Portal de Periódicos da CAPES

Fourth Conference on Epileptogenesis, May 23–26, 2007, Pisa, Italy

2008; Wiley; Volume: 49; Issue: 5 Linguagem: Inglês

10.1111/j.1528-1167.2007.01518_1.x

1528-1167

Yuri Bozzi, Annamaria Vezzani, Michele Simonato, Marco de Curtis, G. Avanzini, Matteo Caleo,

Photoreceptor and optogenetics research

The Fourth Conference on Epileptogenesis was aimed at discussing new and controversial issues in the area of epileptogenesis research in the format of a workshop. Epileptogenesis is a particularly integrative field of research, in which experimental data and clinical evidence are brought together to provide insight into pathological plasticity phenomena and, more broadly, into the mechanisms of brain adaptation to environmental stimuli. The conference was opened by Lamberto Maffei, who presented an overview on the mechanisms regulating the postnatal development and plasticity of the rodent visual system. Visual system development and plasticity have been shown to depend largely on a proper sensory experience during early postnatal life. Recent studies have demonstrated the positive effects of environmental enrichment on the maturation and function of the visual system (Sale et al., 2007). These studies have profound implications for understanding both physiological and pathological development of CNS. The first session addressed the role of ion channels and transporters in epileptogenesis. Aristea Galanopoulou presented a study on the effects of the expression of KCC2 on GABA function in brain development. During development, GABAA receptors mediate depolarizing neuronal responses, whereas in the adult they are hyperpolarizing. The switch from depolarizing to hyperpolarizing GABAA receptor signaling is triggered through the developmental shift in the balance of intracellular chloride cotransporters. The maturation of GABAA signaling follows sex-specific patterns correlated with expression profiles of chloride cotransporter KCC2 (Galanopoulou, 2007), suggesting that GABAA receptor activation may be differentially regulated in males and females. This finding has particular implications in epilepsy, since GABAA receptor activation by antiepileptic drugs may differentially influence brain development in males and females. The role of calcium channels in epileptogenesis was discussed by Heinz Beck. In the pilocarpine model of temporal lobe epilepsy (TLE), a conversion from regular to burst firing (that is likely to play a role in TLE ictogenesis) was observed in hippocampal CA1 pyramidal neurons. Specific changes in the expression and properties of Cav3.2 T-type Ca2+ channel subunits were also detected (Su et al., 2002). Patch-clamp analyses of CA1 neurons from pilocarpine-treated mice revealed that the burst-firing mode does not occur in mice lacking Cav3.2 channels. Moreover, the frequency and severity of chronic seizures, as well as the neuropathological sequelae were significantly attenuated in chronic epileptic Cav3.2 knockout mice, suggesting that Cav3.2 up-regulation after SE may contribute to initiation of seizure activity and neuronal cell death in chronic epilepsy. Potassium channel subunits mediate neuron repolarization following action potential discharge, and their mutations give rise to different types of seizure disorders in humans and in mouse models. The mechanisms underlying potassium channelopathies, discussed by Jeff Noebels, involve modified patterns of sustained repetitive firing properties in both excitatory cells and interneurons. Examples of mouse models include targeted deletion of the Kcna1, Kcnc2, Kcnq2/3, and the Kcnmb4 genes (mediating Kv1.1, Kv 3.2, Kv7.2/3, and BK currents, respectively). Since human epilepsy is often the product of complex or multigenic inheritance, it is crucial to analyze how mutations in more than one channel may interact. One new example of a digenic mouse model, involving Kcn1a and Cacna1a genes reveals multiple levels of physiological interaction leading to partial masking of the seizure phenotypes in these mutants (Glasscock et al., 2007). HCN1/HCN2 channel subunits responsible for hyperpolarization-activated current (Ih) are expressed at high density in dendrites and regulate overall dendritic excitability. As discussed by Daniel Johnston, Ih is partly active at resting potential and acts globally on dendritic inputs. These properties of Ih determine