber sprouting is present in all animals with spontaneous seizures, its presence is not necessarily associated with the occurrence of spontaneous seizures.

Status epilepticus causes necrotic damage in the mediodorsal nucleus of the thalamus in immature rats

Status epilepticus (StE) in immature rats causes long-term functional impairment. Whether this is associated with structural alterations remains controversial. The present study was designed to test the hypothesis that StE at an early age results in neuronal loss. StE was induced with lithium-pilocarpine in 12-d-old rats, and the presence of neuronal damage was investigated in the brain from 12 hr up to 1 week later using silver and Fluoro-Jade B staining techniques. Analysis of the sections indicated consistent neuronal damage in the central and lateral segments of the mediodorsal nucleus of the thalamus, which was confirmed using adjacent cresyl violet-stained preparations. The mechanism of thalamic damage (necrosis vs apoptosis) was investigated further using TUNEL, immunohistochemistry for caspase-3 and cytochrome c, and electron microscopy. Activated microglia were detected using OX-42 immunohistochemistry. The presence of silver and Fluoro-Jade B-positive degenerating neurons in the mediodorsal thalamic nucleus was associated with the appearance of OX-42-immunopositive activated microglia but not with the expression of markers of programmed cell death, caspase-3, or cytochrome c. Electron microscopy revealed necrosis of the ultrastructure of damaged neurons, providing further evidence that the mechanism of StE-induced damage in the mediodorsal thalamic nucleus at postnatal day 12 is necrosis rather than apoptosis. Finally, these data together with previously described functions of the medial and lateral segments of the mediodorsal thalamic nucleus suggest that some functions, such as adaptation to novelty, might become compromised after StE early in development.

Association between the density of mossy fiber sprouting and seizure frequency in experimental and human temporal lobe epilepsy

PURPOSE

If the sprouting of granule cell axons or mossy fibers in the dentate gyrus is critical for the generation of spontaneous seizures in temporal lobe epilepsy (TLE), one could hypothesize that epileptic animals or humans with increased sprouting would have more frequent seizures. This hypothesis was tested by analyzing the data gathered from experimental and human epilepsy.

METHODS

In experiment I (rats with "newly diagnosed" TLE), self-sustained status epilepticus was induced in rats by electrically stimulating the amygdala. Thereafter, the appearance of spontaneous seizures was monitored by continuous video-electroencephalography (EEG) until the animal developed two spontaneous seizures and for 11 d thereafter. Rats were perfused for histology, and mossy fibers were stained using the Timm method. In experiment II (rats with "recently diagnosed" TLE), status epilepticus was induced in rats and the development of seizures was monitored by video-EEG for 24 h/d every other day for 60 days. All animals were then perfused for histology. In experiment III (rats with "chronic" TLE), animals were monitored by video-EEG for 24 h/d every other day for 6 months before histologic analysis. To assess mossy fiber sprouting in human TLE, hippocampal sections from 31 patients who had undergone surgery for drug-refractory TLE were stained with an antibody raised against dynorphin.

RESULTS AND CONCLUSIONS

Our data indicate that the density of mossy fiber sprouting is not associated with the total number of lifetime seizures or the seizure frequency in experimental or human TLE.

GABA(A)-mediated toxicity of hippocampal neurons in vitro

In the present study, we examined whether the elevation of GABA by gamma-vinyl-GABA protects cultured rat fetal hippocampal neurons against toxicity induced by a 20-min incubation with 100 microM L-glutamate. Neither a 24-h pretreatment nor posttreatment with gamma-vinyl-GABA (100 microM) had any neuroprotective effects, as determined by counting microtubule-associated protein-2 positive cells and lactate dehydrogenase assay 24 h after the glutamate treatment. Unexpectedly, gamma-vinyl-GABA alone induced a 20% loss of microtubule-associated protein-2-positive cells in a culture that was grown in medium containing 25 mM KCl. The toxic effect of gamma-vinyl-GABA was mimicked by a 24-h treatment with GABA (100 microM) and the GABA(A) receptor agonist, muscimol (10 microM), but not the GABA(B) receptor agonist, baclofen (10 microM). The GABA(A) receptor antagonist, bicuculline (10 microM), protected against gamma-vinyl-GABA and GABA-evoked toxicity. Neither gamma-vinyl-GABA nor GABA was toxic in culture medium containing 15 mM KCl. These data indicate that, under depolarizing conditions, an increased GABA level is toxic for a subpopulation of developing hippocampal neurons in vitro. The effect is GABA(A) receptor-mediated. These data provide a new view for understanding neurodegenerative processes, and raise a question of the safety of therapies aimed at increasing GABA concentration following brain insults, especially in immature brains.

