Abstract
Commentary
A combination of molecular and physiological experiments reported by Terunuma and colleagues examined the mechanisms underlying this GABAA receptor-deficient state accompanying SE. They used pilocarpine to induce unremitting grade 5 seizures in mice (a commonly used model of SE); they then compared the hippocampi of these animals with those of control animals that received saline and experienced no seizures. The investigators first documented a shift of GABAA receptor subunit composition following 1 hour of SE, with a reduction of beta (β) and gamma (γ) subunits and an increase in delta (8). The functional implication is that synaptic inhibition should be impaired, since the benzodiazepine-sensitive GABAA receptors at inhibitory synapses most typically contain γ subunits and not δ. Indeed, Terunuma et al. nicely documented correlating electrophysiological evidence, revealing a postsynaptic defect of GABAA receptor responses. Additional data revealed that hypoactivity of conventional protein kinase C (PKC) isoforms is responsible for deficient GABAA receptor phosphorylation, resulting in endocytosis of these receptors. PKC normally binds to the β3 subunit of the GABAA receptor and phosphorylates it, thereby preventing its association with the clathrin adaptor AP2—a process critical for receptor internalization. If phosphorylation fails to occur, the result is decreased GABAA receptor expression at the surface that is due to excessive endocytosis and results in defective GABAergic inhibition.
Cellular proteins, be they channels, receptors, or enzymes, are all under constant modulation and regulation by processes that drive their phosphorylation and dephosphorylation. Kinases will phosphorylate proteins by transferring a phosphate group from a nearby molecule of ATP, while phosphatases will remove the phosphate group, thereby dephosphorylating the protein in question. The presence or absence of a phosphate group on a critical amino acid within the protein will alter the protein's configuration and have functional consequences of either enhancing or impairing the activity of the channel, receptor, or enzyme. And, depending on which amino acids within which channels and receptors serve as the target of phosphorylation, the functional impact on the cell could be excitatory or inhibitory. PKC is one of the more prominent kinases, commonly activated by various G-protein–linked metabotropic receptors.
One might expect neuronal excitation underlying seizures to be associated with increased PKC activity. Although most papers indeed report elevated PKC activation with both seizure activity and SE, Terunuma and colleagues describe a decrease in PKC, with no comment addressing the conflicting data. Interpretation of this diminished PKC activation is unclear: it is possible that defective activation of PKC underlies the development or progression of SE; alternatively, prolonged excitation associated with SE may trigger autoregulatory processes, such as desensitization, preventing excessive continued PKC activation. But, let the reader beware: the accurate explanation for the discrepancy most likely lies in the details. For example, PKC responses to the same event may yield different responses among cell types. Guglielmetti et al. reported kainate-induced seizures elicit reduced PKC mRNA in hippocampus but increased levels in dentate gyrus (2). Furthermore, PKC comes in various isoforms, each isoform having distinct physiological roles and potentially responding differently to the same trigger. The current paper by Terunuma et al. focused on the conventional isoforms (i.e., α, β, and γ) and found all three to be reduced in both level and activity in hippocampal slices after 1 hour of SE. By contrast, the reduction of PKC mRNA in CA1 and CA3 that Guglielmetti and coworkers reported was limited to the γ isoform, and lasted 1 to 2 days. Furthermore, the increased PKC levels in the dentate granule cells and their mossy fiber axons were limited to the epsilon (ε) isoform and persisted for months following kainate-induced seizures, which may contribute to the epileptogenic mossy fiber sprouting and synaptic reorganization induced by SE. Additional factors should be considered as well, including: the stage of brain development; brain region or subregion being analyzed; precise timing of analysis relative to seizure onset, length, and termination; and whether the PKC measurement was based on total level versus activated PKC (also called phosphorylated or membrane bound) and protein versus mRNA—all these factors can influence results and need to be carefully assessed to interpret the findings accurately.
The authors boldly conclude their report by suggesting that enhancing phosphorylation of the GABA receptor β3 sub-unit may have therapeutic value in SE; one could achieve this end with agents that activate PKC. Activation of PKC also prevents the induction of group I metabotropic glutamate receptor (mGluR)-driven epileptogenesis in vitro (3), suggesting that agents that activate PKC may be clinically useful in the suppression of both SE and epileptogenesis. Yet, recent data indicate that PKC activation elicits ictal discharges in vitro, sustained in part by enhanced mGluR5 responses (4). PKC-induced phosphorylation has additional seizure-promoting effects: phosphorylation of NMDA receptor subunits promotes trafficking of the NMDA receptors to the cell surface (5); and phosphorylation of presynaptic voltage-gated calcium channels enhances calcium entry, thereby boosting glutamate release (6). Over-activation of PKC may have negative consequences on cognition as well: seizures driven by hyperstimulated NMDA receptors produce sustained PKC elevation associated with reduced hippocampal long-term potentiation, which could account for memory impairment postictally and beyond (7). Furthermore, in prefrontal cortex, increased PKC activation, driven by excessive noradrenergic activation, has been associated with impaired cognitive functioning, with possible relevance to the cognitive dysfunction seen in bipolar disorder and schizophrenia (8).
So, while it is not yet sound to prescribe generalized PKC activators or inhibitors to patients based on any of these data, the article by Terunuma and coworkers opens the possibility that further detailed and carefully analyzed studies along these lines may permit greater knowledge; similarly, technological advances may open the way for more targeted approaches to intervening in PKC-relevant pathways. For now, there is probably but one antiepileptic agent currently in use that works in part by interfering with PKC: the broad-spectrum anticonvulsant valproate, which has been reported to have a modest inhibitory effect on PKCε. This action has been credited with conferring part of the anticonvulsive and mood-stabilizing efficacy of valproate (9), and if the sustained elevation of PKCε in the dentate is indeed necessary for SE-induced mossy fiber sprouting (2), it suggests a possibility that valproate may suppress epileptogenesis. While generalized PKC activation may be ictogenic (4), targeted localized PKC activation driven by selective activation of a specific mGluR subtype may have antiepileptogenic potential as well (3). Additional data are necessary to validate these suppositions, but it certainly encourages continued pursuit in this area of research to tease apart the tangled details relating PKC to epilepsy.