its ability to influence synaptic potentials in both hyperpolarizing and depolarizing membrane potential ranges. Temporal summation of excitatory synaptic potentials is reduced by Ih, regardless of the dendritic location of synapses. Long-term changes in the expression of Ih in pyramidal neurons of CA1 and entorhinal cortex have been demonstrated in acute and chronic models of TLE. Ih decreases and redistributes from dendrites to soma, significantly altering the oscillatory properties of pyramidal neurons, both in conditions of normal excitability (during long-term potentiation) and during epileptogenesis (Shah et al., 2004). Mutations in nicotinic receptor channels have also been associated with the epileptic phenotype. In particular, a missense mutation in the β2 subunit of neuronal nicotinic receptor (CHRNB2) was demonstrated in patients affected by autosomal dominant nocturnal frontal lobe epilepsy (ADNFLE), a focal form of epilepsy characterized by seizures occurring during sleep and originating from the frontal lobe (De Fusco et al., 2000). Irene Manfredi presented a new transgenic ADNFLE mouse model generated by using the TET-OFF system. EEG recordings from transgenic mice expressing the mutant β2 subunit reveal frequent spikes and spontaneous seizures. Preliminary evidence shows the possibility to revert the epileptic phenotype by chronic oral administration of doxycycline, a compound known to silence the transgene. Neurons with long-range connections are essential for transiently “binding” different brain regions and could also play a key role in pathological synchronizations, such as epileptic seizures. Christophe Bernard reported that transient fast oscillations at seizure onset in the immature hippocampus result from the high-frequency firing of most of the GABAergic interneuron population (Cossart et al., 2005). A specific class of long-range interneurons (hippocampus-septum cells), whose axon arborizes in the whole hippocampus and in the septum, is necessary to entrain other interneurons into high frequency firing. Destruction of these cells abolishes fast oscillations but not seizures, providing evidence that neurons with long-range connections control synchronized oscillations. These studies suggest that fast oscillations are not necessary for ictogenesis, but rather constitute a signature of network activity. The second session addressed different aspects of brain development implicated in epileptogenesis, Renzo Guerrini presented data on genetic disorders of cortical development associated with epilepsy. X-linked periventricular nodular heterotopia (PNH) consists of confluent nodules of gray matter located along the lateral ventricles and is often associated with focal epilepsy. Filamin (FLN1) mutations have been reported in both familial and sporadic cases. A recessive form of PNH due to ARGEF2 gene mutations has also been reported in children with microcephaly, severe delay and early-onset seizures. Lissencephaly-pachygyria and subcortical band heterotopia (SBH) result from mutations of either the LIS1 or XLIS gene. Finally, X-linked lissencephaly with corpus callosum agenesis and ambiguous genitalia is associated with mutations of the ARX gene (Barkovich et al., 2005; Guerrini & Marini, 2006). Antonio Simeone discussed the role of homeobox-containing transcription factors of the Otx family in glutamatergic versus GABAergic differentiation during development. Genetic ablation of Otx2 in glutamatergic progenitors of the thalamus results in the expression of GABAergic markers in these cells. Reported data suggest that Otx2 may prevent GABAergic fate switch by repressing the Mash1 gene in glutamatergic progenitors (Puelles et al., 2006). Genetic mechanisms regulating the generation of GABAergic neurons were also addressed by Inma Cobos, who focused on the role of Dlx homeobox transcription factors. Cortical interneurons express Dlx1 and Dlx2 during development and into adulthood, and their differentiation depends on the number of expressed copies of these genes. Dlx1-/-;Dlx2-/- double mutant mice show defects in the differentiation and survival of GABAergic interneurons, and a migration arrest of immature interneurons from the subpallium to neocortex and hippocampus. Conversely, interneurons in Dlx1-/- mutants show no migration defects, but subsets of them fail to survive. Dlx1-/- mice exhibit generalized