Abstract
CaV 2.1 Ablation in Cortical Interneurons Selectively Impairs Fast-Spiking Basket Cells and Causes Generalized Seizures
Rossignol E, Kruglikov I, van den Maagdenberg AM, Rudy B, Fishell G. Ann Neurol 2013;74:209–222.
OBJECTIVE: Both the neuronal populations and mechanisms responsible for generalized spike-wave absence seizures are poorly understood. In mutant mice carrying loss-of-function (LOF) mutations in Cacna1a, which encodes the a1 pore-forming subunit of CaV 2.1 (P/Q-type) voltage-gated Ca2+ channels, generalized spike-wave seizures have been suggested to result from excessive bursting of thalamocortical cells. However, other cellular populations including cortical inhibitory interneurons may contribute to this phenotype. We investigated how different cortical interneuron subtypes are affected by the loss of CaV 2.1 channel function and how this contributes to the onset of generalized epilepsy. METHODS: We designed genetic strategies to induce a selective Cacna1a LOF mutation in different cortical γ-aminobutyric acidergic (GABAergic) and/or glutamatergic neuronal populations in mice. We assessed the cellular and network consequences of these mutations by combining immunohistochemical assays, in vitro physiology, optogenetics, and in vivo video electroencephalographic recordings. RESULTS: We demonstrate that selective Cacna1a LOF from a subset of cortical interneurons, including parvalbumin (PV)+ and somatostatin (SST)+ interneurons, results in severe generalized epilepsy. Loss of CaV 2.1 channel function compromises GABA release from PV+ but not SST+ interneurons. Moreover, thalamocortical projection neurons do not show enhanced bursting in these mutants, suggesting that this feature is not essential for the development of generalized spike-wave seizures. Notably, the concurrent removal of CaV 2.1 channels in cortical pyramidal cells and interneurons considerably lessens seizure severity by decreasing cortical excitability. INTERPRETATION: Our findings demonstrate that conditional ablation of CaV 2.1 channel function from cortical PV+ interneurons alters GABA release from these cells, impairs their ability to constrain cortical pyramidal cell excitability, and is sufficient to cause generalized seizures.
Commentary
Absence seizures and other types of generalized seizures are characteristic of a number of classic epilepsy syndromes, such as childhood absence epilepsy and juvenile myoclonic epilepsy. Major advances in understanding the underlying neuro-anatomy, pathophysiology, and molecular genetics of absence epilepsies have been made through decades of extensive clinical and basic research. However, significant unknowns and controversies about the pathogenesis of generalized seizures still exist, especially regarding the mechanisms of initiation of the seizures. A variety of rational candidates and hypotheses have been advanced in identifying the primary anatomic site, cell populations, and molecular mediators responsible for generating these seizures, but a unifying mechanism has not been established.
The principal neuroanatomical networks mediating absence seizures are thought to involve reciprocal circuits between the thalamus and cortex. However, where within these circuits absence seizures are initially generated has been extensively debated, with different theories vacillating over time between cortex and thalamus (1). Early ideas, dating back to Penfield and Jasper in the mid-1900s, promoted the “centren-cephalic” hypothesis that the thalamus could drive widespread cortical spike and wave discharges. The primary role of the thalamus was later supported by a variety of experimental studies implicating specific thalamic nuclei in generating both normal physiological rhythms, such as sleep spindles, and pathological oscillations of absence seizures (2). In contrast, other experimental evidence suggested that the cortex can be the primary initiator of generalized spikes and absence seizures. Combined cortical and subcortical depth electrode recordings indicated an earlier onset of spikes within the cortex in primate models of generalized epilepsy (1). More detailed quantitative analysis of cortical and thalamic recordings in a rodent genetic model of absence seizures has identified a focal region of somatosensory cortex as being the apparent trigger of generalized spike and wave activity (3). Finally, the “cortico-reticular” concept of absence seizures incorporated aspects of both cortical and thalamic mechanisms with recurrent oscillations between these two regions, collectively forming the basis of absence seizures but with possible initiation in cortex.
