Differential spatial patterns of Fos induction following generalized clonic and generalized tonic seizures

The Dorsal Bed Nucleus of the Stria Terminalis in Depressed and Non-Depressed Temporal Lobe Epilepsy Patients

OBJECTIVES

Temporal lobe epilepsy (TLE) and depression are common comorbid disorders whose underlying shared neural network has yet to be determined. While animal studies demonstrate a role for the dorsal bed nucleus of the stria terminalis (dBNST) in both seizures and depression, and humans clinical studies demonstrate a therapeutic effect of stimulating this region on treatment-resistant depression, the role for the dBNST in depressed and non-depressed TLE patients is still unclear. Here, we tested the hypothesis that this structure is morphologically abnormal in these epilepsy patients, with an increased abnormality in TLE patients with comorbid depression.

METHODS

In this case-controlled study, three Tesla structural magnetic resonance imaging scans were obtained from TLE patients with no depression (TLEonly), with depression (TLEdep) and healthy comparison subjects (HC). TLE subjects were recruited from the Yale University Comprehensive Epilepsy Center, diagnosed with the International League Against Epilepsy 2014 Diagnostic Guidelines, and confirmed by video electroencephalography. Diagnosis of major depressive disorder was confirmed by a trained neuropsychologist through a Mini International Neuropsychiatric Interview based on the DSM-IV. The dBNST was delineated manually by reliable raters using Bioimage Suite software.

RESULTS

The number of patients and subjects included 35 TLEonly patients, 20 TLEdep patients, and 102 HC subjects. Both TLEonly and TLEdep patients had higher dBNST volumes compared to HC subjects, unilaterally in the left hemisphere in the TLEonly patients (p=0.003), and bilaterally in the TLEdep patients (p<0·0001). Furthermore, the TLEdep patients had a higher dBNST volume than the TLEonly patients in the right hemisphere (p=0.02).

SIGNIFICANCE

Here we demonstrate an abnormality of the dBNST in TLE patients, both without depression (left enlargement) and with depression (bilateral enlargement). Our results demonstrate this region to underlie both temporal lobe epilepsy with and without depression, implicating it as a target to treat the comorbidity between these two disorders.

Amygdala-stimulation-induced apnea is attention and nasal-breathing dependent

OBJECTIVE

Evidence suggests that disordered breathing is critically involved in Sudden Unexpected Death in Epilepsy (SUDEP). To that end, evaluating structures that are activated by seizures and can activate brain regions that produce cardiorespiratory changes can further our understanding of the pathophysiology of SUDEP. Past preclinical studies have shown that electrical stimulation of the human amygdala induces apnea, suggesting a role for the amygdala in controlling respiration. In this study, we aimed to both confirm these findings in a larger group of patients with intractable temporal lobe epilepsy (TLE) and also further explore the anatomical and cognitive properties of this effect.

METHODS

Seven surgical TLE patients had depth electrodes implanted in the amygdala that were used to deliver electrical stimulation during functional mapping preceding resection. Real-time respiratory monitoring was performed in each patient to confirm apnea.

RESULTS

Our data confirm that amygdala stimulation reliably induces apnea (occurring in all 7 patients) and further suggest that apnea can be overcome by instructing the patient to inhale, and can be prevented entirely by breathing through the mouth before electrical stimulation. Finally, stimulation-induced apnea occurred only when stimulating the medial-most amygdalar contacts located in the central nucleus.

INTERPRETATION

These findings confirm a functional connection between the amygdala and respiratory control in humans. Moreover, they suggest specific amygdalar nuclei may be critical in mediating this effect and that attentional state is critical to apnea mediated by amygdala activation-perhaps alluding to future development of strategies for the prevention of SUDEP. Ann Neurol 2018;83:460-471.

The Repeated Flurothyl Seizure Model in Mice

Development of spontaneous seizures is the hallmark of human epilepsy. There is a critical need for new epilepsy models in order to elucidate mechanisms responsible for leading to the development of spontaneous seizures and for testing new anti-epileptic compounds. Moreover, rodent models of epilepsy have clearly demonstrated that there are two independent seizure systems in the brain: 1) the forebrain seizure network required for the expression of clonic seizures mediated by forebrain neurocircuitry, and 2) the brainstem seizure network necessary for the expression of brainstem or tonic seizures mediated by brainstem neurocircuitry. In seizure naïve animals, these two systems are separate, but developing models that can explore the intersection of the forebrain and brainstem seizure systems or for elucidating mechanisms responsible for bringing these two seizure systems together may aid in our understanding of: 1) how seizures can become more complex overtime, and 2) sudden unexpected death in epilepsy (SUDEP) since propagation of seizure discharge from the forebrain seizure system to the brainstem seizure system may have an important role in SUDEP because many cardiorespiratory systems are localized in the brainstem. The repeated flurothyl seizure model of epileptogenesis, as described here, may aid in providing insight into these important epilepsy issues in addition to understanding how spontaneous seizures develop.

Dissociation of spontaneous seizures and brainstem seizure thresholds in mice exposed to eight flurothyl-induced generalized seizures

Objective

C57BL/6J mice exposed to eight flurothyl‐induced generalized clonic seizures exhibit a change in seizure phenotype following a 28‐day incubation period and subsequent flurothyl rechallenge. Mice now develop a complex seizure semiology originating in the forebrain and propagating into the brainstem seizure network (a forebrain→brainstem seizure). In contrast, this phenotype change does not occur in seizure‐sensitive DBA/2J mice. The underlying mechanism was the focus of this study.

Methods

DBA/2J mice were exposed to eight flurothyl‐induced seizures (1/day) followed by 24‐h video‐electroencephalographic recordings for 28 days. Forebrain and brainstem seizure thresholds were determined in C57BL/6J and DBA/2J mice following one or eight flurothyl‐induced seizures, or after eight flurothyl‐induced seizures, a 28‐day incubation period, and final flurothyl rechallenge.

Results

Similar to C57BL/6J mice, DBA/2J mice expressed spontaneous seizures. However, unlike C57BL/6J mice, DBA/2J mice continued to have spontaneous seizures without remission. Because DBA/2J mice did not express forebrain→brainstem seizures following flurothyl rechallenge after a 28‐day incubation period, this indicated that spontaneous seizures were not sufficient for the evolution of forebrain→brainstem seizures. Therefore, we determined whether brainstem seizure thresholds were changing during this repeated‐flurothyl model and whether this could account for the expression of forebrain→brainstem seizures. Brainstem seizure thresholds were not different between C57BL/6J and DBA/2J mice on day 1 or on the last induction seizure trial (day 8). However, brainstem seizure thresholds did differ significantly on flurothyl rechallenge (day 28), with DBA/2J mice showing no lowering of their brainstem seizure thresholds.

Significance

These results demonstrate that DBA/2J mice exposed to the repeated‐flurothyl model develop spontaneous seizures without evidence of seizure remission and provide a new model of epileptogenesis. Moreover, these findings indicated that the transition of forebrain ictal discharge into the brainstem seizure network occurs as a result of changes in brainstem seizure thresholds that are independent of spontaneous seizure expression.

Blockade of T-type calcium channels prevents tonic-clonic seizures in a maximal electroshock seizure model

T-type (Cav3) calcium channels play important roles in neuronal excitability, both in normal and pathological activities of the brain. In particular, they contribute to hyper-excitability disorders such as epilepsy. Here we have characterized the anticonvulsant properties of TTA-A2, a selective T-type channel blocker, in mouse. Using the maximal electroshock seizure (MES) as a model of tonic-clonic generalized seizures, we report that mice treated with TTA-A2 (0.3 mg/kg and higher doses) were significantly protected against tonic seizures. Although no major change in Local Field Potential (LFP) pattern was observed during the MES seizure, analysis of the late post-ictal period revealed a significant increase in the delta frequency power in animals treated with TTA-A2. Similar results were obtained for Cav3.1-/- mice, which were less prone to develop tonic seizures in the MES test, but not for Cav3.2-/- mice. Analysis of extracellular signal-regulated kinase 1/2 (ERK) phosphorylation and c-Fos expression revealed a rapid and elevated neuronal activation in the hippocampus following MES clonic seizures, which was unchanged in TTA-A2 treated animals. Overall, our data indicate that TTA-A2 is a potent anticonvulsant and that the Cav3.1 isoform plays a prominent role in mediating TTA-A2 tonic seizure protection.

