An official website of the United States government
NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.
Noebels JL, Avoli M, Rogawski MA, et al., editors. Jasper's Basic Mechanisms of the Epilepsies. 5th edition. New York: Oxford University Press; 2024. doi: 10.1093/med/9780197549469.003.0026
A growing body of evidence suggests that astrocytes are crucial actors in the initiation and progression of epilepsy. Reactive astrocytes in human and experimental epilepsy display marked morphological, transcriptional, and functional changes, but it is still often unclear whether this represents a causative factor, a consequence, or an adaptive response in epileptogenesis. A distinctive feature of astrocytes is their extensive intercellular coupling via gap junction channels. This allows them to form large syncytium-like functional networks that play essential roles in ion and neurotransmitter homeostasis, gliotransmission, nutrient supply to neurons, and regulation of the extracellular space volume. The first part of this chapter summarizes current knowledge on the role of gap junction channels in epilepsy, their expression and activity in human and experimental epilepsy, and the consequences of their genetic and pharmacological modulation on neuronal excitability and epileptogenesis. Astrocytes also regulate and respond to extracellular glutamate levels in the central nervous system via the Na+-dependent glutamate transporters glutamate transporter-1 (GLT-1) and glutamate aspartate transporter (GLAST) and the metabotropic glutamate receptors (mGluR) 3 and mGluR5. Both impaired astrocytic glutamate clearance and changes in mGluR signaling could contribute to the development of epilepsy. The second part of this chapter summarizes the changes in astrocyte glutamate receptors and transporters in epilepsy. Overall, both astrocytic gap junction channels and glutamate transporters and receptors could serve as novel therapeutic targets for epilepsy.
Current antiepileptic treatments are largely symptomatic and do not prevent or cure the disease. Thus, the prevailing concept considering neurons as critical targets for the development of antiepileptic drugs (AEDs) has been questioned, and increasing evidence suggests that dysfunction of glial cells, rather than neurons, might initiate epileptogenesis (Binder and Steinhäuser, 2021). Indeed, expression and function of various channels, transporters, receptors, and signaling molecules in astrocytes are altered in epilepsy. In this chapter we provide evidence for a crucial role of impaired astroglial gap junction coupling and glutamate signaling in the etiology of some epilepsy forms such as mesial temporal lobe epilepsy (MTLE) and tuberous sclerosis complex (TSC). The findings suggest that astrocytes represent promising targets for the development of alternative, more efficient antiepileptogenic therapies.
Astrocytes, the most abundant glial cell population in the brain, are electrically and metabolically connected to each other via gap junction (GJ) channels, forming large syncytium-like functional networks. GJ channels are composed of protein subunits called connexins (Cxs). Six Cxs oligomerize into a hemichannel (connexon), and two connexons from adjacent cell membranes dock to form a complete intercellular channel permeable to ions and small molecules up to 1 kDa, including second messengers, nucleotides, neurotransmitters, and energy metabolites (Giaume et al., 2020, 2010). Astrocytes express Cx43 (encoded by GJA1 gene) and Cx30 (encoded by GJB6 gene) (Nagy and Rash, 2000). The relative expression levels of the two GJ proteins vary considerably across developmental stages and brain regions. For instance, astrocytes in the hippocampus communicate primarily via Cx43 (Deshpande et al., 2017; Gosejacob et al., 2011), whereas Cx30 prevails in the thalamus (Claus et al., 2018; Griemsmann et al., 2015). The two Cx isoforms differ in their biophysical properties and mode of regulation. Cx30 channels exhibit higher unitary conductance, charge selectivity, and sensitivity to transjunctional voltage than channels composed of Cx43 (Valiunas et al., 1999). Of the two, only Cx43 is regulated by phosphorylation. The C-terminal cytosolic domain of Cx43 contains several phosphorylation sites for various protein kinases that control channel gating and trafficking, protein life cycle, and interactions with scaffolding proteins (Solan and Lampe, 2009).