seizures and seizure-induced reorganization, linking the Dlx1 mutation to delayed-onset epilepsy associated with interneuron loss (Cobos et al., 2005, 2007). Rafael Gutierrez discussed the significance of the glutamate/GABA switch in epilepsy. As a consequence of seizures, glutamatergic granule cells of the dentate gyrus (DG) produce GABA and GABAergic markers (GAD67, vGAT), and elicit GABAA-receptor-mediated responses in CA3. During the first postnatal weeks, GABAergic input to dendrites of CA3 pyramidal cells in the stratum lucidum (SL) evokes depolarizing postsynaptic potentials, due to the persistent expression of the NKCC1 cotransporter in CA3 dendrites. Conversely, in adult rats, the DG tonically inhibits CA3 activity through GABA-mediated signaling. Thus, the release of GABA from mossy fibers may contribute to increased seizure susceptibility in the early postnatal period (Gutierrez, 2005; Trevino et al., 2007). Giulia Curia described altered GABAergic function in fmr1 knockout (KO) mice, a model of fragile X syndrome. In the subiculum, GABAA receptor-mediated phasic and tonic components are different in fmr1 KO as compared to wild-type mice. In addition, the expression of subunits involved in GABAergic tonic inhibition is down-regulated in the subiculum of fmr1 KO mice. Thus, alterations in GABAergic transmission occurs in fragile X animals, suggesting that different expression of GABAA currents may lead to hyperexcitability and epilepsy in fragile X patients. Tamar Chachua presented data on the effect of thalamic stimulation on the generation of seizure activity. Some GABAergic neurons of the thalamic reticular nucleus (TRN) fire at high frequency during the silent periods of clonic discharge and at the end of generalized seizures induced by hippocampal kindling. TRN stimulation during hippocampal kindling induces marked suppression of limbic motor seizures due to potentiated activity of TRN inhibitory neurons (Nanobashvili et al., 2003). The last two presentations in this session addressed the role of synaptic vesicle proteins in epilepsy. Mutations in the synapsin 1 and 2 genes have been identified in families of patients with partial temporal lobe or frontal lobe epilepsy (Garcia et al., 2004), and deletion of synapsin genes is associated with epilepsy in mice (Rosahl et al., 1995). Fabio Benfenati showed that patch-clamp recordings on cultured hippocampal neurons from synapsin KO mice reveal a reduced amplitude of evoked inhibitory postsynaptic currents. These results are reflected by a marked increase in bursting activity of synapsin KO networks in culture, as compared to wild-type networks. These data indicate an involvement of synapsins in the balance between inhibitory and excitatory transmission and suggest that they play a role in the etiology of human epilepsy. Flavia Antonucci discussed the neuroprotective effects of botulinum neurotoxin E (BoNT/E, a protease that blocks neurotransmitter release via cleavage of the synaptic protein SNAP-25) in a mouse model of TLE. Intrahippocampal delivery of BoNT/E after kainic acid-induced status epilepticus (SE) delays epileptogenesis without preventing the occurrence of chronic seizures. However, BoNT/E significantly reduces granule cell dispersion, CA1 cell loss and reelin down-regulation that occur after SE (Antonucci et al., 2008). These findings suggest that specific morphogenetic changes following SE are not implicated in epileptogenesis. In the third session, innovative strategies to inhibit ictogenesis and possibly epileptogenesis were presented. Karen Gale discussed the effects of electroconvulsive shock (ECS) preconditioning on the long-term outcomes of SE. Chronic, but not acute, exposure to minimal ECS prevents neuronal damage induced by subsequent SE, presumably via transcriptional up-regulation of fibroblast growth factor-2 (Kondratyev et al., 2002). Evidence on the role of chromatin modifications (such as phosphorylation of histone variant H2A.X; Crowe et al., 2006) in epileptogenesis after ECS-preconditioning and SE was also presented. Michael Rogawski described the potential use of convection-enhanced delivery (CED) for the treatment of focal epilepsy. CED consists in delivering therapeutic substances to a localized brain region by slowly infusing a solution under positive pressure through a fine cannula. N-type calcium channel toxins and botulinum