Spatiotemporal differences in the c-fos pathway between C57BL/6J and DBA/2J mice following flurothyl-induced seizures: A dissociation of hippocampal Fos from seizure activity

Significant differences in seizure characteristics between inbred mouse strains highlight the importance of genetic predisposition to epilepsy. Here, we examined the genetic differences between the seizure-resistant C57BL/6J (B6) mouse strain and the seizure-susceptible DBA/2J (D2) strain in the phospho-Erk and Fos pathways to examine seizure-induced neuronal activity to uncover potential mechanistic correlates to these disparate seizure responsivities. Expression of neural activity markers was examined following 1, 5, or 8 seizures, or after 8 seizures, a 28 day rest period, and a final flurothyl rechallenge. Two brain regions, the hippocampus and ventromedial nucleus of the hypothalamus (VMH), had significantly different Fos expression profiles following seizures. Fos expression was highly robust in B6 hippocampus following one seizure and remained elevated following multiple seizures. Conversely, there was an absence of Fos (and phospho-Erk) expression in D2 hippocampus following one generalized seizure that increased with multiple seizures. This lack of Fos expression occurred despite intracranial electroencephalographic recordings indicating that the D2 hippocampus propagated ictal discharge during the first flurothyl seizure suggesting a dissociation of seizure discharge from Fos and phospho-Erk expression. Global transcriptional analysis confirmed a dysregulation of the c-fos pathway in D2 mice following 1 seizure. Moreover, global analysis of RNA expression differences between B6 and D2 hippocampus revealed a unique pattern of transcripts that were co-regulated with Fos in D2 hippocampus following 1 seizure. These expression differences could, in part, account for D2's seizure susceptibility phenotype. Following 8 seizures, a 28 day rest period, and a final flurothyl rechallenge, ∼85% of B6 mice develop a more complex seizure phenotype consisting of a clonic-forebrain seizure that uninterruptedly progresses into a brainstem seizure. This seizure phenotype in B6 mice is highly correlated with bilateral Fos expression in the VMH and was not observed in D2 mice, which always express clonic-forebrain seizures upon flurothyl retest. Overall, these results illustrate specific differences in protein and RNA expression in different inbred strains following seizures that precede the reorganizational events that affect seizure susceptibility and changes in seizure semiology over time.

Segregation of seizure traits in C57 black mouse substrains using the repeated-flurothyl model

Identifying the genetic basis of epilepsy in humans is difficult due to its complexity, thereby underlying the need for preclinical models with specific aspects of seizure susceptibility that are tractable to genetic analyses. In the repeated-flurothyl model, mice are given 8 flurothyl-induced seizures, once per day (the induction phase), followed by a 28-day rest period (incubation phase) and final flurothyl challenge. This paradigm allows for the tracking of multiple phenotypes including: initial generalized seizure threshold, decreases in generalized seizure threshold with repeated flurothyl exposures, and changes in the complexity of seizures over time. Given the responses we previously reported in C57BL/6J mice, we analyzed substrains of the C57BL lineage to determine if any of these phenotypes segregated in these substrains. We found that the generalized seizure thresholds of C57BL/10SNJ and C57BL/10J mice were similar to C57BL/6J mice, whereas C57BL/6NJ and C57BLKS/J mice showed lower generalized seizure thresholds. In addition, C57BL/6J mice had the largest decreases in generalized seizure thresholds over the induction phase, while the other substrains were less pronounced. Notably, we observed only clonic seizures during the induction phase in all substrains, but when rechallenged with flurothyl after a 28-day incubation phase, ∼80% of C57BL/6J and 25% of C57BL/10SNJ and C57BL/10J mice expressed more complex seizures with tonic manifestations with none of the C57BL/6NJ and C57BLKS/J mice having complex seizures with tonic manifestations. These data indicate that while closely related, the C57BL lineage has significant diversity in aspects of epilepsy that are genetically controlled. Such differences further highlight the importance of genetic background in assessing the effects of targeted deletions of genes in preclinical epilepsy models.

N (w) -propyl-L-arginine (L-NPA) reduces status epilepticus and early epileptogenic events in a mouse model of epilepsy: behavioural, EEG and immunohistochemical analyses

We investigated the anticonvulsant and neurobiological effects of a highly selective neuronal nitric oxide synthase (nNOS) inhibitor, N (w) -propyl-l-arginine (L-NPA), on kainic acid (KA)-induced status epilepticus (SE) and early epileptogenesis in C57BL/6J mice. SE was induced with 20 mg/kg KA (i.p.) and seizures terminated after 2 h with diazepam (10 mg/kg, i.p). L-NPA (20 mg/kg, i.p.) or vehicle was administered 30 min before KA. Behavioural seizure severity was scored using a modified Racine score and electrographic seizure was recorded using an implantable telemetry device. Neuronal activity, activity-dependent synaptogenesis and reactive gliosis were quantified immunohistochemically, using c-Fos, synaptophysin and microglial and astrocytic markers. L-NPA treatment reduced the severity and duration of convulsive motor seizures, the power of electroencephalogram in the gamma band, and the frequency of epileptiform spikes during SE. It also reduced c-Fos expression in dentate granule cells at 2 h post-KA, and reduced the overall rate of epileptiform spiking (by 2- to 2.5-fold) in the first 7 days after KA administration. Furthermore, treatment with L-NPA suppressed both hippocampal gliosis and activity-dependent synaptogenesis in the outer and middle molecular layers of the dentate gyrus in the early phase of epileptogenesis (72 h post-KA). These results suggest that nNOS facilitates seizure generation during SE and may be important for the neurobiological changes associated with the development of chronic epilepsy, especially in the early stages of epileptogenesis. As such, it might represent a novel target for disease modification in epilepsy.

Analysis of flurothyl-induced myoclonus in inbred strains of mice

Myoclonus is often observed in epilepsy. It is characterized by sudden involuntary shock-like movements of the body (myoclonic jerks, MJs). This study examined whether epileptic myoclonus was under genetic control. Inbred strains of mice were administered eight daily flurothyl exposures, a 28-day rest period, and a final flurothyl retest. For all trials, the latency to the first MJ (threshold) and the number of MJs (MJ#) were recorded. The inbred strains that we examined exhibited significant variability in initial myoclonic response, and myoclonus across the eight flurothyl exposures. C57BL/6J and DBA/2J mice displayed significantly different initial latencies to a MJ, MJ# preceding a generalized seizure (GS), and changes in MJ threshold and MJ# across the eight seizure trials. [C57BL/6J x DBA/2J] F1-hybrid mice showed an initial MJ threshold and decreases in MJ threshold over the eight trials, which were similar to C57BL/6J; however, F1-hybrids had an initial MJ# and trend in MJ# over the eight trials that were similar to DBA/2J. Decreases in MJ threshold and MJ# following multiple seizure trials, observed in C57BL/6J mice, were dependent on the expression of GSs and not on MJ occurrence. Our study is the first to document the potential for genetic heterogeneity of myoclonus in mice; we show that significant alterations in myoclonic behavior occur after GSs. These results indicate that multiple GSs affect MJ thresholds. An understanding of the genetics of myoclonus will be important for determination of the brain areas responsible for myoclonus as well as for identification of candidate genes.

Dissociation of seizure traits in inbred strains of mice using the flurothyl kindling model of epileptogenesis

Previous seizure models have demonstrated genetic differences in generalized seizure threshold (GST) in inbred mice, but the genetic control of epileptogenesis is relatively unexplored. The present study examined, through analysis of inbred strains of mice, whether the seizure characteristics observed in the flurothyl kindling model are under genetic control. Eight consecutive, daily generalized seizures were induced by flurothyl in mice from five inbred strains. Following a 28-day rest period, mice were retested with flurothyl. The five strains of mice demonstrated inter-strain differences in GST, decreases in GST across seizure trials, and differences in the behavioral seizure phenotypes expressed. Since many of the seizure characteristics that we examined in the flurothyl kindling model were dissociable between C57BL/6J and DBA/2J mice, we analyzed these strains in detail. Unlike C57BL/6J mice, DBA/2J mice had a lower GST on trial 1, did not demonstrate a decrease in GST across trials, nor did they show an alteration in seizure phenotype upon flurothyl retest. Surprisingly, [C57BL/6JxDBA/2J] F1-hybrids had initial GST on trial 1 and GST decreases across trials similar to what was found for C57BL/6J, but they did not undergo the alteration in behavioral seizure phenotype that had been observed for C57BL/6J mice. Our data establish the significance of the genetic background in flurothyl-induced epileptogenesis. The [C57BL/6JxDBA/2J] F1-hybrid data demonstrate that initial GST, the decrease in GST across trials, and the change in seizure phenotype differ from the characteristics of the parental strains, suggesting that these phenotypes are controlled by independent genetic loci.