In dissecting the critical thalamocortical networks in more detail, the role of specific cell types, neurotransmitter systems, and ion channels in generating oscillatory physiological activity and generalized spike and wave discharges have been investigated. Long-range reciprocal connections between excitatory thalamocortical relay neurons in thalamic nuclei and widespread cortical pyramidal neurons projecting back to the thalamus constitute the primary circuit (2). Within the thalamus, additional local reciprocal connections between inhibitory GABAergic interneurons in the reticular nucleus and thalamocortical relay neurons may form a core unit capable of generating oscillations. In particular, cyclical bursting behavior in thalamocortical relay neurons may be triggered by low-voltage-activated T-type calcium channels in response to a rebound of the membrane potential following GABA interneuron-induced hyperpolarization. While intrinsic thalamic networks are firmly established as causing oscillatory, bursting patterns that promote normal sleep spindle activity, whether similar mechanisms also cause generalized spike and wave activity of absence seizures is hypothesized but not definitely proven. An increase in T-type calcium currents and tonic GABA inhibition in thalamic neurons associated with abnormal bursting in the thalamus has been detected in animal models as a potential cause of absence seizures (4, 5). However, even if thalamic-mediated bursting and oscillations promote absence seizures, it is still possible that a cortical source provides the initial stimulus triggering these oscillations.
Compared with the thalamus, local circuits and mechanisms in the cortex that may generate corticothalamic oscillations are not as well defined. In addition to electrophysiological evidence for an earlier onset of generalized spike-and-wave activity in focal areas of the cortex, some mouse models of absence epilepsy exhibit reduced cortical inhibition, suggesting a defect in GABAergic interneurons (6). Furthermore, the identification of mutations in GABA receptors in human genetic absence epilepsies emphasizes the clinical relevance of GABAergic systems for generalized seizures (7, 8). With this background, the recent study by Rossignol and colleagues directly addresses the hypothesis that impairment of GABAergic interneurons in the cortex can lead to absence and other types of generalized seizures. They use targeted genetic techniques to inactivate the Cav2.1 P/Q-type calcium channel from specific populations of cortical GABAergic interneurons in transgenic mice. In comparison with the T-type calcium channel involved in generating oscillations, the P/Q-type channels are high-voltage gated calcium channels that primarily mediate presynaptic vesicular neurotransmitter release. In addition to localization within cortical neurons, Cav2.1 P/Q-type channels are abundantly found in cerebellar Purkinje neurons. As a result, patients and mouse models with Cav2.1 mutations exhibit ataxia, as well as absence seizures (9, 10). The study by Rossignol and colleagues bypasses the cerebellar involvement and focuses on cortical networks by utilizing advanced genetic inactivation strategies that specifically target cortical interneurons.
The primary finding of this study that mice with loss of Cav2.1 channel function in cortical interneurons develop absence and other generalized seizures convincingly establishes that cortical interneuron dysfunction is sufficient to cause generalized seizures. In fact, one specific population of cortical GABAergic interneurons—parvalbumin-positive fast-spiking neurons—was implicated in causing generalized seizures. These fast-spiking interneurons exhibited impaired GABA release leading to disinhibition of cortical pyramidal neurons in these mice, whereas another major subtype of cortical interneuron was unaffected. In addition, there was no evidence of abnormal bursting of thalamocortical neurons, further emphasizing the primary importance of cortex over thalamus in this model.
Although this study provides compelling evidence implicating cortical interneurons in generalized seizures, there are some caveats: Interestingly, the mouse model exhibited not only absence seizures but also a variety of other generalized seizure types, including myoclonic, tonic, and tonic–clonic seizures. Thus, this is not a pure model of absence epilepsy and may not involve the same thalamocortical mechanisms that have been implicated in many other models primarily involving absence seizures, as indicated by the lack of abnormal bursting of thalamocortical neurons in the present model. In fact, this study should prompt deeper investigation of the role of cortical interneurons in contributing to absence seizures in other animal models that have focused more on thalamocortical oscillations.
Ultimately, patients with absence seizures represent a heterogeneous group, with a variety of underlying genetic etiologies and associated epilepsy syndromes. Attempting to find unifying mechanisms for generalized seizures is a worthwhile but perhaps unrealistic goal. Between different patients and even within the same patient, it is likely that multiple mechanisms are differentially involved in various situations in generating generalized spike-and-wave discharges and generalized seizures. Multiple hubs within a diffuse, distributed network—including both the cortex, thalamus, and other subcortical structures—may be capable of triggering seizures within the network at any given time. The work by Rossignol and colleagues highlights the importance of the cortical interneuron and suggests potential therapeutic strategies for targeting this focal point.