Through the network organization, astrocytes are able to effectively control and synchronize large neuronal populations. Indeed, GJ-based coupling between astrocytes mediates or facilitates several astrocytic functions, including buffering of extracellular K+ ([K+]o), glutamate clearance from the synapse, regulation of the extracellular space volume, supply of nutrients to neurons, intercellular Ca2 + wave propagation, and adult neurogenesis (Giaume et al., 1997, 2020; Orkand, 1986; Pannasch et al., 2011; Philippot et al., 2021; Scemes and Giaume, 2006; Zhang et al., 2018). Given the essential impact of the astrocyte network on neuronal structure and function, it is not surprising that several neurological disorders have been associated with alterations in astrocyte coupling. Examples include stroke, migraine, gliomas, Alzheimer disease, and epilepsy (Giaume et al., 2020). In the latter disorder, dysfunction of regulatory processes mediated by the astroglial network can in principle both enhance or dampen neuronal excitability. Decreased excitability derives from the contribution of GJ-coupled glial networks to the buffering of [K+]o. According to the spatial K+ buffering concept, excessive [K+]o caused by neuronal activity is passively taken up by astrocytes through inwardly rectifying K+ (Kir) channels, redistributed within the GJ connected astrocytic network and released at distal regions of lower [K+]o. This process does not require energy, as the driving force arises from the difference between the local K+ equilibrium potential and the more negative glial membrane potential (Kofuji and Newman, 2004; Orkand et al., 1966; Walz, 2000) (Fig. 26–1). Such an intracellular redistribution has also been proven to be more efficient than diffusion through the extracellular space (Gardner-Medwin, 1983). In this scenario, loss of GJ coupling between astrocytes would result in [K+]o accumulation and, consequently, in neuronal depolarization and enhanced neuronal excitability. Experimental support for the importance of GJs in [K+]o clearance has been gained in hippocampal slices from transgenic mice with coupling-deficient astrocytes (Pannasch et al., 2011; Wallraff et al., 2006), and from pharmacological disruption of GJ communication between hippocampal and neocortical astrocytes in situ and in vivo (Bazzigaluppi et al., 2017; Breithausen et al., 2020; EbrahimAmini et al., 2021). Work on Cx43 and Cx30 double knockout mice revealed that not only K+ buffering but also glutamate clearance is impaired when astrocyte connexins are absent (Pannasch et al., 2011). Whether GJ channels promote glutamate clearance by mediating its spatial redistribution through the astrocytic network or by preventing astrocyte depolarization and thus maintaining a higher Na+ gradient is unknown. In any case, less efficient removal of glutamate from the synapse would promote seizure activity due to prolonged activation of excitatory synapses. Astrocytes in Cx30–/–, Cx43fl/fl:hGFAP-Cre mice displayed also strong activity-dependent swelling, which can be explained by enhanced water influx due to intracellular K+ and glutamate accumulation. Enhanced astrocyte swelling and the resulting reduction in extracellular space volume, in turn, further increase the concentration of extracellular ions and neurotransmitters, exacerbating the seizure-promoting effect of impaired spatial K+ and glutamate buffering (Pannasch et al., 2011). Although these findings strongly suggest an antiseizure function of the coupled astroglial network, there is also evidence for its pro-seizure function. First, astroglial GJs are essential for the maintenance of synaptic activity under pathological conditions, by mediating activity-dependent intercellular trafficking of energetic metabolites from blood vessels to sites of high energy demand (Philippot et al., 2021; Rouach et al., 2008). Thus, astrocyte coupling might be required to sustain (rather than initiate) seizure activity. A second possible seizure-promoting function arises from the involvement of GJ channels in the propagation of astrocytic Ca2+ waves. These waves can propagate either through a paracrine purinergic pathway or via gap junctional diffusion of IP3 and Ca2+ (Giaume et al., 2020) and mediate nonsynaptic synchronization of large neuronal populations via release of gliotransmitter (Fellin et al., 2004; Tian et al., 2005; Zorec et al., 2012). In this way, GJs might support hypersynchronous neuronal firing that characterizes epilepsy.