Metabolic environment in substantia nigra reticulata is critical for the expression and control of hypoglycemia-induced seizures

Seizures represent a common and serious complication of hypoglycemia. Here we studied mechanisms of control of hypoglycemic seizures induced by insulin injection in fasted and nonfasted rats. We demonstrate that fasting predisposes rats to more rapid and consistent development of hypoglycemic seizures. However, the fasting-induced decrease in baseline blood glucose concentration cannot account for the earlier onset of seizures in fasted versus nonfasted rats. Data obtained with c-Fos immunohistochemistry and [14C]2-deoxyglucose uptake implicate a prominent involvement of the substantia nigra reticulata (SNR) among other structures in the hypoglycemic seizure control. This is supported by data showing that fasting decreases the SNR expression of K(ATP) channels, which link metabolism with activity, and is further confirmed with microinfusions of K(ATP) channel agonist and antagonist. Data obtained with whole-cell and perforated patch recordings from SNR neurons in slices in vitro demonstrate that both presynaptic and postsynaptic K(ATP) channels participate in the failure of the SNR to control hypoglycemic seizures. The results suggest that fasting and insulin-induced hypoglycemia can lead to impairment in the function of the SNR, leading thus to hypoglycemic seizures.

Characterization of the neurotoxicity induced by the extract of Magnistipula butayei (Chrysobalanaceae) in rat: Effects of a new natural convulsive agent

This study was designed to document convulsant and neurotoxic properties of extracts of a tropical tree, Magnistipula butayei subsp. Montana, and to investigate the involvement of the glutamatergic system in these effects. Continuous behavioral observations and electroencephalographic (EEG) records were obtained after per os administration of an aqueous extract of Magnistipula (MBMAE) in rats. MBMAE (800 mg/kg) induced behavioral changes resembling motor limbic seizures: staring and head tremor, automatisms, forelimb clonic movements and violent tonic-clonic seizures leading to death in all animals. Concomitantly, important seizure activity that gradually evolved to epileptiform activity was recorded on the EEG. Moreover, c-Fos immunohistochemistry has revealed an increased c-Fos expression in the dentate gyrus and in piriform, peri- and entorhinal cortices 2 and 4h after treatment. This expression pattern suggested that the mechanism of action for the MBMAE is similar to that observed in glutamate-induced models of epilepsy. The MBMAE increased cell death also in hippocampal cell cultures. Furthermore, the build-up of convulsive activity and epileptic discharges induced by MBMAE in rat were abolished by MK-801, an NMDA receptor antagonist. Our study suggests that MBMAE contains a potent toxin, with a powerful neurotoxic activity in rat, and corresponding to a new natural component(s) that act as an NMDA-mediated convulsant molecule.

Model of infantile spasms induced by N-methyl-D-aspartic acid in prenatally impaired brain

OBJECTIVE

Infantile spasms (a catastrophic epileptic syndrome of childhood) are insensitive to classic antiepileptic drugs. New therapies are limited by lack of animal models. Here we develop a new model of flexion spasms based on prenatal exposure to betamethasone combined with postnatal administration of N-methyl-D-aspartic acid (NMDA) and determine brain structures involved in the induction of flexion spasms.

METHODS

Pregnant rats received two doses of betamethasone on day 15 of gestation. Offspring was injected with NMDA on postnatal day 15. Effects of adrenocorticotropin therapy on the development of age-specific flexion spasms were determined and electroencephalographic correlates recorded. C-fos immunohistochemistry and [14C]2-deoxyglucose imaging identified brain structures involved in the development of flexion spasms.

RESULTS

Prenatal betamethasone exposure sensitizes rats to development of NMDA-induced spasms and, most importantly, renders the spasms sensitive to adrenocorticotropin therapy. Ictal electroencephalogram results correspond to human infantile spasms: electrodecrement or afterdischarges were observed. Imaging studies defined three principal regions involved in NMDA spasms: limbic areas (except the dorsal hippocampus), hypothalamus, and the brainstem.

INTERPRETATION

Despite certain limitations, our new model correlates well with current infantile spasm hypotheses and opens an opportunity for development and testing of new effective drugs.

Pharmacological and genetic manipulation of kappa opioid receptors: Effects on cocaine- and pentylenetetrazol-induced convulsions and seizure kindling

The present study used pharmacological and gene ablation techniques to examine the involvement of kappa opioid receptors (KOPr) in modulating the convulsant effects of two mechanistically different drugs: cocaine and pentylenetetrazol (PTZ; GABA-A receptor antagonist) in mice. Systemic administration of the selective KOPr-1 agonist, U69593 (0.16-0.6mg/kg; s.c.), failed to modify cocaine-evoked convulsions or cocaine kindling. Similarly, no alteration in responsiveness to cocaine was observed in wild-type mice that received the selective KOPr-1 antagonist, nor-binaltorphimine (nor-BNI; 5mg/kg) or in mice lacking the gene encoding KOPr-1. In contrast to cocaine, U69593 attenuated the seizures induced by acute or repeated PTZ administration. Nor-BNI decreased the threshold for PTZ-evoked seizures and increased seizure incidence during the initial induction of kindling relative to controls. Decreased thresholds for PTZ-induced seizures were also observed in KOPr-1 knock out mice. Together, these data demonstrate an involvement of endogenous KOPr systems in modulating vulnerability to the convulsant effects of PTZ but not cocaine. Furthermore, they demonstrate that KOPr-1 activation protects against acute and kindled seizures induced by this convulsant. Finally, the results of our study suggest that KOPr-1 antagonists will not have therapeutic utility against cocaine-induced seizures, while they may prove beneficial in attenuating several actions of cocaine that have been linked to its abuse.

Region specific hypothalamic neuronal activation and endothelial cell proliferation in response to electroconvulsive seizures

BACKGROUND

Major depression is often associated with disturbances in basal biological functions regulated by the hypothalamus. Electroconvulsive therapy (ECT), an efficient anti-depressant treatment, alters the activity of hypothalamic neurons. We have previously shown an increased proliferation of endothelial cells in specific areas of the rat hippocampus in response to electroconvulsive seizure (ECS) treatment, an animal model for ECT. Here we examine the effect of ECS treatment on neuronal activation and endothelial cell proliferation in mid-hypothalamus.

METHODS

Rats received one daily ECS treatment for 5 days and cell proliferation was detected by bromodeoxyuridine (BrdU). The number of cells double-labeled for BrdU and the endothelial cell marker rat endothelial cell antigen-1 was determined. Neuronal activation in response to acute ECS treatment was detected as c-Fos immunoreactivity in an additional experiment.

RESULTS

We demonstrate a correlating pattern of increases in neuronal activation and increased endothelial cell proliferation in the paraventricular nucleus, the supraoptic nucleus, and the ventromedial nucleus of the hypothalamus after ECS treatment.

