The ionotropic glutamate receptor (iGluR) family is divided into four distinct subgroups based on receptor pharmacology and structural homology, including the AMPA receptors (GluA1–GluA4), kainate receptors (GluK1–GluK5), NMDA receptors (GluN1, GluN2A–GluN2D, GluN3A, and GluN3B), and δ receptors (GluD1 and GluD2) (Dingledine et al., 1999). The iGluRs are tetramers with a binding site for glutamate on each subunit that assemble as dimer of dimers, and their composition can be homomeric or heteromeric (Traynelis et al., 2010).
AMPA receptors. The majority of fast excitatory synaptic neurotransmission in the central nervous system (CNS) is mediated by glutamate activation of AMPA receptors (Dingledine et al., 1999). In addition to initiating neuronal firing, AMPA receptors also underlie aspects of synaptic plasticity (Dingledine et al., 1999), such as such as long-term potentiation (LTP) (Selcher et al., 2012). AMPA receptors are tetrameric assemblies of GluA1 to GluA4 subunits (Traynelis et al., 2010). Although homomeric receptors are functional, native AMPA receptors are believed to be heteromers. For example, in hippocampal pyramidal cells of mature rats, the most common subunit configurations are GluA1/GluA2 and GluA2/GluA3 (Dingledine et al., 1999; Traynelis et al., 2010). The permeability to Ca2+ ions of AMPA receptors is governed by the GluA2 subunit. The presence of a GluA2 subunit renders the channel impermeable to Ca2+ due to the posttranslational mRNA editing of a glutamine (Q) to an arginine (R) residue (Seeburg & Hartner, 2003). The majority of GluA2 subunits in the CNS are edited to the GluA2(R) form, thus the principal ions gated by these AMPA receptors are sodium (Na+) and potassium (K+) (Dingledine et al., 1999). Conversely, AMPA receptors lacking GluA2- or containing an unedited (Q form) of GluA2 are rendered permeable to Ca2+ ions (Sommer et al., 1991). This editing is catalyzed by the enzyme adenosine deaminase acting on RNA (ADAR2). This enzyme has also shown to edit the Kv1.1 potassium channel, conferring a loss in 4-aminopyridine sensitivity to kainic acid–induced seizures (Streit et al., 2011). The imbalance between AMPA receptors and CP-AMPA receptors has been related to epileptogenesis (Tanaka et al., 2000), the processes by which a normal brain becomes epileptic (McNamara et al., 2006). Krestel et al. (2004) showed in adult rat brains that seizure susceptibility is increased by the presence of the CP-AMPAR (GluA2[Q] receptors), and that these types of receptors play a role in circuit hyperexcitability. In addition, susceptibility to hypoxia-induced seizures occurs in the brain during developmental stages when there is an increased expression of CP-AMPA receptors. Consequently, perinatal hypoxia-induced seizures increase the expression of CP-AMPA receptors and the capacity for an AMPA receptor–mediated epileptogenesis (Sanchez et al., 2001). Moreover, studies have found a reduction in the GluA2 subunit in the amygdala, piriform cortex, and limbic forebrain in rats after amygdala kindling (Prince et al., 1995, 2000). Furthermore, decreased mRNA expression of GluA2 was found in the hippocampal CA1 region and dentate gyrus in rats experiencing pentylenetetrazol (PTZ) –induced seizures (Ekonomou et al., 2001). In the pilocarpine model of TLE, increased levels of CP-AMPAR were observed in the hippocampal membrane, which can contribute to the elevation of intracellular Ca2+ during recurrent burst firing. In addition, AMPAR–mediated currents became inwardly rectifying in the hippocampal CA1 neurons and dentate granule cells (Rajasekaran et al., 2012). Moreover, the R/G editing, another posttranslational modification, of the GluA2 subunit was found to be increased in the hippocampus and temporal cortex from patients with refractory epilepsy. The increased editing at the R/G site in the hippocampal tissue of epilepsy patients may enhance responses to glutamate (Vollmar et al., 2004). Furthermore, a mutation in the Gria4 gene, which codes for the GluA4 subunit, has been identified in C3H/HeJ mice, which exhibit spontaneous absence seizures (Beyer et al., 2008). This mutation results in decreased expression of GluA4, which is one of the main AMPA-receptor subunits expressed in the reticular thalamic nucleus (nRT), and the predominant subunit in corticothalamic neurons (Mineff & Weinberg, 2000). GluA4-containing receptors have the fastest desensitization rate to glutamate (Mosbacher et al., 1994). Therefore, reduced expression of GluA4 can increase the duration of response to excitatory input by glutamate. This event could in turn promote increased burst firing in reticular neurons enhancing circuit synchrony (Meeren et al., 2002).