Status epilepticus-induced neuronal damage in the rat amygdaloid complex: distribution, time-course and mechanisms

The present study was designed to elucidate the distribution, time-course and mechanism(s) of status epilepticus-induced neuronal damage in the rat amygdaloid complex. Status epilepticus was induced with kainate (9 mg/kg, i.p.), and the behavioral and electrographic seizure activity of each rat was monitored via cortical electrodes attached to a continuous video electrocorticogram system. Rats were subsequently perfused 1, 2, 4, 8, 16, 24 or 48 h after kainate injection. The first signs of amygdaloid damage were seen in rats perfused 4 h after kainate injection, though the severity and temporal appearance of damage varied substantially between the different amygdaloid nuclei and their subdivisions. Second, terminal transferase dUTP nick-end labeling (TUNEL)-positive nuclei and laddering of DNA in gel electrophoresis appeared in the amygdala 8 and 16 h after kainate, respectively. The distribution and density of TUNEL-positive nuclei in the different amygdaloid nuclei correlated with the distribution of neuronal damage in Thionin- and silver-stained sections. Third, the immunoreactivity of Bax protein, a promoter of apoptotic neuronal death, increased in the vulnerable medial division of the lateral nucleus prior to the appearance of argyrophilic neurons and TUNEL-positive nuclei. Fourth, the severity of neuronal damage progressed in some, but not all, amygdaloid regions throughout the 48-h follow-up, even though the occurrence of high-amplitude and frequency discharges, which are typically associated with behavioral seizure activity, extinguished after 7 h. These data show that status epilepticus-induced neuronal damage in the amygdala is a dynamic region-specific process, the severity of which depends on the duration of seizure activity. At least one mechanism underlying the damage involves apoptosis, which continues long after the behavioral and electrographic seizures have subsided.

Defensive conditioning-related increase in AP-1 transcription factor in the rat cortex

In the studies reported herein, electrophoretic mobility shift assay (EMSA) and immunocytochemistry have been applied to document increased levels of AP-1 transcription factor, and its major component, c-Fos in the rat brain following behavioral training of two-way active avoidance. A single training session (50 trials) provoked elevation of AP-1 in the visual, sensory and limbic cortex but not in the hippocampus. A session following long term training (10 sessions, up to asymptotic level of performance) had much smaller effect on AP-1 levels in the visual cortex than single training session. The long term training was used to ensure that observed effects were related to acquisition of the reaction rather than simply to behavioral performance. Supershift EMSA analysis with antibodies directed at individual AP-1 components revealed that AP-1 extracted from the brains of trained as well as naive animals is composed of the same proteins, i.e., in order of relative level within the protein family: c-Fos, Fos B, Fra-2, and Jun D, Jun B, c-Jun. These studies reinforce the notion that transcription factors as regulators of gene expression-and AP-1 in particular-may respond to behavioral stimulation and furthermore may play a role in acquisition of behavioral reactions.

Kainate-evoked modulation of gene expression in rat brain

Kainate is a glutamate analog that produces neuronal excitation resulting in seizures within hours following its intraperitoneal injection into adult rats. Then, at 2-3 days after the treatment, neurodegeneration of apoptotic character can be observed in limbic system. As a consequence, plastic reorganization and glial reactivation phenomena occur. These physiological and pathological responses are reflected by specific changes in gene expression, that can be dissected according to their spatio-temporal patterns. The early phase of gene expression observed in all hippocampal subfields appears to reflect a sudden burst of spiking activity. Changes in mRNA levels restricted to dentate gyrus are suggestive of a link to neuronal plasticity. The late gene expression response implies its correlation either to neuronal cell death or glial reactivation, depending on cellular localization of gene products. Thus analysis of the temporal and spatial gene expression pattern in the hippocampus after kainate treatment may provide clues revealing specific phenomena to which gene expression could be attributed.