CONCLUSIONS

Hypothalamic areas with the largest increase in neuronal activation after ECS treatment exhibit increased endothelial cell proliferation. We suggest that similar angiogenic responses to ECT might counteract hypothalamic dysfunction in depressive disorder.

c-Fos immunohistochemical mapping of the audiogenic seizure network and tonotopic neuronal hyperexcitability in the inferior colliculus of the Frings mouse

The Frings mouse is a model of audiogenic seizure (AGS) susceptibility. The genetic locus responsible for the AGS phenotype in the Frings mouse has been named monogenic audiogenic seizure-susceptible (MASS1). MASS1 is unique in that it is one of only two identified seizure loci that are not associated with an ion channel mutation. Furthermore, Frings mice display a robust AGS phenotype demonstrating very high and prolonged susceptibility to sound-induced tonic extension seizures. The purpose of this investigation was to use c-Fos immunohistochemistry to map the brain structures involved in the Frings AGS and to examine neuronal hyperexcitability in the inferior colliculus, the brain structure that is recognized as the site of AGS initiation. AGS mapping revealed that intense seizure-induced neuronal activation was mostly limited to structures involved in a brainstem seizure network, including the external and dorsal nuclei of the inferior colliculus, as observed in other AGS rodents. Acoustically induced c-Fos expression in the central nucleus of the inferior colliculus to sub-AGS threshold tone stimulations displayed a greater level of neuronal activation in AGS-susceptible Frings, DBA/2J and noise-primed C57BL/6J mice compared to AGS-resistant C57BL/6J and CF1 mice. The AGS-susceptible mice also displayed c-Fos immunoreactivity that was more focused within the tonotopic response domain of the inferior colliculus compared to AGS-resistant mice. Furthermore, Frings mice displayed significantly greater tonotopic hyper-responsiveness compared to other AGS-susceptible mice.

Comparative fos immunoreactivity in the brain after forebrain, brainstem, or combined seizures induced by electroshock, pentylenetetrazol, focally induced and audiogenic seizures in rats

To help discern sites of focal activation during seizures of different phenotype, the numbers of Fos immunoreactive (FI) neurons in specific brain regions were analyzed following "brainstem-evoked," "forebrain-evoked" and forebrain/brainstem combination seizures induced by a variety of methods. First, pentylenetetrazol (PTZ, 50 mg/kg) induced forebrain-type seizures in some rats, or forebrain seizures that progressed to tonic/clonic brainstem-type seizures in other rats. Second, minimal electroshock induced forebrain seizures whereas maximal electroshock (MES) induced tonic brainstem-type seizures in rats. Third, forebrain seizures were induced in genetically epilepsy-prone rats (GEPRs) by microinfusion of bicuculline into the area tempestas (AT), while brainstem seizures in GEPRs were induced by audiogenic stimulation. A final set was included in which AT bicuculline-induced forebrain seizures in GEPRs were transiently interrupted by audiogenic seizures (AGS) in the same animals. These animals exhibited a sequence combination of forebrain clonic seizure, brainstem tonic seizure and back to forebrain clonic seizures. Irrespective of the methods of induction, clonic forebrain- and tonic/clonic brainstem-type seizures were associated with considerable Fos immunoreactivity in several forebrain structures. Tonic/clonic brainstem seizures, irrespective of the methods of induction, were also associated with FI in consistent brainstem regions. Thus, based on Fos numerical densities (FND, numbers of Fos-stained profiles), forebrain structures appear to be highly activated during both forebrain and brainstem seizures; however, facial and forelimb clonus characteristic of forebrain seizures are not observable during a brainstem seizure. This observation suggests that forebrain-seizure behaviors may be behaviorally masked during the more severe tonic brainstem seizures induced either by MES, PTZ or AGS in GEPRs. This suggestion was corroborated using the sequential seizure paradigm. Similar to findings using MES and PTZ, forebrain regions activated by AT bicuculline were similar to those activated by AGS in the GEPR. However, in the combination seizure group, those areas that showed increased FND in the forebrain showed even greater FND in the combination trial. Likewise, those areas of the brainstem showing FI in the AGS model, showed an even greater effect in the combination paradigm. Finally, the medial amygdala, ventral hypothalamus and cortices of the inferior colliculi showed markedly increased FND that appeared dependent upon activation of both forebrain and brainstem seizure activity in the same animal. These findings suggest these latter areas may be transitional areas between forebrain and brainstem seizure interactions. Collectively, these data illustrate a generally consistent pattern of forebrain Fos staining associated with forebrain-type seizures and a consistent pattern of brainstem Fos staining associated with brainstem-type seizures. Additionally, these data are consistent with a notion that separate seizure circuitries in the forebrain and brainstem mutually interact to facilitate one another, possibly through involvement of specific "transition mediating" nuclei.

Persistent regional increases in brain-derived neurotrophic factor in the flurothyl model of epileptogenesis are dependent upon the kindling status of the animal

Brain-derived neurotrophic factor (BDNF) appears to be both regulated by and a regulator of epileptogenesis. In the flurothyl (HFE) model of kindling mice exposed to successive flurothyl trials over 8 days express a rapid, long-lasting reduction in generalized seizure threshold and a more slowly evolving change in seizure phenotype in response to subsequent flurothyl exposure. The BDNF genotype of particular mouse strains appears to influence the epileptogenic progression in this model. Thus, we hypothesized that BDNF signaling pathways are altered by flurothyl-induced seizures. Following HFE kindling, fully kindled (eight seizures) adult male C57BI/6J mice had significantly elevated whole brain BDNF levels through at least 28 days after their final seizure. Mice that received only four HFE seizures (not kindled) had elevated BDNF levels, but only at 1 day post-seizure (DPSz), while BDNF levels were not significantly altered in mice receiving just one HFE seizure at any time point studied. Regional expression patterns of BDNF in the hippocampus, hypothalamus, and frontal cortex were also elevated by one DPSz and returned to control values by 14 DPSz in mice that received four HFE seizures. No changes were seen in the cerebellum, striatum, or piriform cortex. In contrast, fully kindled mice had significantly elevated BDNF levels within the hippocampus, hypothalamus, neocortex, and striatum that remained elevated through at least 14 DPSz, while levels were unchanged in the cerebellum and piriform cortex. Regional results were confirmed using anti-BDNF immunohistochemistry (IHC). Despite changes in BDNF levels following HFE kindling, we were unable to demonstrate alterations either in full-length tyrosine kinase receptor B (TrkB) expression (Western blot and IHC) or in truncated TrkB (IHC) expression levels. Together, these data suggest a model of a positive feedback loop involving seizure activity and seizure number and persistently modified BDNF signaling pathways that influences seizure phenotypes within the HFE kindling paradigm. Thus, long-term elevations in BDNF may be responsible in part for epileptogenic processes and the development of human refractory epilepsies.

Suppression of absence seizures by electrical and pharmacological activation of the caudal superior colliculus in a genetic model of absence epilepsy in the rat

Activation of the superior colliculus has been shown to reproduce the antiepileptic effect of the inhibition of the substantia nigra reticulata. A circuit involving neurons of the caudal deep layers of the superior colliculus has been suggested to control brain stem convulsive seizures. The present study was designed to examine whether a similar circuit is also involved in the control of absence seizures. For this, activation of either the rostral or caudal parts of the deep and intermediate layers of the superior colliculus was applied in a genetic model of absence seizures in the rat (GAERS). Single-shock (5 s) electrical stimulation of the rostral and caudal superior colliculus interrupted ongoing spike-and-wave discharges at an intensity (antiepileptic threshold) significantly lower than the intensity inducing behavioral effects. At this intensity, no interruption of licking behavior was observed in water-deprived rats. Repeated stimulations (5 s on/5 s off) at the antiepileptic threshold reduced absence seizures only during the first 10 min. Bilateral microinjection of a GABA antagonist (picrotoxin, 33 pmol/side) significantly suppressed spike-and-wave discharges when applied in the caudal aspect of the superior colliculus. This antiepileptic effect appears dissociated from an anxiogenic effect, as tested in an elevated plus maze test. Finally, bilateral injection of picrotoxin (33 pmol/side) appeared more effective in the superficial and intermediate layers of the caudal superior colliculus, whereas such injections had only weak effects on absence seizures when applied in the deep layers. These results suggest that a specific population of neurons located in the intermediate and superficial layers of the caudal superior colliculus is involved in the inhibitory control of absence seizures. It may constitute an important relay for the control of absence seizures by the basal ganglia via the substantia nigra reticulata.

