Abstracts

Rationale: Developmental cortical malformations (DCM), such as polymicrogyria, have a high incidence of drug-resistant epilepsy, but the underlying mechanisms by which these lesions contribute to the onset of seizure activity remain poorly understood. DCM can be modeled using neonatal freeze-lesion (FL), which has been shown to cause in vitro cortical hyperexcitability following a 2 week latent period. FL-cortex also shows an upregulation of the astrocyte-secreted protein thrombospondin (TSP), prior to the onset of epileptiform activity. TSP is known to induce excitatory synapse formation, which we hypothesize contributes to the pathological reorganization of the FL cortex. The neuronal receptor for TSP is the calcium channel subunit α2δ-1, which is also transiently upregulated following FL. We hypothesized that TSP and α2δ-1 upregulation leads to aberrant excitatory synaptogenesis and pathological network formation and that targeting TSP/ α2δ-1 signaling will be protective against epileptogenic processes following neonatal cortical insult.Methods: To address this hypothesis, we used both a pharmacological and genetic strategy. First, daily, I.P. injections of gabapentin (GBP, an antagonist of TSP/a2d1 signaling) were used to disrupt the interaction of TSP with α2δ-1, for one week to coincide with the time period of TSP/ α2δ-1 upregulation. IHC was used to count synapses and assess anatomical changes in GBP and saline-treated FL animals. Potential neuroprotective effects of GBP were examined using a TUNEL assay to detect cell death. FL was then performed in mice with a germ-line, global knockout of α2δ-1 to address the same question but avoid the potential non-specific drug effects of GBP. In vitro hyperexcitability was assessed by recording evoked cortical fields potentials (fEPSP).Results: In line with the known mechanism of TSP/ α2δ-1-driven synaptogenesis, GBP treatment blocked cortical hyperexcitability and the rise in excitatory synapses following FL. GBP treatment also attenuated anatomical cortical reorganization and cell death. When FL was performed in α2δ-1 KO mice, we found that they had less epileptiform activity compared to FL wild-type (WT) littermates. Interestingly, the FL α2δ-1 KO mice have an intermediate phenotype compared to GBP treated FL WT animals. This suggests that KO of α2δ-1 during development leads to genetic compensations that contribute to pathological network formation following FL. Finally, α2δ-1 KO mice appear insensitive to GBP treatment, providing compelling evidence that the therapeutic effects of GBP are specific to its interactions at α2δ-1.Conclusions: These results shed new light on how hyperexcitable networks are formed after injury and suggest the use of GBP, or other modulators of TSP/α2δ-1 signaling, as potential therapeutic agents to minimize epileptogenesis associated with DCM and other post-traumatic epileptic conditions.