AMPA receptors can interact with a variety of proteins in the postsynaptic membrane that function as auxiliary subunits and can modify their properties. The majority of these interactions occur through the cytoplasmic C-terminal tail of AMPA receptors, with proteins containing postsynaptic density-95 (PSD-95), discs large, zonula occludens (PDZ) domains, including glutamate receptor interacting protein (GRIP), AMPA-receptor binding protein (ABP), protein interacting with C-kinase 1 (PICK1), and synapse-associated protein of 97 kDa (SAP-97), or via non-PDZ domains (Srivastava et al., 1998; Dev et al., 1999; Kim & Huganir, 1999; Xia et al., 1999). Proteins that contain a PDZ domain that have been shown to interact with AMPA receptors include the transmembrane AMPA receptor regulatory proteins (TARPs; Chen et al., 2000), cornichon homologs (CNIH-2, CNIH-3; Schwenk et al., 2009), synapse differentially induced gene 1 (SynDIG1; Kalashnikova et al., 2010), and cystine-knot AMPAR-modulating protein (CKAMP44; von Engelhardt et al., 2010). These proteins differentially regulate AMPA-receptor channel gating and are involved in subunit folding, assembly, surface expression, and clustering and anchoring of AMPA receptors at synapses (Diaz, 2010; Jackson & Nicoll, 2011).
Transmembrane AMPA receptor regulatory proteins: TARPs are a family of proteins— including stargazin (γ2), γ3, γ 4, γ5, γ7, and γ8—with distinct and complementary expression patterns in both neurons and glia in the developing and mature CNS (Tomita et al., 2003; Kato et al., 2010). Stargazin (γ2) was the first TARP identified when a mutation in its gene (Cacng2) was found to cause the stargazer mouse, which manifests spontaneous absence-like seizures with generalized spike-and-wave discharges as well as having cerebellar ataxia (Letts et al., 1998). Stargazin was found to have structural homology to the γ1 subunit of the skeletal muscle voltage-gated Ca2+ channels (VGCCs) (Letts et al., 1998). Structurally, TARPs comprise four transmembrane domains and cytosolic amino- and carboxy-termini containing the PDZ domain. Despite the structural similarity with γ1, the TARPS are not expressed on the skeletal muscle and have minor or no effect on VGCCs (Tomita et al., 2003; Fukaya et al., 2005).
The TARPs regulate many different characteristics of the AMPA receptors such as AMPAR biogenesis, trafficking, anchoring AMPARs at the synapse, as well as modulating the channel kinetics. In the endoplasmic reticulum (ER), AMPA receptors are assembled by the formation of dimers and tetramers. TARPs associate with the tetrameric AMPAR to act as an auxiliary subunit permitting the efficient export of the AMPAR from the ER to the Golgi. Then, nPIST (a Golgi-enriched protein involved with trafficking of transmembrane proteins) binds to the C-terminal tail of stargazin in Golgi and assists the AMPAR–TARP complex to exit the Golgi and traffic it to the cell surface (Chen et al., 2000; Ziff, 2007). The AMPAR-TARP complex diffuses into the postsynaptic density PSD, where PSD-95 binds to the PDZ domain of the C-terminal tail of the TARP to anchor the complex at the synapse (Choi et al., 2002; Tomita et al., 2003, 2004; Ziff, 2007). Phosphorylation of TARPs by protein kinase C (PKC) and CaMKII is important in mediating AMPA-receptor synaptic transmission (Inamura et al., 2006). For example, stargazin phosphorylation increases AMPA-receptor synaptic trafficking, thereby establishing stargazin as critical in controlling synaptic strength (Tomita et al., 2005). Furthermore, TARPs can modify synaptic plasticity by affecting AMPA-receptor biophysical properties such as reduced receptor desensitization, slowed receptor deactivation rates, and increased recovery for desensitization (Priel et al., 2005).