Repeated generalized audiogenic seizures induce plastic changes on acoustically evoked neuronal firing in the amygdala

Repetition of audiogenic seizures (AGS) (AGS kindling) results in increases in the duration of convulsive behavior and the emergence of cortical epileptiform EEG activity. These changes involve expansion of the neuronal network subserving these seizures. The amygdala (AMG) is postulated to become involved in this expanded network, but the neurophysiological basis of this process is unknown. The present study examined changes in chronically-recorded extracellular neuronal firing patterns in the lateral nucleus of AMG (LAMG) induced by AGS kindling in behaving genetically epilepsy-prone rats (GEPR-9s). Before AGS kindling, onset-only (36.1%), onset-delayed (50%) and delayed-only (13.9%) patterns of response to acoustic stimuli were observed. Neuronal firing was greatly suppressed following systemically administered uncompetitive NMDA antagonist (ketamine, 30 mg/kg, i.p.) with complete recovery by 4 h. After AGS kindling, LAMG neurons displayed a significantly increased incidence of onset-only patterns (93.3%, at 0.5 Hz), and mean acoustic responsiveness was also significantly increased (516.2% of control). LAMG neurons fired tonically during tonic convulsions and exhibited burst firing during post-tonic clonus. Greater acoustically-induced synchronization of LAMG firing, as indicated by elevated responsiveness and increased concentration of firing near the stimulus onset, may be critical for mediating the behavioral and EEG changes induced by AGS kindling. LAMG neuronal firing increases induced by AGS kindling may initiate these pathophysiological alterations, in part, by enhanced glutamate receptor-mediated excitation. This possibility is supported by the previously observed ability of an NMDA antagonist to reverse AGS kindling when focally microinjected into AMG, and the blockade of LAMG firing by administration of an uncompetitive NMDA antagonist observed in the present study.

Pharmacological characterization of the 6 Hz psychomotor seizure model of partial epilepsy

Originally described as a model of 'psychomotor seizures' (J. Pharmacol. Exp. Ther. (1953) 107-273), the 6 Hz corneal stimulation model was abandoned shortly after its description because of its lack of sensitivity to phenytoin. This observation is the basis for the present study designed to validate the 6 Hz seizure as a model of therapy-resistant epilepsy. The pharmacological profile of the 6 Hz seizure was determined at varying current intensities using seven established AEDs (phenytoin, carbamazepine, clonazepam, phenobarbital, ethosuximide, trimethadione, valproic acid) and five second-generation AEDs (lamotrigine, levetiracetam, felbamate, tiagabine, topiramate). The immediate early gene c-Fos was used as a marker of seizure-induced neuronal activation to help define those brain structures that were activated by 6 Hz corneal stimulation. At the current intensity required to produce a seizure in 97% of the population (CC97=22 mA), the 6 Hz seizure did not discriminate between clinical classes of AEDs tested. Increasing the current intensity by 50% (i.e. 32 mA) decreased the sensitivity of the 6 Hz seizure to phenytoin and lamotrigine. At a current intensity of 2 x CC97 (i.e. 44 mA), only two AEDs, levetiracetam and valproic acid, displayed complete protection against the 6 Hz seizure, though the efficacy of these drugs was reduced when compared to the lower stimulation intensities. Intense c-Fos staining from 6 Hz seizures induced by 22 and 32 mA stimulus intensities remained localized to the amygdala and piriform cortex. Increasing the stimulus intensity to 44 mA resulted in additional heavy staining of the dentate gyrus. This recruitment of the dentate gyrus may account for the decrease in potency of levetiracetam and valproic acid at 44 mA. The pharmacological results combined with the c-Fos immunohistochemistry suggest that the 6 Hz stimulation may provide a useful model of therapy-resistant limbic seizures.

Behavioral and metabolic features of repetitive seizures in immature and mature rats

Seizure incidence varies significantly with age, with seizure susceptibility particularly high during the first few years of life. Of significant concern is what effects do brief, repetitive seizures have on the developing brain. We approached this issue by examining the change in seizure threshold, and related markers of neuronal activity and metabolic activity (c-fos mRNA and 2-deoxyglucose [2DG]), as a function of repetitive seizure episodes in immature and mature rats. Starting on postnatal day 15 (P15) (immature) or P60 (adult) rats were given two flurothyl seizures a day for 5 days (nine or ten seizures). The seizure latency profile, our measure of threshold, in immature versus adult rats across the 5-day testing period was different. In immature rats, threshold for the second seizure on each day was significantly lower than for the first seizure, suggesting that there was little refractoriness after the first seizure of the day. In contrast, the mature animal had a significantly longer threshold latency to the second seizure for the first 3 days of testing. The immature animal was also more likely than the adult to exhibit tonic extension as a feature of the first seizure of the day. Following repetitive seizures, more regions of the CNS showed c-fos mRNA expression in the immature animal than adults, suggesting that repetitive seizures in the immature animal activated a greater percentage of the brain. Compared with the effects of a single seizure, repetitive seizures resulted in less 2DG labeling in most regions of the brain (except the hippocampus); in the immature brain this difference was more distinct than in adults. The consequences of repetitive seizures in the immature animal results in distinctly different seizure behavior and neuronal activity pattern (c-fos expression) than that observed in the mature animal.

Spatiotemporal analysis of Fos expression associated with cocaine- and PTZ-induced seizures in prenatally cocaine-treated rats

We previously reported that prenatal cocaine exposure (40 mg/kg s.c., E10-E20) increased susceptibility to convulsant-induced seizures later in life, with female rats becoming more sensitive to seizures induced by cocaine and pentylenetetrazol (PTZ), and males more sensitive to PTZ-induced seizures (Snyder-Keller and Keller, 1995, 2000). In order to determine the locus of enhanced seizure susceptibility in the brains of prenatally cocaine-treated rats, we examined the distribution and density of Fos-immunoreactive cells after cocaine- and PTZ-induced seizures in mature rats. Subconvulsive cocaine doses induced c-fos in cortical areas as well as densely dopamine-innervated regions such as striatum and nucleus accumbens. Following cocaine-induced seizures, intense c-fos induction was observed in piriform cortex, amygdala, and hippocampus. Quantification of the number of Fos-immunoreactive cells in the brains of prenatally cocaine-treated versus prenatally saline-treated rats revealed differences in piriform cortex and amygdala that were indicative of a lower threshold in prenatally cocaine-treated female rats. Following PTZ-induced seizures, the same pattern of limbic structures were recruited with increasing seizure severity. Only females exhibited changes in the number of Fos-immunoreactive cells as a result of prenatal cocaine treatment. Pretreatment with the noncompetitive NMDA antagonist MK-801 blocked both cocaine- and PTZ-induced seizures, and Fos expression in limbic areas was also blocked. The dopamine D1 antagonist SCH 23390 blocked cocaine-induced seizures and associated c-fos induction, but not PTZ-induced seizures or Fos. Examination of the pattern of Fos expression at 15-20 min postseizure revealed that the initial site of c-fos induction associated with PTZ-induced seizures appeared to be the piriform cortex, whereas cocaine-induced seizures induced early expression in both piriform cortex and lateral amygdala. These findings suggest that neural alterations residing in the piriform cortex and amygdala are likely to account for the increased seizure susceptibility of prenatally cocaine-treated rats.

c-Fos expression in rat brain after repetitive transcranial magnetic stimulation

We investigated neuronal response to repetitive transcranial magnetic stimulation (rTMS) and electroconvulsive shock (ECS) in terms of c-Fos expression. In rats at postnatal day 49, six rTMS sessions induced widespread nuclear c-Fos-like immunoreactivity in frontal cortex, lateral orbital cortex, striatum, lateral septal nucleus, piriform cortex, dentate gyrus, Ammon's horn, cingulate cortex, parietal cortex, thalamus, occipital cortex, and amygdala; this reactivity was greater than with two sessions of rTMS or sham rTMS. ECS produced even stronger c-Fos expression than six sessions of rTMS in all regions except thalamus (no difference) and striatum (stronger with rTMS). Thus, functional modification of neuroanatomic substrates as demonstrated by c-Fos expression may partially differ between rTMS and ECS.

