Abstract
Hyder SK, Lazarini-Lopes W, Toib J, Williams G, Sukharev A, Forcelli PA. Proc Natl Acad Sci U S A. 2025;122(14):e2419178122. doi:10.1073/pnas.2419178122. PMID: 40163720. Despite a century of development of antiseizure medications, up to a third of people with epilepsy do not achieve seizure freedom with drug therapy. Deep brain stimulation is of growing use, but just as with pharmacotherapy, is not universally effective. Identifying new targets for deep brain stimulation—and in particular sites that are effective against a range of seizure types—may close this gap. Engagement of the basal ganglia experimental seizures was first observed almost 75 y ago. However, the role of the basal ganglia's input nucleus, the striatum, in seizure control is relatively understudied. To address this gap, we used an optogenetic approach to activate and inactivate neurons in the dorsal striatum of rats submitted to the gamma-butyrolactone (GBL) model of absence epilepsy, amygdala kindling model of temporal lobe epilepsy, and pilocarpine-induced Status Epilepticus (SE). Open-loop (continuous light delivery) optogenetic activation of dorsal striatal neurons robustly suppressed seizures in all models. By contrast, open-loop optogenetic silencing increased absence seizure expression and facilitated SE onset but had no effect on kindled seizures. In the GBL model, we also tested the effects of closed-loop modulation (light delivery in response to seizure detection). Closed-loop activation reduced duration of spike-wave discharges (SWDs), while closed-loop inhibition increased SWD duration. These results demonstrated previously unrecognized antiabsence effects associated with striatal neuromodulation. These findings demonstrate a robust, bidirectional role of the dorsal striatum in the control of multiple seizure types, suggesting that the striatum is a site that can exert broad-spectrum control of seizures.
Commentary
Brain stimulation is increasingly used as a palliative option for people with refractory epilepsy. 1 Stimulation targets include the anterior and centromedian nucleus of the thalamus, the hippocampus, and the cortex. These targets are often specific to the type of seizures and seizure-onset zone. While there is good evidence that neurostimulation can lead to seizure reductions and, in rare cases, seizure freedom, too many people continue to have seizures after implantation and parameter optimization.
Many targets have been studied in the past, each with its own levels of evidence and associated risks and adverse effects. One region that has received less clinical attention is the basal ganglia, despite its recognized role in seizure control for several decades. Early work from Gale and colleagues showed that enhancing GABAA receptor-mediated inhibition in the substantia nigra pars reticulata can strongly suppress seizures across multiple models, helping establish the basal ganglia as a seizure-control circuit. The striatum, the principal input nucleus of the basal ganglia, has remained relatively unexplored as a therapeutic target. Although it does not project directly to the cortex, it is embedded within the basal ganglia-thalamo-cortical network, 2 well placed to influence wide-ranging cortical networks. Direct evidence for the striatum's anti-seizure potential comes from studies demonstrating that intrastriatal injections of the dopamine agonist apomorphine protects against pilocarpine-induced seizures, 3 suggesting that dopaminergic modulation of direct and indirect pathways within the striatum may shift seizure threshold. Furthermore, a GABA antagonist (bicuculline), which typically provokes seizures when applied cortically, paradoxically demonstrated anticonvulsant effects when injected into the striatum, 4 further suggesting that the striatum may play an active role in seizure control.
The striatum is one of the largest subcortical structures, sitting at the interface between widespread cortical inputs and deep basal ganglia outputs. Targeting such a network hub could engage broader anti-seizure network effects across brain regions and seizure types. This network-targeted approach is supported by work identifying dynamic driver hubs for seizure generation and termination in human neocortical epilepsy, 5 and by the growing use of neural stimulation to target thalamic hubs such as the anterior and centromedian nuclei. 6
For these reasons, Hyder et al 6 applied bidirectional optogenetic manipulations to test how activating or silencing dorsal striatal neurons influences seizures across epilepsy models. This work builds on the group's prior studies showing that stimulation of midbrain output structures such as the superior colliculus can potently suppress absence seizures. 7 More broadly, preclinical work supports neuromodulation of basal ganglia-brainstem circuits as an effective seizure-control strategy, 8 but clinical translation of this finding has been limited by surgical risks and technical challenges in targeting brainstem and midbrain nuclei. The dorsal striatum provides a more surgically accessible upstream node in this network, serving as a regulator for globus pallidus and substantia nigra output pathways that project to the thalamus and the superior colliculus. These characteristics position the dorsal striatum as a promising candidate for neuromodulation, which Hyder et al have now systematically evaluated for seizure-modulating potential.