Spatial K+ buffering. Activity in a group of neurons has produced local increase in [K+]o to 12 mM (shaded area, left). This provokes a more positive membrane potential (Vm) that passively spreads through the gap-junction coupled astrocytes. The positive (more...)
Undocked Cx hemichannels (HCs) might also function as transmembrane channels to enable bidirectional exchange of ions, metabolites, and signaling molecules between the cytoplasm and the extracellular space (Giaume et al., 2013, 2020). Cx HCs are essentially closed under physiological conditions, but they might open upon strong depolarization, altered intra- and extracellular Ca2+ concentrations, metabolic inhibition, or cytokine release (Giaume et al., 2013; Orellana et al., 2008). In epilepsy, Cx HC activation was proposed to promote neuronal hyperactivity and synchronization through release of ATP, glutamate, or D-serine, which in turn increased synaptic excitability and, in case of ATP, facilitated extracellular Ca2+ wave propagation (Giaume et al., 2020; Gómez-Gonzalo et al., 2010; Steinhäuser et al., 2012; Walrave et al., 2020). It should be noted, however, that the functional characteristics and relevance of HCs are still questioned by many researchers (Nielsen et al., 2017).
Remarkably, in addition to their role as channel proteins, Cxs also display non-channel functions involved in various processes such as cell growth, adhesion, migration, apoptosis, and signal transduction (Giaume et al., 2020). Pannasch and colleagues (2014) reported an epilepsy-relevant, channel-independent function of Cx30. They demonstrated that Cx30 protein regulates excitatory synaptic strength by controlling the insertion of astrocytic processes into synaptic clefts. Indeed, Cx30 deficiency caused astroglial synaptic cleft invasion, resulting in more efficient glutamate clearance and thus decreased excitatory synaptic transmission (Clasadonte and Haydon, 2014; Pannasch et al., 2014). Accordingly, reduced Cx30 protein levels would counteract seizure activity in a channel-independent manner.
Human and rodent astrocytes share similar properties with regard to Cx expression and GJ coupling (Bedner et al., 2020). Several studies investigated epilepsy-related Cx expression changes in epileptic tissue surgically resected from patients with medically untreatable focal epilepsies (Aronica et al., 2001; Collignon et al., 2006; Das et al., 2012; Deshpande et al., 2017; Elisevich et al., 1997; Fonseca et al., 2002; Garbelli et al., 2011; Naus et al., 1991). Here, epileptic tissue samples from patients with focal epilepsies (of various etiologies) were compared with specimens from nonepileptic patients (postmortem, peritumoral, or temporal lobectomy from causes other than epilepsy) or surgical specimens from MTLE patients with hippocampal sclerosis (HS) were compared with specimens from TLE patients with focal temporal lesions but anatomically preserved hippocampal structures (non-HS). Interestingly, the results were quite consistent, as almost all of the above studies found upregulation of Cx43 transcripts or proteins in epileptic (or sclerotic) tissue. The only exception is the study by Elisevich and coworkers (1997), who reported similar Cx43 protein levels and lower Cx43 mRNA levels in hippocampal specimens from patients with intractable epilepsy compared to samples from nonepileptic patients. A detailed analysis of the expression, subcellular distribution, and phosphorylation status of Cxs in human hippocampal specimens with and without HS was performed by Deshpande and colleagues (2017). Their results revealed similar plasma membrane Cx43 levels in human HS, despite increased total Cx43 protein in sclerotic tissue. In contrast, neither total nor membrane-associated Cx30 protein levels were different between the HS and non-HS conditions. Interestingly, the authors also reported subcellular reorganization and altered phosphorylation of Cx43 in HS compared to non-HS, pointing to differences in intercellular coupling (Deshpande et al., 2017). Indeed, it has long been recognized that the expression level of Cxs does not necessarily reflect the extent of functional coupling, as posttranslational modifications, such as phosphorylation, may change gating properties, assembly, or subcellular localization of GJ channels (Deshpande et al., 2017; Steinhäuser et al., 2012). Reliable results can thus only be gathered by performing functional assays. To our knowledge, such functional coupling analyses between human astrocytes in situ have only been performed in one study so far (Bedner et al., 2015). In that study, the authors assessed interastrocytic GJ coupling in hippocampal specimens from patients with medically refractory TLE by analyzing tracer diffusion and found complete loss of coupling in the sclerotic hippocampus, while astrocytes in non-HS specimens were extensively coupled (Bedner et al., 2015). Hence, based on the human data available to date, Cx proteins are abundant in HS, but they do not form functional channels, probably due to abnormal posttranslational processing and/or subcellular reorganization. In lesion-associated (non-HS) TLE, Cxs do form functional channels, but whether the extent of astrocytic coupling is altered compared to nonepileptic hippocampal tissue cannot be deduced, as the latter is not available for functional studies.