toxins were delivered using CED in fully kindled animals, resulting in a dose-dependent increase in the afterdischarge threshold and a decrease in its duration. Behavioral seizure score and duration were also decreased up to 1 week for calcium channel toxins and up to 50 days for botulinum toxins (Gasior et al., 2007). Jana Veliskova described the anticonvulsant and neuroprotective effects of β-estradiol. Administration of β-estradiol to ovariectomized females in doses producing physiological concentrations, delays seizure onset and prevents SE-induced damage of hippocampal neurons. Neuroprotection includes the damage-sensitive subpopulation of neuropeptide Y (NPY)-containing hilar interneurons. Parallel effects of β-estradiol and NPY suggest a possible estrogen-NPY interaction (Velísková & Velísek, 2007). Mireille Lerner-Natoli addressed the role of angiogenesis and blood-brain barrier (BBB) function in epileptogenesis. Neovascularization and loss of BBB integrity was found in hippocampi from TLE patients, as compared to hippocampi from nonepileptic subjects (Rigau et al., 2007). These alterations positively correlated with seizure frequency. Vascular endothelial factor (VEGF) and tyrosine kinase receptors were highly expressed by neurons and endothelial cells, respectively, and might be implicated in neovascularization. Accordingly, VEGF overexpression and BBB impairment occurred early after experimental seizures in rats, followed by a progressive increase in vascularization which was maintained in chronically epileptic tissue. Robert Schwarcz illustrated the functional changes in glial cells occurring in the brain of patients suffering from chronic epilepsy and in various animal models of the disease. These changes, which affect both microglial cells and astrocytes, have been traditionally viewed as being secondary to neuronal loss and of little if any functional significance. Recent studies, however, indicate that selective glial changes frequently precede seizure activity and seizure-induced neurodegeneration; activated glial cells produce and release neuroactive metabolites which can influence seizure threshold and neuronal viability, and abnormal glial cells can be targeted selectively for the focal delivery of antiepileptic principles (Schwarcz & Pellicciari, 2002). Barbara Gagliardi described the interleukin-1 system as a novel pathway involved in ictogenesis and possibly in epileptogenesis, which may offer a nonconventional antiepileptic approach (Vezzani & Baram, 2007). IL-1β is rapidly synthesized by glia in rodent brain during acute seizures, and its production persists during epileptogenesis and in chronic epileptic tissue. When injected into the rodent hippocampus, IL-1 has proconvulsant effects, while inhibition of its signal transduction pathway or blockade of its endogenous synthesis affords significant anticonvulsant and antiepileptogenic effects. Eleonora Palma, using membrane extracts from the brain of epileptic patients microtransplanted into Xenopus oocytes, provided electrophysiological evidence supporting the hypothesis that rundown of GABAA receptors is a pathologically relevant dysfunction for epilepsy (Palma et al., 2005). The fourth session of the meeting focused on neuronal damage and repair in acquired epilepsies. Asla Pitkänen investigated whether changes in magnetic resonance imaging (MRI) could be used to predict posttraumatic epileptogenesis and cognitive decline in the traumatic brain injury (TBI) model. From all MRI parameters analyzed, changes in postinjury diffusion trace (Dav) in the hippocampus were most consistently associated with the long-term outcome (spike activity, mossy fiber sprouting, memory impairment) (Kharatishvili et al., 2007). Thus, damage severity measured with MRI may predict seizure susceptibility and cognitive outcome after TBI. Many preclinical studies on acquired epilepsies use convulsive SE as an epileptogenic stimulus. However, convulsive SE fails to reproduce many important features of the human disease. To develop better animal models, Robert Sloviter hypothesized that human-pattern hippocampal sclerosis (HS) is the result of less than maximally intense excitation—a level of excitation that activates the hippocampus but stays sequestered within temporal circuits (in contrast to convulsive SE, which evokes seizure activity predominantly in extrahippocampal