The authors investigated optogenetic stimulation of the dorsal striatum in three distinct and complementary rat models of epilepsy: the gamma-butyrolactone model of absence epilepsy, the amygdala kindling model of temporal lobe epilepsy, and pilocarpine-induced status epilepticus. They leveraged bidirectional optogenetic techniques to sequentially stimulate and inhibit light-activated channels ChR2 and ArchT, respectively. They evaluated both continuous open-loop and closed-loop stimulation. They surveyed the effects of bidirectional stimulation in each model, measuring epileptiform activity and seizures, both in terms of frequency and severity.
Their findings demonstrate concordance across models, supporting a beneficial effect of striatal activation. Excitatory neuronal stimulation suppressed seizures or epileptiform activity in all three models, a highly encouraging result for a novel stimulation target that suggests a broader, generalizable mechanism for seizure control rather than one dependent on specific electrographic or structural onset zones. High-frequency (100 Hz) stimulation consistently outperformed low-frequency (5 Hz) stimulation, reducing epileptiform spike-wave discharges and preventing electrographic and behavioral seizures in 60% and 80% of kindled animals, respectively. When employing closed-loop stimulation similar to clinical responsive neurostimulation, the bidirectional effects became particularly clear, with high-frequency activation shortening epileptiform discharges while inhibitory stimulation prolonged them.
These complementary findings reveal the striatum as an active bidirectional modulator of seizures, with activation suppressing and inhibition exacerbating seizure activity. This positions the striatum as a potentially more accessible therapeutic target than brainstem nuclei. While movement-related side effects are a concern given the striatum's motor functions, the authors found no noticeable locomotor abnormalities in their basic screening. Human studies provide additional reassurance: Phase I trials of caudate DBS for treatment-resistant tinnitus demonstrated no significant cognitive or neurological changes attributable to chronic stimulation, with adverse events being primarily surgical or device-related and transient. 9 Though epilepsy-specific safety evaluations would be needed, existing human experience with striatal neuromodulation is encouraging.
The striatum's effectiveness across absence and temporal lobe seizure models challenges traditional views of the basal ganglia as primarily motor and reward structures. The authors propose a mechanistic framework where direct pathway activation predominates, leading to inhibition of the substantia nigra pars reticulata and subsequent seizure suppression. This positions the striatum as a gating structure that modulates diverse seizure networks via basal ganglia outputs. However, the separate contributions of D1 and D2 pathways remain to be determined. While these mechanistic details await clarification, translating this optogenetic proof-of-concept into clinical practice will pose distinct challenges.
In humans, the striatum's elongated shape and curvature would likely require multiple electrodes for adequate coverage, unlike the compact rodent structures. More fundamentally, electrical stimulation lacks the cell-type specificity of optogenetics, and while light-activated cation channels allow precise neuronal activation, electrical stimulation produces mixed effects, including depolarization block. Whether electrical high-frequency stimulation can recapitulate the selective excitatory effects seen with optogenetic activation remains uncertain and would require careful validation.
Despite these challenges, this work strongly supports the dorsal striatum as a promising therapeutic target for drug-resistant epilepsy, particularly for patients with mixed seizure types. The bidirectional control demonstrated here, with activation suppressing and inhibition worsening seizures, reveals the striatum as an underappreciated gatekeeper in seizure networks. Next steps include determining whether these effects can be reproduced with clinically available interventions, such as electrical stimulation, and identifying optimal targeting strategies within the complex anatomy of the human striatum. If successful, striatal neuromodulation could offer broad-spectrum seizure control that has thus far eluded current neurostimulation approaches.