Focal microinjection of carbachol into the periaqueductal gray induces seizures in the forebrain of the rat

Previous studies have reported that the repetition of running-bouncing and tonic-clonic seizures mediated by brainstem structures eventually elicits seizure activity in the forebrain. The purpose of the present study was to determine if the periaqueductal gray (PAG) region is a component of the neural network through which brainstem seizures elicit forebrain seizures. Bilateral microinjection of 40 nmol carbachol into the PAG region of rats induced arrested, staring behavior accompanied by epileptiform electrocorticogram (ECoG) afterdischarge recorded from the parietal cortex. In two animals limbic seizure activity similar to kindled amygdala seizures was also induced. The carbachol effect was dose-related as the 40 nmol dose induced a significantly greater duration of ECoG afterdischarge than a 20 nmol dose. The carbachol effect was mediated by muscarinic receptors as bilateral 50 nmol atropine microinjection 1 min prior to 40 nmol carbachol microinjection inhibited all seizure activity. Immunohistochemical detection of the proto-oncogene c-fos was used to verify that seizure activity was induced in forebrain regions. Rats with seizures induced by PAG carbachol microinjections exhibited dense c-fos-like immunoreactivity in the dentate gyrus but not the CA(1) or CA(3) regions, amygdala, piriform cortex, perirhinal cortex or hypothalamus. In addition, PAG microinjection of 10 nmol N-methyl-D-aspartic acid (NMDA) induced wild-running convulsions while 400 pmol bicuculline induced clonic spasms, myoclonic activity or limbic seizures. These results indicate that stimulation of the PAG, a brainstem structure, is sufficient to induce forebrain seizures. Since the forebrain seizures were induced by a single carbachol administration, it is proposed that the PAG serves as a pathway for caudal-rostral seizure generalization.

Electroshocks delay seizures and subsequent epileptogenesis but do not prevent neuronal damage in the lithium-pilocarpine model of epilepsy

Electroconvulsive therapy, which is used to treat refractory major depression in humans increases seizure threshold and decreases seizure duration. Moreover, the expression of brain derived neurotrophic factor induced by electroshocks (ECS) might protect hippocampal cells from death in patients suffering from depression. As temporal lobe epilepsy is linked to neuronal damage in the hippocampus, we tested the effect of repeated ECS on subsequent status epilepticus (SE) induced by lithium-pilocarpine and leading to cell death and temporal epilepsy in the rat. Eleven maximal ECS were applied via ear-clips to adult rats. The last one was applied 2 days before the induction of SE by lithium-pilocarpine. The rats were electroencephalographically recorded to study the SE characteristics. The rats treated with ECS before pilocarpine (ECS-pilo) developed partial limbic (score 2) and propagated seizures (score 5) with a longer latency than the rats that underwent SE alone (sham-pilo). Despite this delay in the initiation and propagation of the seizures, the same number of ECS- and sham-pilo rats developed SE with a similar characteristic pattern. The expression of c-Fos protein was down-regulated by repeated ECS in the amygdala and the cortex. In ECS-pilo rats, c-Fos expression was decreased in the piriform and entorhinal cortex and increased in the hilus of the dentate gyrus. Neuronal damage was identical in the forebrain areas of both groups, while it was worsened by ECS treatment in the substantia nigra pars reticulata, entorhinal and perirhinal cortices compared to sham-pilo rats. Finally, while 11 out of the 12 sham-pilo rats developed spontaneous recurrent seizures after a silent period of 40+/-27 days, only two out of the 10 ECS-pilo rats became epileptic, but after a prolonged latency of 106 and 151 days. One ECS-pilo rat developed electrographic infraclinical seizures and seven did not exhibit any seizures. Thus, the extensive neuronal damage occurring in the entorhinal and perirhinal cortices of the ECS-pilo rats seems to prevent the establishment of the hyperexcitable epileptic circuit.

Fos expression and 2-deoxyglucose uptake following seizures in developing genetically epilepsy-prone rats

Juvenile genetically epilepsy-prone rats (GEPR)-3s display one of three types of seizures in response to sound: a typical class 3 seizure consisting of an explosive running/bouncing episode followed by a clonic seizure (audiogenic response score, ARS-3); an ARS-3 seizure followed by a forebrain seizure that includes facial and forelimb (F&F) clonus with rearing (ARS-3f); or, a running/bouncing episode followed by a severe tonic seizure with complete hindlimb extension (ARS-9) not accompanied with subsequent F&F clonus. The adult seizure phenotype, manifest in all GEPR-3s by age 45 days of age, consists of an ARS-3 not followed by F&F clonus or tonic extension. The present studies sought to determine the neuronal networks activated during these various developmental convulsive patterns by examining anatomical patterns of [(14)C]2-deoxyglucose (2-DG) uptake or immediate-early-gene (Fos) expression subsequent to seizures. Many, but not all, brain areas of control rats showed age-related increases in Fos expression in response to the acoustic stimulation. An age effect was not observed in 2-DG uptake. In GEPRs, the profiles of Fos expression and 2-DG uptake following seizures were often parallel; however, there were notable exceptions. For example, increased 2-DG uptake in the cochlear nuclei, central region of the inferior colliculi, and the substantia nigra were not accompanied by increased Fos expression in these areas regardless of the seizure phenotypes. Reciprocally, other regions, particularly in the amygdala, ventromedial hypothalamus and parabrachial areas, displayed intense seizure related Fos labeling without detectable increases in 2-DG uptake. Fos and 2-DG uptake patterns in response to acoustic stimulation varied according to brain region, seizure phenotype and severity. In general, the degree of 2-DG uptake correlated with seizure severity. For example, the ARS-9 seizures, being the most intense, resulted in significant increases in 2-DG uptake in almost all brain regions examined. 2-DG uptake following the ARS-3f and ARS-3 seizures, although increased, did not reach statistical significance in most brain areas. In contrast to the 2-DG findings, a seizure-severity dependent effect was not seen with Fos. Rather, the induction of Fos associated with acoustic stimulation and seizure was more associated with age and seizure-phenotype. Thus, the developmental profiles of Fos expression and 2-DG uptake in response to seizures are distinctly different and concurrent examination of both markers is useful in the identification of brain circuitry involved in seizure development.

Molecular serotonergic mechanisms appear to mediate genetic sensitivity to cocaine-induced convulsions

Cocaine-induced convulsions appear to be mediated by serotonin (5-HT) neurotransmission, acting primarily at 5-HT(2) receptors. However, this effect of cocaine is attenuated by cocaine binding at sigma and muscarinic M(1) and M(2) sites. This study examined whether the aforementioned neural sites mediate the nearly two-fold difference in sensitivity to cocaine-induced convulsions across C57BL/6J (6J) and C57BL/6ByJ (6ByJ) mice. Experiment 1 compared 5-HT transporter densities across several brain regions of 6J and 6ByJ mice and cocaine-induced convulsions following pretreatment with the 5-HT reuptake inhibitor fluoxetine. Experiment 2 compared 5-HT(2) receptor densities across these mice and cocaine-induced convulsions following pretreatment with the 5-HT(2) antagonist cinanserin. There were no differences in 5-HT transporter densities, however, fluoxetine produced a greater facilitation of cocaine-induced convulsions in 6ByJ relative to 6J mice, suggesting that sensitivity to convulsions is mediated postsynaptically. Indeed, 5-HT(2) density was higher in 6ByJ relative to 6J mice in the amygdaloid ridge, hypothalamus, and midbrain. In addition, cinanserin attenuated convulsions more potently in 6J relative to 6ByJ mice. There were no differences in the densities or affinities of 5-HT(1), muscarinic, or sigma receptors across these strains, suggesting that density of these latter sites does not mediate genetic sensitivity to cocaine-induced convulsions. Since 6ByJ mice are less sensitive to convulsions despite the fact that they have more 5-HT(2) receptors, we hypothesized that these mice may exhibit a weaker linkage of 5-HT(2) sites to their second-messenger system relative to 6J mice. However, in experiment 3 we demonstrated that 5-HT(2)-receptor mediated phosphoinositide hydrolysis was higher in 6ByJ relative to 6J mice in the same regions also displaying higher 5-HT(2) densities. This study suggests that 5-HT(2) receptors mediate genetic sensitivity to cocaine-induced convulsions, further supporting the role of these sites in mediating this toxic effect of cocaine.