pathways). Sloviter described two new animal models that approximate the features of human TLE, both models involving perforant path stimulation and having the advantages of minimal variability, minimal lethality, spontaneous hippocampal-onset seizures, and human-pattern HS (Sloviter et al., 2007). If cell loss is a critical factor for the development of epilepsy, one may expect that neurogenesis represents an insufficient attempt to halt the process by repairing the damage. Jack Parent is studying the influence of neurogenesis in the epileptic adult hippocampus. Increased precursor cell proliferation and ectopic location of mature granule cells in the hilus and in the molecular layer was found in pilocarpine-treated adult rats (Parent et al., 2007). These ectopic cells look remarkably similar in epileptic human and rat DG. Anatomical studies indicated that ectopic adult-born neurons integrate long-term and persistently exhibit abnormal properties (Jessberger et al., 2007). Thus, neurogenesis in the epileptic hippocampus is not only insufficient, but also altered, and likely contributes to seizures. This conclusion leads to the idea that endogenous neurogenesis should be supplemented and/or guided toward a therapeutic aim (tissue repair). One approach for supplementation is the transplant of cells. Ashok Shetty used hippocampal fetal cell (HFC) grafting for repairing damage and treating chronic TLE. He quantified survival and antiseizure effects of HFC grafts transplanted into the hippocampi of rats displaying spontaneous recurrent seizures (SRS) following kainate-induced SE. HFCs treated and grafted with neurotrophic factors and a caspase inhibitor (but not standard HFC grafts) survived well, differentiated into neurons and blunted the progression of TLE (Rao et al., 2007). Thus, grafting of precursor cells into the hippocampus might be effective in restraining SRS in chronic TLE. The alternative approach—guiding the endogenous progenitors to survival and integration—has been explored by Beatrice Paradiso. Once epileptogenic damage was established, she injected in the hippocampus a herpes vector engineered to locally supplement two neurotrophic factors, FGF-2 and BDNF (a mitogenic and a neuro-differentiating agent for neural progenitors). This treatment increased neuronogenesis, repaired neuronal damage and prevented epileptogenesis (Paradiso et al., 2008). Another interesting vector type may be based on adeno-associated virus (AAV). Merab Kokaia reported advancements on one such vector supplementing NPY, a homeostatic agent for synaptic transmission and plasticity. When overexpressed in the rat hippocampus using recombinant AAV (rAAV) vectors, NPY exerts antiepileptic and antiepileptogenic effects. Interestingly, hippocampi overexpressing NPY have a partial reduction of LTP magnitude, due to activity-dependent release of transgene-mediated NPY, thus decreasing glutamate release. These studies suggest that a rAAV-based gene therapy approach using NPY could be useful for treating intractable forms of epilepsy. Finally, Marzena Stefaniuk presented preliminary data from a large microarray screening study aimed at identifying novel epileptogenesis-related genes in the rat brain. Samples were collected at 1, 4, or 14 day after kainic acid or pilocarpine-induced SE. Recent experimental evidences demonstrated that glial cells exert a primary role in the control of neuronal excitability. An update of the new findings on glia and the control of epileptic activity was the topic of the last session of the conference. Giorgio Carmignoto focused on the interactions between glia, neurons, and brain vessels (Haydon & Carmignoto, 2006). Following neuronal activity, glutamate-mediated Ca2+ elevations in astrocyte processes are transferred to astrocyte endfeet that contact cerebral blood vessels and control their tone (Zonta et al., 2003). Interictal and ictal discharges exert different effects on astrocytes in entorhinal cortical slices. Ictal events triggered a highly synchronized Ca2+ response in astrocyte endfeet along the entire length of the blood vessel, followed by a rapid arteriole response. In contrast, interictal events triggered only discrete Ca2+ oscillations in endfeet, accompanied by rare and spatially restricted responses of the arteriole. Astrocytes thus may regulate cerebral blood flow in the epileptic brain. Glial cells can, by multiple