Stargazer mice, which harbor a mutation in the stargazin gene that results in a decrease in expression of the stargazin protein, lack functional AMPA receptors in the cerebellum (Letts et al., 1998). In the thalamocortical synapses, AMPA receptors are reduced in the nRT, due to a reduction of GluA4 and GluA2/3 subunit expression (Barad et al., 2012). In contrast, the levels of AMPA receptors in the cortex remain unchanged, probably because stargazin is not the predominant TARP expressed in the cerebral cortex and the possible compensatory effect of the other TARPs expressed in the cortex (Tomita et al., 2003; Fukaya et al., 2005). In the Genetic Absence Epilepsy Rats from Strasbourg (GAERS), a model of genetic generalized epilepsy that manifests spontaneous absence seizures associated with generalized spike-and-wave discharges on electroencephalography (EEG), without any other neurologic deficit, stargazin expression is increased in the somatosensory cortex and thalamus (Powell et al., 2008). This increase in stargazin expression was also associated with an increase in the cortical membrane expression of AMPA-receptor subunits GluA1 and GluA2 (Kennard et al., 2011). The findings in these two rodent models illustrate that either a decrease or an increase in stargazin expression can be associated with an absence seizures phenotype, highlighting that perturbations in AMPA-mediated synaptic transmission, either positively or negatively, can result in an epileptic network.
Drugs that modulate AMPA receptors: Drugs that can inhibit AMPA-receptor activity have the potential to reduce excessive excitatory responses and may be promising future antiepileptic drugs (Meldrum & Rogawski, 2007; Rogawski, 2011). Perampanel, a noncompetitive and highly selective AMPA-receptor antagonist, is currently in phase III development as an adjunctive treatment for drug-resistant partial-onset seizures (Krauss et al., 2012). Parampanel also decreases intracellular Ca2+ concentration induced by AMPA-receptor activation, which would have the net effect of decreasing excitability (Hanada et al., 2011; Ceolin et al., 2012). Perampanel has been shown to have antiseizure effects in a broad range of rodent models, but showed no effect in the GAERS model, suggesting that it is less likely to be effective clinically against absence seizures (Hanada et al., 2011).
N-methyl-d-aspartate receptors. NMDA receptors play a crucial role in excitatory neurotransmission regulation in the CNS. NMDA receptors are cationic channels permeable to Na+, K+, and Ca2+ (Perez-Otano & Ehlers, 2005). The Ca2+ influx through NMDA receptors is the critical factor that mediates many of its roles in health and disease (Cull-Candy & Leszkiewicz, 2004). The NMDA receptors are tetramers composed of different subunits: GluN1, GluN2A–GluN2D, and GluN3A–GluN3B (Traynelis et al., 2010). GluN1 is a necessary subunit of all NMDARs; genetic deletion of this subunit causes death in neonatal stages (Forrest et al., 1994). The C-terminus region of GluN1 regulates NMDAR trafficking and binding to proteins, including calmodulin, CaMKII, yotiao, alpha-actinin, tubulin, neurofilaments, and downstream regulatory element antagonist modulator (DREAM; Cull-Candy & Leszkiewicz, 2004; Horak & Wenthold, 2009).
Both GluN2A and GluN2B subunits are highly expressed in the cortex and hippocampus, where they are reported to control synaptic plasticity and metaplasticity (Monyer et al., 1994). Mutations in the GRIN2A gene, which codes for the human GluN2A, have been identified in patients with idiopathic epilepsy (Endele et al., 2010). Moreover, haplotypes of GRIN1, the gene encoding for the human GluN1 subunit, have been associated with infantile spasms, further indicating an involvement of NMDARs in epilepsy (Ding et al., 2010). Decreased GluN2B mRNA expression has been found in pyramidal neurons of TLE patients with hippocampal sclerosis, whereas an up-regulation of this subunit was found in pyramidal cells of nonsclerotic hippocampi from epileptic patients (Mathern et al., 1998).