Selective susceptibility to inhibitors of GABA synthesis and antagonists of GABA(A) receptor in rats with genetic absence epilepsy

Thalamocortical spike-and-wave discharges characterize the nonconvulsive absence seizures that occur spontaneously in genetic absence epilepsy rats from Strasbourg (GAERS), a selected strain of Wistar rats. GABA is crucial in the generation of absence seizures. The susceptibility to convulsions induced by threshold doses of various GABA receptor antagonists and inhibitors of GABA synthesis, kainic acid and strychnine, was compared in GAERS and in nonepileptic rats from a selected control strain (NE). The brain structures involved in the drug-elicited convulsive seizures were mapped by c-Fos immunohistochemistry. Injection of various antagonists of the GABA(A) receptor, bicuculline and picrotoxin, and inverse agonists of the benzodiazepine site (FG 7142 and DMCM) induced myoclonic spike-and-wave discharges followed by clonic or tonic-clonic seizures with high paroxysmal activity on the cortical EEG. The incidence of the convulsions was dose-dependent and was higher in GAERS than in NE rats. Mapping of c-Fos expression showed that the frontoparietal cortex was constantly involved in the convulsive seizures elicited by a threshold convulsant dose, whereas limbic participation was variable. In contrast, GAERS were less susceptible than NE rats to the tonic-clonic convulsions induced by the inhibitors of glutamate decarboxylase, isoniazide and 3-mercaptopropionic acid. The GABA(B) receptor antagonist CGP 56999 and kainic acid induced a similar incidence of seizures in GAERS and NE rats and predominantly activated the hippocampus. No difference in the tonic seizures elicited by strychnine could be evidenced between the strains. These results suggest that an abnormal cortical GABAergic activity may underlie absence seizures in GAERS.

FOS induction in brain associated with seizure and sustained cortical vasodilation in anesthetized rat

PURPOSE

By estimating the anatomical distribution of neurons expressing c-fos protein, we sought to establish whether the intrinsic neural systems known to be implicated in the cerebrovascular regulation were activated during the increase in cortical blood flow associated with epileptic seizures.

METHODS

A single unilateral microinjection of the cholinergic agonist, carbachol, in the thalamic generalized convulsive seizure area was used in anesthetized rats to elicit recurrent episodes of electrocortical epileptiform activity and an increase in cortical blood flow. Neuronal expression of Fos protein was analyzed to identify activated brain regions.

RESULTS

We identified two cortical vasodilatory responses: a sustained cortical vasodilatory response associated with the continuous low-frequency, high-amplitude spiking and a transient cortical vasodilatory response invariably related to the recurrent spike-burst activity. The sustained cortical blood flow began to increase at 55-65 min, remaining significantly (p < 0.05) increased and reaching at the end of the experiment < or =182+/-17% of the prestimulated control. The electrocortical epileptic activity and the cerebral cortical vasodilation were associated with a marked increase in Fos immunoreactivity in the entorhinal and piriform cortices, the dentate gyrus, the hippocampus, and the amygdala. Fos-positive neurons also were found in specific thalamic nuclei, the cerebral cortex, the caudate-putamen, the hypothalamus, the pontine parabrachial nuclei, the dorsal raphe, and the rostral ventrolateral medulla.

CONCLUSIONS

These results provide evidence that convulsive seizures elicited by cholinergic stimulation of the thalamus, in addition to limbic and somatic motor systems, activate central autonomic nuclei and their pathways, including those implicated in cerebrovascular regulation.

Spatial and temporal relationships between C-Fos expression and kindling of audiogenic seizures in Wistar rats

In a strain of Wistar rats selected in our laboratory, audiogenic seizures (AS), characterized by a wild running phase followed by a tonic seizure, can be elicited by exposure to sound. In these animals repeated daily stimulations induce permanent changes which reflect the extension of seizure activity from the brainstem to the forebrain. C-Fos immunoreactivity was used to further characterize the sound-susceptibility of the strain and to specify the spatiotemporal relationships between c-Fos expression and development of AS kindling. AS susceptible rats appeared to be more sensitive to a subthreshold sound as compared to controls. Sound-evoked wild running induced a similar pattern of c-Fos as a full AS in naive rats, confirming the epileptic nature of this early component. AS-induced c-Fos labeling in the auditory pathways of the brainstem extended to the forebrain with repetition of AS and marked increases in c-Fos expression sequentially occurred in the amygdala and perirhinal cortex, followed by the frontoparietal cortex, the piriform cortex, and finally the hippocampus and entorhinal cortex. These results show that the kindled AS preferentially propagate from the brainstem, through the amygdala and the perirhinal cortex, to the motor cortex, with the piriform cortex and hippocampus as secondary targets. No more c-Fos expression was detected 24 h after an AS. A down-regulation of cortical c-Fos induction was observed 1 and 2 days after daily exposure to kindled AS, with full recovery of c-Fos expression after a 5-day seizure-free period. This suggests a regulatory function of c-Fos expression in development of kindling.

Modulation of the audiogenic seizure network by noradrenergic and glutamatergic receptors of the deep layers of superior colliculus

Recent studies suggest that the deep layers of superior colliculus (DLSC) play a role in the network for audiogenic seizures (AGS) in genetically epilepsy-prone rats (GEPR-9s). The present study examined the role of glutamatergic and noradrenergic receptors in DLSC in modulation of AGS susceptibility. The study examined effects of a competitive NMDA receptor antagonist [dl-2-amino-7-phosphonoheptanoic acid (AP7)] or an alpha1 noradrenergic agonist (phenylephrine) focally microinjected into DLSC as compared to effects in the inferior colliculus (IC) and pontine reticular formation (PRF), which are major established components of the AGS network. The results demonstrated that blockade of NMDA receptors in DLSC suppressed AGS susceptibility. AP7 microinjection was effective at relatively low doses in IC, but required higher doses in DLSC and PRF. The DLSC was relatively more sensitive to seizure reduction by the alpha1 noradrenergic agonist as compared to the IC and PRF. The anticonvulsant effect of AP7 was longer-lasting than phenylephrine in the DLSC and IC but not in the PRF. These data suggest that neurons in the DLSC are a requisite component for the neuronal network for AGS in GEPR-9s and that NMDA and alpha1 adrenoreceptors in this site may play important roles in the modulation of AGS propagation. The relatively greater sensitivity of DLSC to phenylephrine as compared to IC and PRF indicates that norepinephrine may be more important in the modulation of AGS in DLSC, which contrasts to the role of glutamate modulation. These data support recent neuronal recording data, which indicate that DLSC neurons play a critical role in AGS.

Inducible and constitutive transcription factors in the mammalian nervous system: control of gene expression by Jun, Fos and Krox, and CREB/ATF proteins

This article reviews findings up to the end of 1997 about the inducible transcription factors (ITFs) c-Jun, JunB, JunD, c-Fos, FosB, Fra-1, Fra-2, Krox-20 (Egr-2) and Krox-24 (NGFI-A, Egr-1, Zif268); and the constitutive transcription factors (CTFs) CREB, CREM, ATF-2 and SRF as they pertain to gene expression in the mammalian nervous system. In the first part we consider basic facts about the expression and activity of these transcription factors: the organization of the encoding genes and their promoters, the second messenger cascades converging on their regulatory promoter sites, the control of their transcription, the binding to dimeric partners and to specific DNA sequences, their trans-activation potential, and their posttranslational modifications. In the second part we describe the expression and possible roles of these transcription factors in neural tissue: in the quiescent brain, during pre- and postnatal development, following sensory stimulation, nerve transection (axotomy), neurodegeneration and apoptosis, hypoxia-ischemia, generalized and limbic seizures, long-term potentiation and learning, drug dependence and withdrawal, and following stimulation by neurotransmitters, hormones and neurotrophins. We also describe their expression and possible roles in glial cells. Finally, we discuss the relevance of their expression for nervous system functioning under normal and patho-physiological conditions.

The role of the ventromedial nucleus of the hypothalamus in epileptogenesis

Flurothyl kindling initiates a time-dependent process that results in a facilitated propagation from the forebrain to the brainstem seizure system and in an increase in the complexity of behavioral seizure expression. We investigated the involvement of the ventromedial nucleus of the hypothalamus (VMH) in mediating this facilitated propagation between these seizure systems. Bilateral ibotenic acid lesions of the VMH, but not the dorsomedial nucleus of the hypothalamus (DMH), resulted in a disruption in the propagation of seizure activity from the forebrain to the brainstem. Moreover, VMH lesioned mice were able to express brainstem seizures following minimal corneal electroconvulsive shock (mECS). Together, our results indicate that the VMH is a critical substrate involved in propagating seizure activity between the forebrain and brainstem, but is not involved in the expression systems necessary for forebrain or brainstem seizure manifestations.

