One Site to Rule Them All: Toward a Master Regulator of Ictal Activity

Commentary

Optogenetic Activation of Superior Colliculus Neurons Suppresses Seizures Originating in Diverse Brain Networks.

Soper C, Wicker E, Kulick CV, N'Gouemo P, Forcelli PA. Neurobiol Dis 2016;87:102–115

Because sites of seizure origin may be unknown or multifocal, identifying targets from which activation can suppress seizures originating in diverse networks is essential. We evaluated the ability of optogenetic activation of the deep/intermediate layers of the superior colliculus (DLSC) to fill this role. Optogenetic activation of DLSC suppressed behavioral and electrographic seizures in the pentylenetetrazole (forebrain+brainstem seizures) and Area Tempestas (forebrain/complex partial seizures) models; this effect was specific to activation of DLSC, and not neighboring structures. DLSC activation likewise attenuated seizures evoked by gamma butyrolactone (thalamocortical/absence seizures), or acoustic stimulation of genetically epilepsy prone rates (brainstem seizures). Anticonvulsant effects were seen with stimulation frequencies as low as 5 Hz. Unlike previous applications of optogenetics for the control of seizures, activation of DLSC exerted broad-spectrum anticonvulsant actions, attenuating seizures originating in diverse and distal brain networks. These data indicate that DLSC is a promising target for optogenetic control of epilepsy.

Ictogenic diversity presents a puzzle to seizure intervention. Stimulation of certain sites can produce positive therapeutic outcomes across patients with different ictogenic loci. For example, a recent report of patient outcomes receiving anterior thalamic stimulation had encouraging results toward controlling a variety of seizure types (though not all patients responded to treatment) (1). This effect across multiple seizure types raises the enticing prospect of a master regulator of ictal activity—a brain region with privileged connectivity capable of resetting pathological activity in any type of epilepsy. While this extreme outcome (one site for all seizures) is unlikely, identifying areas with broad enough applicability that they could be used to intervene in cases with multiple, diffuse, unknown, or progressing seizure foci would be clinically advantageous. Clinical seizure intervention currently relies on electrical stimulation or pharmacology; however, experimentally, optogenetics is a powerful research tool: genetic markers drive expression of light-activated proteins capable of depolarizing or hyperpolarizing cells, allowing bidirectional, cell-type specific, rapid control of neural activity. Optogenetic manipulation of cells within the ictogenic locus and, importantly, outside the locus, has been reported to inhibit seizures (2–5). For example, stimulation of the cerebellum has been separately reported to disrupt both temporal lobe (5) and thalamocortical (absence) (6) seizures. These findings that optogenetic stimulation of the cerebellum inhibits seizures in models of two distinct types of epilepsy suggest that the cerebellum may function as a master regulator of ictal activity.

One efferent target of the cerebellum is the superior colliculus (SC), which Soper and colleagues investigated as another potential master regulator. In a methodical series of experiments, the researchers addressed the therapeutic reach of optogenetic seizure intervention targeting the deep and intermediate layers of the superior colliculus (DLSC). The authors injected a neuron-specific virus carrying the light-activated, excitatory protein channelrhodopsin to control DLSC neurons, subsequently testing the anti-ictal potential of DLSC activation in four seizure models: systemic pentylenetetrazole (PTZ), systemic gamma butyrolactone, focal injection of bicuculline methiodide into piriform cortex (Area Tempestas; AT), and genetically seizure-prone GEPR-3 rats. While pharmacological and electrical seizure intervention of the DLSC has been previously examined, Soper and colleagues extended these findings by comparing DLSC activation among multiple models and also by conducting these experiments with specificity unachievable with previous technologies. The increased temporal specificity allowed for examination of stimulation parameters beyond that of pharmaceutical intervention, while the spatial specificity targeted a more defined population of cells than could electrical stimulation. Illustrating its potential as a master regulator, DLSC activation achieved seizure inhibition in all models tested. Importantly, while DLSC neurons—similar to cerebellar neurons—are key regulators of motor output, observed seizure inhibition included reducing electrographic seizures, indicating that DLSC-targeted intervention does not simply attenuate the behavioral manifestations of seizures but broadly reduces ictal activity.