Cx expression studies performed in experimental models of seizure and epilepsy yielded much less consistent results (for review, see Bedner and Steinhäuser, 2015; Giaume et al., 2010; Walrave et al., 2020). This inconsistency most probably arises from the large variety of used models as well as from differences in seizure duration and investigated brain regions (Steinhäuser et al., 2012). However, in animal models of TLE (post–status epilepticus (SE) models and kindling), Cx43 and Cx30 mRNA and/or proteins were found either unchanged or upregulated (Condorelli et al., 2003, 2002; Deshpande et al., 2017; Hussein et al., 2017; Kinjo et al., 2014; McCracken and Roberts, 2006; Pannasch et al., 2019; Söhl et al., 2000; Takahashi et al., 2010; Wu et al., 2015), which is in accordance with the human studies. Similar to findings from human specimens, in the intracortical kainate mouse model of TLE total Cx43 protein levels (but not Cx30) were upregulated in the sclerotic hippocampus compared to the contralateral hippocampus, while the membrane-associated fractions of both astrocytic GJ proteins were not different between both sides. Furthermore, and again like in the sclerotic human hippocampus, in the intracortical kainate mouse model ipsilateral Cx43 displayed pronounced subcellular redistribution, altered phosphorylation, and loss of functional coupling (Deshpande et al., 2017). Intriguingly, in this model, loss of astrocyte coupling and the resulting impairment in K+ clearance temporally precede neuronal death and onset of spontaneous seizure activity (Fig. 26–2), pointing to a causal role of this malfunction in the initiation of TLE (Bedner et al., 2015). Reduced astrocytic GJ coupling has also been documented in mouse models of TSC (Xu et al., 2009) and hyperthermia-induced febrile seizures (Khan et al., 2016). In contrast, increased coupling was found 1 week following systemic kainate-induced SE in rats by Takahashi and coworkers (Takahashi et al., 2010).
Decreased tracer coupling and impaired K+ clearance in the latent period after unilateral intracortical kainate injection. A. Representative example showing reduced tracer coupling in the hippocampus 4 hours post injection (hpi). Scale bar = 100 µm. (more...)
In conclusion, most functional coupling analyses in specimens from human and experimental TLE with HS indicated an antiseizure effect of the astroglial network.
Mice with coupling-deficient astrocytes were generated by Wallraff and coworkers by crossing conditional Cx43-deficient mice (Cx43fl/fl:hGFAP-Cre mice; Theis et al., 2003) with mice globally lacking Cx30 (Cx30–/–; Teubner et al., 2003). The resulting double knockout (dKO) mice displayed indeed complete absence of tracer coupling between hippocampal astrocytes (Wallraff et al., 2006). Importantly, acute hippocampal slices from dKO mice exhibited spontaneous epileptiform events, which is most probably a consequence of disturbed K + and glutamate clearance as mentioned above. Furthermore, in slices with disconnected astrocytes, epileptiform activity could be evoked by low-intensity Schaffer collateral stimulation and occurred more frequently and with shorter latency in Mg2+-free solution than in slices from wild-type mice (Wallraff et al., 2006). In a later work, the same mouse line was used to investigate consequences of Cx deficiency on pentylenetetrazole (PTZ)-induced seizures in vivo. Here, dKO mice exhibited a higher number but less severe seizures during short-term (30 min) EEG/video recording. The authors attributed this finding to increased excitability but decreased neuronal release probability and synchronization during activity and suggested that specific blockade of astrocytic GJs could be of therapeutic benefit in seizures (Chever et al., 2016). However, in the intracortical kainate model that covers not only initiation but also sclerotic stages of TLE, dKO mice showed substantially increased seizure and interictal spike activity during the chronic phase (Deshpande et al., 2020). Interestingly, despite elevated epileptic activity, the severity of HS was weaker in dKO mice. These data indicate that loss of GJ coupling between astrocytes promotes chronic seizures but attenuates seizure-induced histopathological changes (Deshpande et al., 2020). Though, the data from dKO mice should be interpreted with some caution, as Cxs are constitutively deleted in these mice and the results may therefore be biased by compensatory developmental changes or by the fact that (in contrast to the situation in human HS) not only functional coupling but also Cx proteins are missing.