functions, influence neuronal excitability. Uve Heinemann reported on spatial K+ buffering, i.e., the facilitated redistribution of potassium from sites of maximal K+ release to remote sites where extracellular potassium is low. Astrocytes expressing glutamate transporters (Glu-T type) are involved in spatial K+ buffering, mostly through Kir 4.1 and 2P channels. Spatial K+ buffering is strongly reduced when Kir channels are blocked, an effect which can be augmented by additional blockade of 2P channels. This effect is often missing in chronic epileptic tissue, likely as a result of altered BBB permeability. Following artificial BBB opening, astrocyte activation correlates with a strong down-regulation of Kir channels and disturbed K+ buffering. As a result, increased K+ accumulation and appearance of epileptiform discharges can occur (Ivens et al., 2007). The relevance of altered functional astrocyte properties to seizure generation in postsurgical specimens of human hippocampi was addressed by Christian Steinhäuser. Astrocytes can be differentiated in two populations, one coupled via gap junctions and expressing glutamate transporters (GluT-type), and the other expressing AMPA-type glutamate receptors and receiving direct synaptic inputs from neurons (GluR-type). In human hippocampal specimens obtained from patients with pharmaco-resistant TLE, GluR- and GluT-like cells could be detected in nonsclerotic tissues. In contrast, in the hippocampus of patients with HS, GluT-type astrocytes disappeared, and GluR-cells displayed slower receptor desensitization (Seifert et al., 2004). Thus, in HS, the ability of astrocytes to clear the extracellular space of glutamate and K+ is dramatically impaired, therefore facilitating the generation and spread of seizure activity (Seifert et al., 2006). Glial cells express receptors for neurotransmitters that are activated through spillover or release from extrasynaptic sites. Dwight Bergles provided evidences that neurons in the hippocampus, cortex, and cerebellum form direct synaptic junctions with glial progenitor cells known as “NG2+ cells.” These junctions may provide a means to couple neuronal activity to changes in the proliferation and behavior of these glial progenitors. Alternatively, abnormal activity may lead to changes in the local environment, which alter the fate of these cells. Chemically induced seizures cause the appearance of periodic bursts of EPSCs in NG2+ cells in the hippocampus, suggesting that abnormal neuronal activity may trigger changes in these progenitors, altering both their association with neuronal elements and their fate in the mature CNS (Paukert & Bergles, 2006; Ziskin et al., 2007). Giuseppe Biagini described the modulatory effects of neurosteroids on epileptogenesis. Neurosteroids are mainly contained in glial cells. In brains from animals experiencing SE, the limiting enzyme for neurosteroid production (cholesterol side-chain cleavage cytochrome P450, P450scc), is up-regulated in astrocytes in the hippocampus and extrahippocampal areas. P450scc immunostaining of oligodendrocytes and microglial cells increased progressively after the SE. Treatment of SE animals with finasteride (5α-reductase inhibitor of neurosteroid synthesis) determined an earlier occurrence of seizures in comparison to the vehicle-treated animals (Biagini et al., 2006). Finally, Laura Uva discussed the role of BBB impairment in pilocarpine-induced SE (Marchi et al., 2007). When pilocarpine was applied by arterial perfusion to the in vitro isolated guinea pig brain, no epileptiform activity was induced, unless drugs that increase BBB permeability, such as histamine and bradykinin were coperfused. The data discussed suggest that an increase of BBB permeability could play a crucial role in promoting seizure activity, possibly through an increase of pilocarpine concentration in brain parenchyma. The conference was made possible by the generous contribution of Fondazione Pierfranco and Luisa Mariani (Milan, Italy), International Brain Research Organization (IBRO), Scuola Normale Superiore (Pisa, Italy), Fondazione Cariplo (Milan, Italy) and Istituto Neurologico Carlo Besta (Milan, Italy). The administrative assistance from International School of Neurological Sciences (San Servolo, Venice, Italy), Fondazione Pierfranco and Luisa Mariani and Scuola Normale Superiore was greatly appreciated.