Abstracts

Rationale: Posttraumatic epilepsy (PTE) is a long-term negative consequence of traumatic brain injury (TBI) in which spontaneous recurrent seizures (SRS) occur after the initial head injury. PTE develops according to a poorly-defined process during which circuitry reorganization in the brain causes permanent hyperexcitability. Clinically relevant models of PTE are key to understanding the mechanisms underlying the development of PTE. Aquaporin-4 (AQP4), a water channel primarily expressed on astrocyte endfeet contacting blood vessels, is a key epileptogenic protein. Previously, AQP4 has been shown to be mislocalized in resected epileptic human sclerotic hippocampi. Here, we aim to determine the role of AQP4 in the development of PTE in mice at various time points after TBI.  Methods: Adult male CD1 wild-type (WT) and aquaporin-4 knockout (AQP4 KO) mice were used. Mice were subjected to TBI in the right frontal cortex using the controlled cortical impact (CCI) injury model. Sham mice received craniectomy only. 10 days prior to each final time point (14, 30, 60, and 90 days post injury (dpi)), mice received an indwelling electrode in their hippocampus and underwent 1 week of continuous video-electroencephalographic recording. Western blot analysis and immunohistochemistry for AQP4 were then performed for both genotypes at each time point.  Results: TBI mice exhibited nonconvulsive SRS. The total number of SRS are 1/8 (13%), 3/10 (30%), 4/12 (33%), and 2/7 (29%) for WT mice and 4/7 (57%), 5/13 (38%), 3/6 (50%), and 0/4 (0%) for AQP4 KO mice at 14, 30, 60, and 90 dpi, respectively. The total number of mice that developed PTE are 1/8 (13%), 2/10 (20%), 3/11 (27%), and 1/7 (14%) for WT mice and 2/7 (29%), 4/13 (31%), 2/6 (33%), and 0/4 (0%) for AQP4 KO mice at 14, 30, 60, and 90 dpi, respectively. AQP4 KO mice also displayed significantly longer seizure duration compared with WT mice (p=0.0085). Power spectral density analysis of the seizures revealed a significant difference in the beta frequency band between genotypes (p=0.0002). Qualitative Morlet wavelet analysis revealed heterogeneity in the seizures. Sham mice did not develop PTE. Western blot analysis showed a significant increase in AQP4 in the frontal cortex (p=0.0381) and hippocampus (p=0.0034) of mice with PTE compared with mice without PTE. Confocal images revealed AQP4 dysregulation primarily in mice with PTE at 14 dpi where the frontal cortex exhibited marked loss of AQP4 and in the hippocampus AQP4 immunoreactivity was largely expressed in the soma and major processes of astrocytes and blood vessels were denuded of perivascular AQP4 in the hippocampus. Perivascular was abundant in the frontal cortex and hippocampus of sham mice.  Conclusions: Overall, we have: 1) obtained the highest yield of PTE after CCI reported to date; 2) found significant differences in the beta frequency band of seizures between WT and AQP4 KO mice; and 3) observed AQP4 dysregulation, specifically upregulation and mislocalization of perivascular AQP4 in mice that developed PTE. Collectively, these findings suggest that PTE may be modulated by AQP4 and that AQP4 may be a target for antiepileptogenic strategies after TBI.  Funding: No funding