Mice lacking only Cx30 develop less severe behavioral seizures during the first 2 h after intraperitoneal injection of kainate (Pannasch et al., 2019). In hippocampal slices it was shown that Cx30 deficiency tunes down neuronal network activity by enhancing the efficiency of synaptic glutamate clearance. This phenomenon was not detected in mice with defective Cx30 channel pores but intact membrane targeting (Cx30 T5M mice), demonstrating that it is mediated by a channel-independent function of Cx30. This is in agreement with the aforementioned role of the protein in controlling invasion of astrocytic processes into the synaptic cleft (Pannasch et al., 2019, 2014).
Taken together, most data derived from dKO mice are in favor of an antiseizure function of the astroglial network, supporting the view that loss of astrocyte coupling represents a crucial event in epileptogenesis (Bedner et al., 2015). Since preserved or increased Cx30 expression has been demonstrated in epileptic tissue (Condorelli et al., 2002; Deshpande et al., 2017; Pannasch et al., 2019), non-channel functions of the GJ protein might also contribute to the progression of epilepsy.
Several studies have examined consequences of pharmacological disruption/activation of GJ communication on seizure activity in a variety of in vitro and in vivo epilepsy models (for review, see Bedner and Steinhäuser, 2015; Walrave et al., 2020). The majority of these studies reported that GJ blockers, such as carbenoxolone (CBX), halothane, or octanol have anticonvulsant effects (Bostanci and Bağirici, 2006; Gajda et al., 2003; Gigout et al., 2006; Jahromi et al., 2002; Medina-Ceja et al., 2008; Nilsen et al., 2006; Ömer Bostanci and Bağirici, 2007; Szente et al., 2002; Vincze et al., 2019; Walrave et al., 2020). However, conclusions from these studies should be drawn with caution, as GJ blockers are usually not Cx isoform-specific, and it remains therefore largely unclear whether the observed anticonvulsive effects were caused by inhibition of electrical synapses between neurons or by blockage of the glial network (Steinhäuser et al., 2012). Moreover, the blockers do not only inhibit intercellular GJs but also Cx HCs, which might differently affect excitability (Bedner and Steinhäuser, 2013; Steinhäuser et al., 2012). In addition, CBX, the most commonly used inhibitor, has many off-target effects. For example, it inhibits voltage-gated Ca2+ channels (Vessey et al., 2004), purinergic P2X7 receptors (Suadicani et al., 2003), tonic GABAA receptor currents (Ransom et al., 2017), and, accordingly, reduces synaptic transmission and neuronal network activity (Rouach et al., 2003; Tovar et al., 2009), which questions its suitability for assessing the role of gap junctions in epilepsy. Apparently consistent with GJ blockade studies are findings showing that drugs enhancing gap junction communication through intracellular alkalization, such as trimethylamine (TMA), increase epileptiform activity (Köhling et al., 2001; Perez-Velazquez et al., 1994; Vincze et al., 2019). However, TMA increases coupling between neurons (Perez-Velazquez et al., 1994), while evidence for effects on astrocytic coupling has not been provided yet. Indeed, in a recent study, Onodera and colleagues demonstrated that astrocytes react to epileptiform activity with intracellular alkalization, which in turn inhibits GJ coupling. Suppression of alkalization via Na+/HCO3− cotransporter blockade prevented astrocyte uncoupling and neuronal hyperactivity in vitro and in vivo (Onodera et al., 2021). Thus, pharmacological manipulation of GJ coupling provides an inconsistent picture of the role of the astroglial network in the pathophysiology of epilepsy and its suitability as a target for antiepileptogenic drugs. More specific GJ modulators are needed to clarify this issue.