Kainate-evoked changes in dystrophin messenger RNA levels in the rat hippocampus

Dystrophin and dystroglycan messenger RNAs are expressed in specific brain areas, including regions of the cortex and the hippocampus, and in such neurons dystrophin has been localized to postsynaptic densities. In the present study we examined by in situ hybridization the effect of neuronal activation and neurotoxicity induced by kainate and pentylenetetrazole administered in vivo on dystrophin and dystroglycan expression in the rat brain. Kainate injection resulted in a transient but dramatic decrease in dystrophin transcript levels in the dentate gyrus granule cells, neurons not affected by kainate neurotoxicity, 6 h after injection. There was also a strong, concomitant increase in dystrophin messenger RNA levels in the CA3 subfield. At 24-72 h after kainate injection, the dystrophin transcript in the dentate granule cells returned to control levels, while it decreased gradually in the CA subfields, coinciding with the neurodegeneration observed in these areas. Comparable results were obtained with pan-dystrophin probes and probes specific to the short, G-dystrophin (Dp71) isoform that predominates in the dentate gyrus. This indicates that any dystrophin transcript that might be expressed in these areas responds to kainate in the same manner. In contrast, kainate insult had no significant effect on the dystroglycan messenger RNA levels in these hippocampal areas at 6 h post-injection. At later times. however, there was a gradual decrease in the dystroglycan messenger RNA in those areas which respond to the kainate insult with extensive neuronal death. For comparison, seizures which are not associated with progressive neurodegeneration were induced by pentylenetetrazole: in this situation the dystrophin and dystroglycan messenger RNA levels remained unchanged in all areas of the hippocampal formation. Since activation of glutamate receptors is thought to be involved in some forms of synaptic plasticity in the adult hippocampus, our data indicate that the dystrophin gene behaves as a candidate plasticity-related gene responding to glutamate.

Kainate-evoked modulation of gene expression in rat brain

Kainate is a glutamate analog that produces neuronal excitation resulting in seizures within hours following its intraperitoneal injection into adult rats. Then, at 2-3 days after the treatment, neurodegeneration of apoptotic character can be observed in limbic system. As a consequence, plastic reorganization and glial reactivation phenomena occur. These physiological and pathological responses are reflected by specific changes in gene expression, that can be dissected according to their spatio-temporal patterns. The early phase of gene expression observed in all hippocampal subfields appears to reflect a sudden burst of spiking activity. Changes in mRNA levels restricted to dentate gyrus are suggestive of a link to neuronal plasticity. The late gene expression response implies its correlation either to neuronal cell death or glial reactivation, depending on cellular localization of gene products. Thus analysis of the temporal and spatial gene expression pattern in the hippocampus after kainate treatment may provide clues revealing specific phenomena to which gene expression could be attributed.

Differential expression of syntrophins and analysis of alternatively spliced dystrophin transcripts in the mouse brain

Expression of syntrophin genes, encoding members of the dystrophin-associated protein complex, was studied in the mouse brain. In the hippocampal formation there is distinctive co-localization of specific syntrophins with certain dystrophin isoforms in neurons, e.g. alpha1-syntrophin with the C-dystrophin in CA regions and beta2-syntrophin with the G-dystrophin in the dentate gyrus. Expression of the alpha1-syntrophin is predominant in CA regions and the olfactory bulb and it is also present in the cerebral cortex and the dentate gyrus. The beta2-syntrophin mRNA is most abundant in the dentate gyrus and is also evident in the pituitary, the cerebral cortex and in Ammon's horn and in traces in the caudate putamen. The choroid plexus was labelled by both alpha1- and beta2-syntrophin-specific probes. The expression of syntrophins in the brain correlates with expression of dystrophins and dystroglycan. There are brain areas such as the cerebral cortex where several different syntrophins and dystrophins are expressed together. Syntrophin expression co-localizes with utrophin in the choroid plexus and caudate putamen. Finally, no syntrophin was detected in the cerebellar Purkinje cells where the specific dystrophin isoform (P-type) is present. This specific distribution of syntrophins in the brain is particularly interesting, as muscle syntrophin interacts with neuronal nitric oxide synthase. This may suggest that the dystrophin-associated protein complex may be involved in synaptic organisation and signal transduction machinery in both muscle and neurons. The dystrophin isoform, with exons 71-74 spliced out and hence lacking syntrophin binding sites, had been believed to be predominant in the brain, but our analyses using in situ hybridization, S1 nuclease protection and the semi-quantitative polymerase chain reaction revealed that this alternatively spliced mRNA is a minor, low abundance form in the brain.