Expression of Fos in the superior lateral subdivision of the lateral parabrachial (LPBsl) area after generalized tonic seizures in rats

Generalized tonic-clonic seizures of brain stem origin in rats are associated with acute induction of neuronal Fos in several discrete regions of the brain. One particular site in the dorsal pons shows remarkable Fos induction following generalized tonic seizures induced by maximal electroshock in normal rats or by audiogenic stimulation in genetically epilepsy-prone rats (GEPRs). Although this area shows the most intense Fos induction of any brain area following generalized tonic seizures, its identity has been uncertain. Based on its general location, we hypothesized that this nucleus was either 1) a component of the pedunculopontine tegmentum nucleus-pars compacta (PPTn-pc) or 2) the superior lateral subnucleus of lateral parabrachial area (LPBsl). The present study used Fos-protein immunocytochemistry in combination with the reduced form of nicotinamide-adenine dinucleotide phosphate (NADPH)-diaphorase histochemistry, cholecystokinin (CCK) immunocytochemistry, and neuronal tract-tracing to determine the identity of this cluster of Fos-immunoreactive neurons in the dorsal pons. Following maximal electroshock seizure (MES), Fos labeling was compared to NADPH diaphorase staining (a marker for cholinergic neurons of the PPTn-pc); retrograde transport of wheat germ agglutinin-horseradish peroxidase (WGA-HRP) injected into the ventromedial nucleus of the hypothalamus (VMH; to identify the LPBsl) or CCK immunoreactivity (also a marker for LPBsl neurons). Results showed this cluster of Fos immunoreactive (FI) neurons to be closely associated, but not overlapping, with the lateral and most caudal aspect of the PPTn-pc. Alternatively, WGA-HRP retrograde-labeled neurons corresponded precisely with the seizure-induced FI neurons. Additionally, the location of CCK immunoreactive neurons directly overlapped with the FI neurons, although they were not nearly as prevalent. These results demonstrate that the seizure-induced FI neurons in this area are neurons of the LPBsl and not cholinergic neurons of the PPTn-pc. This is the first report of seizure-induced Fos expression specifically localized to the superior lateral subnucleus of the lateral parabrachial area.

Regional analysis of the spatial patterns of Fos induction in brain following flurothyl kindling

We have recently demonstrated that eight, daily flurothyl-induced generalized clonic seizures, followed by a four week stimulus-free interval, results in a long-lasting reduction in generalized seizure threshold and a change in the type of seizure expressed in response to flurothyl from clonic to tonic. There is a progressive increase in the probability that a mouse will express a tonic seizure during the four week interval, suggesting that prior flurothyl seizures initiate a proepileptogenic process that requires time to develop. In this study, the immunohistochemical detection of the c-fos protein (Fos) was used to evaluate whether seizure-induced epileptogenesis resulted in regional differences in the degree of neuronal activation. Fos immunoreactivity was examined 1.5 h following either a single generalized seizure, the last of eight consecutive daily seizures or a retest seizure evoked two weeks after the last of eight seizures. In each condition, generalized seizure behaviours were elicited in C57BL/6 mice using flurothyl and classified as either "forebrain" (face and forelimb clonus) or "brainstem" (running/bouncing, treading, tonic extension). The spatial distribution of Fos induction was compared on the basis of the seizure phenotype and the seizure history. The predominant differences in Fos distribution were found to be related to the type of seizure expressed regardless of the seizure history. Furthermore, the different motor components that make up a "brainstem" seizure could not be distinguished by the pattern of Fos labelling suggesting that multiple convulsive behaviours are mediated by one anatomical system. Finally, Fos induction in the ventromedial hypothalamic nucleus preceded and predicted the change in seizure type from "forebrain" to "brainstem". These data support the concept that separate anatomical systems mediate the expression of the two generalized seizure phenotypes. In addition, the ventromedial nucleus of the hypothalamus may be a point of interaction between the systems and may play a role in seizure-induced neural reorganization.

A role for the bilateral involvement of perirhinal cortex in generalized kindled seizure expression

The perirhinal cortex (PRh) has been suggested as a substrate for the expression of generalized clonic seizures in the late stages of kindling development (stages 4-5). Using the induction of Fos as a marker of neuronal activation, the PRh region was investigated after kindling or nonkindling electrical stimulation. Nonkindling electrical stimulation of the PRh elicited stimulus-locked behaviors, without afterdischarge. These behaviors were characterized by rearing and bilateral forelimb clonus which were terminated upon electrical stimulus offset in half of the rats displaying this behavior (with the other half expressing self-sustained seizures). In these animals, Fos immunoreactivity was found throughout neocortical and subcortical structures in the hemisphere ipsilateral to the stimulating electrode. By contrast, Fos-immunoreactivity in the contralateral hemisphere was localized primarily in the PRh and frontal motor cortex. Likewise, similar patterns of Fos immunoreactivity were observed in both hemispheres of rats following kindling to one generalized clonic seizure from several limbic and paleocortical structures. These results suggest that the bilateral involvement of the PRh is critical in producing the bilateral behaviors associated with generalized clonic seizure expression. In support of this interpretation, infusion of 3 M KCl directly into the contralateral PRh of rats kindled to a single stage 4-5 (generalized clonic) seizure from the ipsilateral amygdala reduced seizure manifestations from a generalized clonic seizure (stage 4-5) to a unilateral clonic seizure (stage 3) without affecting measures of focal excitability. Taken together, these data indicate a role for the bilateral involvement of the PRh in generalized clonic seizure expression whether evoked from the naive or kindled state. These results further indicate that bilateral behaviors require the bilateral involvement of the structures necessary for the expression of these behaviors.

Decreased brainstem seizure thresholds and facilitated seizure propagation in mice exposed to repeated flurothyl-induced generalized forebrain seizures

We recently have described a novel model of epileptogenesis utilizing the inhalant chemoconvulsant, flurothyl (Applegate et al., 1997; Samoriski and Applegate, 1997). The hallmark feature of this model is a change in behavioral seizure phenotype from a forebrain seizure, observed during the initial flurothyl exposures, to a brainstem seizure, elicited by flurothyl, after a 28-day stimulation free incubation period. In this study, we sought to establish the basis for this change in behavioral seizure response. To this end, we examined the effects of exposure to this paradigm on the generalized brainstem seizure threshold and on the propagation of forebrain seizures onto the brainstem seizure substrate. Ten mice were given flurothyl-induced generalized forebrain seizures on 8 consecutive days (induction phase). The other ten mice were not exposed to the flurothyl induction paradigm and served as controls. Minimal corneal electroconvulsive shock (mECS--20 mA) was used to assay whether there was any change in the animals' generalized brainstem seizure thresholds at 3, 14 and 28 days following the last flurothyl seizure trial. Mice that were exposed to flurothyl exhibited a progressive increase in the percentage of animals having a mECS-induced brainstem seizure when tested at 3 (40%), 14 (70%) and 28 (90%) days following the last flurothyl seizure. Control mice rarely had a brainstem seizure at any of the three time points tested, mostly forebrain seizures were observed. These results suggest that there is a significant progressive lowering of the brainstem seizure threshold, during the incubation phase of the flurothyl paradigm, which is coincident with the previously reported time course of change in the behavioral seizure phenotype observed using this flurothyl model (Applegate et al., 1997; Samoriski and Applegate, 1997). Following mECS testing, mice were implanted with bipolar electrodes and kindled from the olfactory bulb (OB). Mice exposed to the flurothyl paradigm demonstrated significantly faster kindling rates, longer afterdischarge durations. and longer durations of and latencies to stage 5 seizures compared to controls. Furthermore, animals exposed to the flurothyl protocol demonstrated an increase in the expression of brainstem seizures after focally-elicited OB afterdischarges. These results suggest that there is an increased interaction between the forebrain and brainstem seizure systems after exposure to this model of epileptogenesis. Together, results indicate that the change in behavioral seizure phenotype observed following exposure to our flurothyl paradigm are promoted by both decreases in brainstem seizure thresholds and facilitated forebrain seizure propagation onto the brainstem seizure system.