In accordance with the proposed seizure-promoting consequences of HC opening, inhibition of Cx43 HCs with the Cx mimetic peptide TAT-Gap19, which seems to block HC activity without reducing GJ coupling, yielded anticonvulsant effects in mice and rats in the pilocarpine model of epilepsy and in two electrical seizure mouse models (Walrave et al., 2018).
Astroglial glutamate transporter-1 (GLT-1) is the transporter responsible for the majority of glutamate clearance in the central nervous system (CNS). Maintaining low levels of extracellular glutamate is critical to permit efficient and localized synaptic transmission. Mice that lack GLT-1 die from lethal spontaneous seizures at an early age while transgenic mice that overexpress GLT-1 have higher seizure thresholds following convulsant challenge, suggesting that GLT-1 plays an important role in preventing seizures and its function is essential for normal neuronal transmission (Kong et al., 2012; Tanaka et al., 1997). Extracellular glutamate levels in the epileptic hippocampus increase significantly during seizures in patients with epilepsy (During and Spencer, 1993). Basal glutamate levels are also abnormally elevated in patients with medication-refractory epilepsy in the epileptogenic region and lesional cortical sites (Çavuş et al., 2016). These data suggest that glutamate homeostasis is altered in the epileptic brain. Dysregulation in glutamate uptake could be contributing to the excess glutamate observed in these patients and could play a role in the development of epilepsy. GLT-1 protein levels are regulated in preclinical models and patients with TLE. Perisynaptic GLT-1 protein levels have been shown to be downregulated post-SE induce by kainate in a mouse model of TLE (Clarkson et al., 2020; Peterson and Binder, 2019). GLT-1 protein levels have also been found to be downregulated in the epileptic hippocampus of TLE patients with HS (Proper et al., 2002; Sarac et al., 2009). Glutamate aspartate transporter (GLAST), another member of the family of Na+-dependent glutamate transporters, has also been shown to be dysregulated in epileptogenesis. GLAST is found in the dorsal forebrain and is homogeneously distributed in astrocytic soma and endfeet (Schreiner et al., 2014). Homozygous mice deficient in GLAST have significantly longer seizures than wild-type mice following convulsant challenge, suggesting that GLAST also plays a role in seizure susceptibility (Watanabe et al., 1999). GLAST protein levels have also been shown to be downregulated in both a spontaneous epileptic rat genetic model of epilepsy and in TLE patients with HS (Guo et al., 2010; Proper et al., 2002). These data suggest that glutamate uptake capacity is diminished in the epileptic brain.
Glutamate clearance by astrocytes is not only crucial for localized glutamatergic neurotransmission but also important for glutamate-glutamine cycling. Astrocytic glutamine synthetase (GS) is responsible for the rapid conversion of intracellular glutamate to glutamine and is a prerequisite for efficient glutamate clearance from the extracellular space. Rats infused with a nonspecific GS inhibitor, namely methionine sulfoximine, develop recurrent seizures and display neuropathology similar to HS (Wang et al., 2009). Congenital GS deficiency leads to extensive brain malformations, generalized seizures, and early death (Eid et al., 2012). Loss of GS has also been observed in patients with TLE, which could impact glutamate clearance by astrocytic glutamate transporters (Eid et al., 2004).
