Abstracts

Rationale: The potent anticonvulsant effect of pharmacological inhibition of the substantia nigra pars reticulara (SNpr), originally identified in the early 1980s by Gale and colleagues, has long been hypothesized to function by disinhibition of the deep and intermediate layers of the superior colliculus (DLSC). The SNpr is a principle source of GABAergic inhibitory input to the DLSC, and consistent with this hypothesis, activation of DLSC is also potently anticonvulsant. Selective targeting of nigrotectal projections (to the exclusion of projections from SNpr to other regions such as thalamus or pedunculopontine nucleus) was not achievable using pharmacological approaches. The advent of optogenetics has enabled direct targeting of this pathway, and thus direct testing of the DLSC disinhibition hypothesis. We have previously reported that optogenetic stimulation of the DLSC is potently anticonvulsant in various models of seizures and epilepsy (Soper et al., 2016). Methods: Adult, male Sprague-Dawley rats or adult female genetically epilepsy-prone rats (GEPR-3s) were injected with an AAV coding for the inhibitory opsin, ArchT, into the SNpr and fiber optics were implanted either into SNpr (to silence cell bodies) or within the DLSC (to silence nigrotectal terminals). Seizures were evoked by systemic administration of pentylenetetrazole (which can trigger both forebrain and hindbrain seizures), focal microinjection of bicuculine into the "area tempestas" (AT; to trigger forebrain seizures), systemic administration of gamma butyrolactone (to trigger absence-like seizures) and by auditory stimulation to trigger generalized tonic-clonic seizures in the GEPRs. Animals were tested within subject with and without optogenetic silencing. Results: In the absence of optogenetic silencing, the median seizure severity evoked from AT was a score of 4; this was reduced to a score of 1 by optogenetic silencing of cell bodies within the SNpr [unmodulated light, (n=13, p=0.0002); 100hz light, (n=11, p=0.001)]. When silencing nigrotectal terminals the median seizure score was reduced from 3 to 1 with unmodulated light delivery (n=11, p=0.001) and from 3.5 to 1.5 with 100hz light (n=8, p=0.008). Consistent with these data, inhibition of either SNpr cell bodies or the nigrotectal terminals with either unmodulated light or 100Hz pulsed light were effective at reducing seizure severity in each of the other models examined. Consistent with our hypothesis, the magnitude of seizure suppression was similar between sites of optogenetic silencing. Conclusions: These data show that selective suppression of the nigrotectal pathway is sufficient to account for anticonvulsant effects achieved by inhibition of the SNpr. The enhanced specificity provided by this approach may further enhance the translational utility and our understanding of this endogenous seizure suppressive circuit. Funding: KL2TR001432 and a Research Grant from the American Epilepsy Society to PAF