PTZ-evoked seizures engage a diverse network, including thalamocortical, forebrain, and hindbrain areas, that results in tonic–clonic activity. Optogenetic stimulation of DLSC neurons began immediately after PTZ administration and resulted in a dramatic reduction of both behavioral and electrographic seizures. In a model of absence seizures, induced by systemic gamma butyrolactone and characterized by generalized spike-and-wave discharges (GSWDs), DLSC stimulation reduced the number and duration of electrographically recorded GSWDs. Similarly, in the AT forebrain limbic seizure model (induced by focal injections, in contrast to the systemic administration of PTZ and gamma butyrolactone), seizures also benefitted from DLSC stimulation, reducing electrographic seizures as well as the frequency and severity of behavioral seizures. In genetically epilepsy-prone GEPR-3 rats that exhibit behavioral seizures in response to acoustic stimulation, DLSC activation reduced behavioral seizure severity, duration, and increased the latency to seizure onset. These four seizure models were well chosen to represent ictogenic diversity: behavioral and electrographic seizures from diffuse and focal chemoconvulsant activity, pharmacologically induced absence seizures, and a genetic model of epilepsy. The same intervention proved effective for these four distinct seizure models, which speaks to the therapeutic potential of modulating a single, common network with widespread influence.

Two important findings in particular highlight key considerations of DLSC optogenetic seizure intervention: In the PTZ model, postmortem analysis revealed that a subset of rats had the virus and fiber optic cable misplaced into the inferior colliculus (IC). In these rats, stimulation either worsened or had no effect on seizure outcome, emphasizing the importance of site specificity in seizure intervention. Indeed, even within the superior colliculus (SC), there is important ictal-related heterogeneity. Different areas of the SC are involved with distinct functions and exhibit heterogeneous projections to downstream networks. The DLSC neurons targeted in this study are a portion of a larger dorsal midbrain anticonvulsant zone that has been previously described to have desynchronizing effects on distal cortical activity and multiple possible routes by which to influence seizure activity. For example, DLSC neurons are interconnected with the substantia nigra and have an established anticonvulsant role via this pathway during electroshock seizures (7). Additionally, DLSC neurons project to areas of the thalamus associated with seizure control (8, 9). It is highly possible that different projections contribute differently to the therapeutic effects in each seizure model. Untangling specific network contributions is an important next step in the development of this therapy. Implementing intersectional targeting strategies that allow stimulation of spatially restricted, cell-type, and projection-specific neurons may provide insight to therapeutic contributions of particular networks.

Secondly, in the focal AT seizure model, the researchers conducted a revealing experiment in which they illuminated the issue of seizure detection and stimulation timing. In contrast to the interventions examined by Soper and colleagues described earlier, which commenced before seizure occurrence, optogenetic activation of DLSC neurons after behavioral seizure onset provided no therapeutic relief, emphasizing previous reports of effects regarding the timing of seizure intervention (3, 10). This experiment demonstrates the importance of timing of seizure intervention relative to seizure onset, highlighting an ictal-vulnerable period before seizure activity exceeds therapeutic reach. Utilizing an on-demand, automated seizure detection and intervention system (3) could elaborate these DLSC studies into epileptic models with spontaneously occurring seizures.

Broadly, Soper and colleagues' findings continue an exploration of the multifaceted therapeutic potential of optogenetic seizure intervention. While additional challenges remain before an optogenetic approach could enter clinical applications (such as safe and effective opsin expression in humans), the authors' results further the exciting prospect that targeting certain networks will yield benefits for diverse epilepsies. While isolating a single master regulator of all seizure activity may be overly ambitious, defining the reaches of site-specific stimulation will allow further investigation into mechanisms and network contributions, representing an important step forward in understanding and applying seizure intervention technology.