Targeting glutamate uptake by upregulation of GLT-1 has been shown to have neuroprotective and antiepileptic effects (Peterson and Binder, 2020). It has been hypothesized that ceftriaxone, a β-lactam antibiotic, is a transcriptional activator of GLT-1. Ceftriaxone treatment has been shown to suppress seizures and elevate GS activity and GLT-1 expression in acute phases of epileptogenesis (Hussein et al., 2016; Ramandi et al., 2021). Pharmacological upregulation of GLT-1 using ceftriaxone has also been shown to improve pilocarpine-induced cognitive impairments in epileptic animals (Ramandi et al., 2021). Preventing GLT-1 degradation using pharmacologic inhibition of the heat shock protein Hsp90 suppresses spontaneous recurrent seizures, suggesting that upregulation of GLT-1 in astrocytes may be a potential therapeutic target for the treatment of epilepsy (Sha et al., 2016).
Metabotropic glutamate receptors (mGluRs) are G protein–coupled receptors (GPCRs) expressed throughout the CNS in both neurons and glia that modulate neuronal excitability and synaptic transmission. These receptors can be divided into three separate families: Group I, Group II, and Group III, based on their structure and downstream function (Spampinato et al., 2018). Group I receptors mGluR1 and mGluR5 are expressed by many excitatory neurons while astrocytes dominantly express mGluR5 and mGluR3 receptors, and differential regulation of these glutamate receptors has been observed in epilepsy (Aronica et al., 2000, 2003; Kandratavicius et al., 2014; Peterson and Binder, 2020). Neuronal mGluR5, coupled to Gαq proteins, is important for dendritic morphogenesis and synaptic function/plasticity (Conn and Pin, 1997; She et al., 2009), while astrocytic mGluR5 signaling plays a vital role in astrocytic motility, ensheathment and glutamate transport in the developing brain (Bernardinelli et al., 2014; Devaraju et al., 2013). Acute activation of mGluR5 has been shown to increase astrocytic glutamate uptake, while sustained activation leads to a reduction in astrocytic glutamate transporters and glutamate uptake (Aronica et al., 2003).
Dysregulation of mGluR5 has been observed in preclinical models and in patients with epilepsy. In particular, mGluR5 levels have been shown to be increased in murine seizure models and patients with TLE (Aronica et al., 2000, 2003). mGluR5 expression in reactive astrocytes is persistently upregulated following electrically induced SE (Aronica et al., 2000). Multiple studies have also shown that mGluR5 expression is increased in TLE patients with and without HS (Kandratavicius et al., 2013; Notenboom et al., 2006). Expression of astrocytic mGluR5 is typically limited to the first few weeks of development (Cai et al., 2000), but upregulation of mGluR5 has been observed in surviving neurons and astrocytes in tissue resected from patients with TLE (Aronica et al., 2003; Notenboom et al., 2006). There is also a correlation between mGluR5 expression levels and seizure frequency in patients with TLE. mGluR5 expression levels are negatively correlated with seizure frequency, suggesting that upregulation of mGluR5 may contribute to hyperexcitability in these patients (Kandratavicius et al., 2013).
Group II mGluR3 receptor activation, coupled to Gαi proteins, has been shown to have neuroprotective and neurotrophic effects. Activation of mGluR3 leads to increased glutathione synthesis, increased glutamate uptake, and increased formation of the neuroprotective factor TGFβ (Aronica et al., 2003; Bruno et al., 1998; Zhou et al., 2006). mGluR3 receptors are found in astrocytes and in the presynaptic terminals of glutamatergic neurons (Bruno et al., 2017). Astrocytic mGluR3 expression levels have been shown to be differentially expressed in distinct animal models of TLE. For example, in the kindling model of TLE, astrocyte-specific mGluR3 expression is elevated in CA3 and hilar region as early as 1 week after SE but reduced in the primarily astrocytic layers, molecular layer, and stratum lacunosum moleculare, 3 months after induction of SE (Aronica et al. 2000). In a mouse model of epilepsy, mGluR2/3 expression is also markedly decreased 2 months following pilocarpine-induced SE (Tang et al., 2004). Tang et al. also found that mGluR2/3 expression is reduced in the CA1 stratum lacunosum moleculare, suggesting degenerating mGluR2/3 immunoreactive axons and terminals in patients with mesial TLE (Tang et al., 2004). Interestingly, mGluR2/3 total protein expression examined by Western blot analysis was shown to be increased in temporal lobe resection tissue collected from TLE patients compared to sudden death controls, suggesting that regulation of these receptors, thought to negatively regulate excitatory neurotransmission, might be regionally different (Das et al., 2012).