Drugs that alter NMDA receptor properties: Felbamate was the first new-generation antiepileptic drug introduced in the 1990s, and it is highly effective against a broad range of seizure types (Sachdeo et al., 1992), but has had its clinical use severely limited following the discovery that it rarely results in the potential fatal adverse effects of aplastic anemia and liver failure (Pellock, 1999). Felbamate inhibits NMDA-evoked responses, preferentially in NMDA receptors containing GluN2B subunits, as well as potentiates GABA-evoked responses (Sofia, 1994).
Remacemide is a low-affinity noncompetitive antagonist of NMDA receptors (Subramaniam et al., 1996; Norris & King, 1997) but has also been shown to inhibit sustained repetitive firing in cultured neurons by blocking voltage-activated Na+ channels (Garske et al., 1991). Remacemide is effective against a broad range of animal seizure models (Palmer et al., 1991). However, clinical trials of remacemide have overall shown relatively disappointing efficacy (Cramer et al., 1994; Brodie et al., 2002; Wesnes et al., 2009). Other potential new antiepileptic drugs in development that have been found to have effects in at least part by modulating NMDA-receptor action include losigamone (Srinivasan et al., 1997), which also reduces potassium-evoked release of glutamate and aspartate from cortical slices (Draguhn et al., 1997) and enhances GABAergic synaptic transmission (Dimpfel et al., 1995). Huperzine A, which shows protective activity against partial seizures in animal models (Schneider et al., 2009), is an NMDA-receptor antagonist (Zhang et al., 2002) and can reduce glutamate-induced toxicity (Ved et al., 1997).
Kainate receptors. Kainate receptors (KARs) are members of the iGluR receptor family that show distinctive characteristics. Unlike AMPA and NMDA receptors, KARs are not predominantly found in excitatory postsynaptic complexes (Contractor & Swanson, 2008). Instead, KARs act as modulators of synaptic transmission and neuronal excitability. Noticeably, they link to metabotropic (G-protein–mediated) signaling pathways in addition to operating as ionotropic receptors (Rodriguez-Moreno & Lerma, 1998). KARs are distributed throughout the brain pre- and postsynaptically and are involved in the regulation of activity of synaptic networks by postsynaptic depolarization at a subset of excitatory synapses, in presynaptic regulation of neurotransmitter release, in presynaptic modulation of both excitatory and inhibitory transmission, in refinement of synaptic strength during development, and in enhancement of neuronal excitability (Contractor et al., 2011). The KARs are tetramers formed by combinations of low-affinity, GluK1–GluK3, and high affinity, GluK4–GluK5 subunits (Collingridge et al., 2009). All subunits contribute to the formation of the ionic pore and each one contains a glutamate-binding site (Fisher & Mott, 2011).
It is notable that in a rat model of TLE, aberrant synaptic kainite receptors dramatically expand the temporal window for synaptic integration in dentate granular cells. This introduces a change in the input–output operation of dentate granule cell that switches their firing from a sparse to an abnormal sustained and rhythmic mode (Artinian et al., 2011). Moreover, in the entorhinal cortex, both GluK1 and GluK2 have been implicated in regulating the network oscillations and enhancing both glutamate and GABA release in this critical area for epileptogenesis in mesial TLE (Bartolomei et al., 2005; Chamberlain et al., 2012).
Drugs that modulate kainate receptors: There are no antiepileptic drugs in clinical practice, or in advanced clinical development, that have been shown to have effects primarily on kainate receptors. However, the experimental compound, LY293558, which inhibits GluK1-containing kainate receptors, has been shown to be effective in reducing the duration of SE in rats (Figueiredo et al., 2011). This drug is a nonselective antagonist, having effects on kainate and AMPA receptors (Alt et al., 2006; Jane et al., 2009).