Seizure related changes in the regulation of opioid genes and transcription factors in the dentate gyrus of rat hippocampus

An in situ hybridization study showed that limbic seizures induced by kainate strongly augmented the prodynorphin and proenkephalin messenger RNA levels in granular cells of the rat hippocampal dentate gyrus. Pentylenetetrazole increased the level of proenkephalin messenger RNA, but slightly decreased that of prodynorphin messenger RNA in the dentate gyrus. Administration of kainate to rats caused a profound increase in messenger RNAs of the transcription factor genes c-fos and c-jun in the dentate gyrus, followed by an increase in the level of the transcriptional complex activator protein-1 in hippocampal neurons. Pentylenetetrazole also elevated the formation of activator protein-1, but the effect appeared earlier than that induced by kainate. Thus, recurrent limbic seizures activate both prodynorphin and proenkephalin genes, whereas generalized clonic-tonic seizures seem to decrease the prodynorphin and increase the proenkephalin gene expression in the dentate gyrus. Furthermore, our present results suggest that the transcription factors, c-fos, c-jun and activator protein-1 complex may be involved in the process of inducing the hippocampal proenkephalin gene, while these factors might be differently involved in regulation of prodynorphin gene expression.

Modulation of ischemic signal by antagonists of N-methyl-D-aspartate, nitric oxide synthase, and platelet-activating factor in gerbil hippocampus

Cerebral ischemia in the gerbil results in early hippocampal changes, which include transient activation and/or translocation of protein kinase C (PKC), increased enzymatic activity of ornithine decarboxylase (ODC), and elevated DNA binding ability of activator protein-1 (AP1). The time-course of all three of these postischemic responses was found to be almost parallel, peaking at 3 hr after the ischemic insult. The effectiveness of known modulators of postischemic morphological outcome (MK-801, L-NAME, and gingkolides BN 52020 and BN 52021) in counteracting the induction of PKC, ODC, and AP1 formation was tested. These drugs were administrated as followed: MK-801 (a noncompetitive inhibitor of NMDA channel), 0.8 mg/kg i.p., 30 min before ischemia, and 5 min after the insult; L-NAME (competitive inhibitor of NO synthase), 10 mg/kg i.p., 30 min before ischemia, and 5 mg/kg, 5 min after ischemia; BN52020 and BN52021 (inhibitors of platelet-activating factor: PAF receptors) were administered as a suspension in 5% ethanol in water by oral route, 10 mg/kg for 3 days before ischemia. Three of these drugs, MK-801, L-NAME, and BN52021, significantly reduced ischemia-elevated activity of PKC and ODC, whereas AP1 formation was only partially attenuated. Our observations implicate the existence of different mechanism(s) for postischemic PKC and ODC activation, which in turn is engaged in AP1 induction.

Inducible and constitutive transcription factor NF-kappa B-like DNA binding activities in rat brain cells cultured in vitro

Description of constitutive and inducible transcription factors in brain cells is a prerequisite for understanding phenomena of long-term neuronal plasticity. In this report we report that basal levels of NF-kappa B-like transcription factor DNA binding activity (measured by electrophoretic mobility shift assay) are observed in two kinds of cultured rat brain cells, namely neurons of fetal cerebral hemispheres and astroglia derived from newborn brains. In the latter culture, phorbol ester induces additional forms of DNA binding activity, whereas L-glutamate fails to do so. In neurons, neither treatment is effective in inducing NF-kappa B-like DNA binding activity over the basal level. These results are in contrast to the fact that L-glutamate in both neurons and glia elevates DNA binding activity of AP-1, another transcription factor. These data, besides describing behavior of NF-kappa B-like DNA binding activities, also provide some evidence that L-glutamate exerts its modulatory functions on neurons and glia through specific transcription factors like AP-1.

AP-1 and CRE DNA binding activities in rat brain following pentylenetetrazole induced seizures

Pentylenetetrazole (PTZ) evoked seizures, known to be dependent on stimulation of excitatory amino acids (EAA) receptors, serve as a useful model to study genomic responses to increased brain activity. It is believed that these responses form the basis for long term modifications in neuronal functions. Formation of the AP-1 transcription factor genes and proteins in hippocampal cells is the best known example of a genomic response to PTZ seizures and to an activation of the EAA receptors. In the studies reported herein electrophoretic mobility shift assay (EMSA) was employed to investigate levels of AP-1 DNA binding activity in various regions of the rat brain following PTZ seizures and these levels were compared to the cyclic AMP responsive element (CRE) DNA binding activity. A dramatic increase of the AP-1 DNA binding activity was observed in the hippocampus and in sensory and limbic cortices, and to much lesser extent in the cerebellum. The EMSA supershift method provided an evidence that Jun B and c-Fos and probably Fos B are major components of AP-1 at 2 h after the seizures. In none of the structure investigated, clear modulation of CRE DNA binding activity was noted. These data are discussed in the context of CRE and AP-1 DNA binding crossreactivity.