Current antiseizure drugs (ASDs) work primarily by targeting neurons directly through modulation of ion channels, enhancement of inhibitory neurotransmission, or attenuation of excitatory neurotransmission. Unfortunately, approximately 25% of patients develop refractory epilepsy resistant to currently available ASDs (Xue-Ping et al., 2019). Therefore, non-neuronal targets, including glial cells, are an attractive alternative approach. Since mGluR5 activation increases excitability, inhibiting mGluR5 could be a potential therapeutic target to treat uncontrolled seizures (Table 26–1). Administration of the mGluR5 antagonist MPEP suppressed seizures in the pilocarpine model of TLE (Smolders et al., 2004). MPEP has also been shown to decrease neuronal death when administered after SE (Ding et al., 2007, p. 20007). Chronic administration of the mGluR5 antagonist CTEP reduced seizure frequency and total seizure time in a genetic mouse model of TSC (Kelly et al., 2018). Acute treatment with positive allosteric modulators of mGluR1 and mGluR5 have also been shown to reduce the frequency of spike-and-wave discharges in a rat model of absence epilepsy, suggesting that inhibition of mGluR5 signaling might have antiseizure effects (D’Amore et al., 2014). A Group II mGluR2/3 agonist, R,4R-4-aminopyrrolidine-2,4-dicarboxylate (APDC), dramatically suppressed the incidence of clonic and generalized tonic-clonic seizures following i.c.v. infusion of DL-homocysteic acid (HCA) in a rat model of seizures (Folbergrová et al., 2001), but in the pilocarpine seizure model, APDC could not reduce seizures (Tang et al., 2004). Low-dose DCG IV, a mGluR2/3 agonist, has been shown to reduce seizures, while high-dose DCG IV had pro-epileptic effects in the DL-HCA rat model of seizures (Folbergrová et al., 2001; Smolders et al., 2004). DCG IV also increased seizure threshold in seizure-susceptible animals, demonstrating the anticonvulsant effect of group II mGluR activation (Attwell et al., 1998). These studies indicate that the therapeutic outcome of glutamate receptor agonists/antagonists is highly dependent on multiple factors, including dose, route of administration, and model. Overall, the available preclinical studies suggest that targeting glutamate metabolism and signaling is a promising therapeutic avenue to treat patients with refractory epilepsy.
Positive and Negative Outcomes of Glutamate Transporter Modulation and mGluR Agonists/Antagonists in Preclinical Seizure Models.
It has become increasingly clear that astrocytes play a critical role in the development and progression of epilepsy. Astrocytic gap junction coupling and glutamate uptake is dysregulated in both preclinical models and in patients with epilepsy. Thus, targeting astrocyte gap junctions and glutamate transporters and receptors might be effective as novel antiepileptogenic (disease-modifying prevention of development of epilepsy after epileptogenic insults) and antiseizure (controlling seizures in pharmacoresistant epilepsies) therapies, with fewer side effects compared to traditional suppression of glutamatergic neurotransmission in neurons. Future studies have to reveal further details of the underlying mechanisms.
Work of the authors is supported through grants from the EU (H2020-MSCA-ITN project 722053 EU-GliaPhD; to CS) and BMBF (16GW0182 CONNEXIN, 01DN20001 CONNEX; to CS), and National Institutes of Health (U54HD104461, NS114516, NS111552; to DKB).
The authors declare no relevant conflicts.
This is an open access publication, available online and distributed under the terms of a Creative Commons Attribution-Non Commercial-No Derivatives 4.0 International licence (CC BY-NC-ND 4.0), a copy of which is available at https://creativecommons.org/licenses/by-nc-nd/4.0/. Subject to this license, all rights are reserved.
Your browsing activity is empty.
Activity recording is turned off.
See more...