Glutamate receptor-driven activation of transcription factors in primary neuronal cultures

We have used primary neuronal cultures prepared from fetal cerebral hemispheres to investigate the effects of different glutamate receptor agonists and antagonists on the expression of transcription factor encoding genes, such as c-fos, fosB, c-jun, junB, junD, c-myc, and zif/268. The addition of glutamate (100 microM) to the culture medium rapidly activated c-fos, fosB, c-jun, junB and zif/268 gene expression, reaching the maximal level at 30-60 minutes for zif/268 and at 60 minutes for the other genes. The onset of fosB mRNA accumulation was slightly delayed in comparison to the other genes. No clear induction was found for junD and c-myc. Different glutamate receptor agonists, such as NMDA, kainate, quisqualate, trans-(+/-)-1-aminocyclopentane-1,3-dicarboxylic acid (t-ACPD) and alpha-amino-3-hydroxy-5-methyl-4-isoxazole-propionate (AMPA) were able to increase c-fos, c-jun, and zif-268 mRNA levels with rapid and transient kinetics similar to those observed after glutamate treatment. Similar results were obtained for junB and fosB after kainate and quisqualate stimulation. Pretreatment with MK-801, a non competitive NMDA antagonist, produced an almost complete inhibition of glutamate-driven expression of transcription factor genes, thus suggesting that NMDA receptor plays a major role in glutamate induced-gene expression. On the contrary the kainate/AMPA receptor antagonist, DNQX, did not influence glutamate induced-gene expression. Under the conditions used in the present study, NMDA was effective in inducing the simultaneous activation of several IEGs even when added to the culture medium containing millimolar concentration of magnesium. When experiments were performed in Krebs solution, NMDA was effective in stimulating zif/268 and c-fos mRNAs only in the absence of Mg2+, while glutamate activated c-fos and zif/268 both in the presence and absence of magnesium ions. As expected, NMDA effect was fully inhibited by MK-801. The level of AP-1 DNA binding activity, as measured by electrophoretic mobility shift assay, increased after addition of glutamate and NMDA to cultured neurons and such increase was antagonized by the pretreatment with MK-801.

Induction of primary response genes by excitatory amino acid receptor agonists in primary astroglial cultures

We have characterized the genomic response of astroglial cells to excitatory amino acids by using selective agonists and antagonists for the various receptor subtypes and by analyzing different primary response genes, such as members of the Fos (c-fos and fosB) and Jun (c-jun, junB, and junD) families, zif/268, and c-myc. A rapid and transient elevation of mRNA levels for c-fos, fosB, c-jun, junB, and zif/268 was observed after addition of glutamate to cultured astrocytes, whereas junD and c-myc expression was not affected. The level of AP-1 DNA binding activity, as measured by the electrophoretic mobility shift assay, also increased after addition of glutamate to cultured astrocytes. Glutamate-induced c-fos expression was not affected by the N-methyl-D-aspartate receptor antagonists MK-801 and D-2-amino-5-phosphonopentanoate, by the kainate/alpha-amino-3-hydroxy-5- methylisoxazole-4-propionate (AMPA) receptor antagonist 6,7-dinitroquinoxaline-2,3-dione (DNQX), or by the broad-spectrum antagonist kynurenate. Kainate and AMPA were also effective in inducing primary response gene expression, and their actions were antagonized by kynurenate and DNQX but not by MK-801. 1S,3R-1-Aminocyclopentane-1,3-dicarboxylic acid, a selective agonist for the metabotropic glutamate receptor, induced primary response gene expression, but its action was not antagonized by different glutamate antagonists, including L-2-amino-3-phosphonopropionate. In conclusion, our data suggest that cultured astrocytes express both kainate/AMPA ionotropic receptors and metabotropic receptors coupled to the rapid and coordinated activation of different classes of transcriptional factor genes.