Abstracts

Rationale: Febrile status epilepticus (FSE) in children has been linked to cognitive deficits and epileptogenesis later in life. Recent evidence in the experimental hyperthermia animal model of FSE (eFSE) has shown that spatial memory deficits may be a result of structural and functional changes to specific cell types in the hippocampal circuit as well as the synaptic connections between regions of the hippocampus. The corresponding loss of functional connectivity between CA1 and CA3 hippocampal subfields may lie at the root of differential cognitive outcomes post FSE. We used slow and mid-frequency gamma measures as proxies for CA3-CA1 and Entorinal Cortex–CA1 inputs in relation to performance of a complex spatial task. We hypothesized that cognitive outcome post FSE would correlate with the efficacy of gamma oscillations as a function of both task demands and movement speed. Methods: Postnatal day 10–11 male Sprague-Dawley rat pups (N=7) underwent procedures for eFSE hyperthermia induction. Pups were subjected to a continuous stream of warm air until hyperkinesis and chewing automatisms were identified. Seizure behaviors typically progressed to clonic movements and eventual tonic extension. After approximately 60 min of hyperthermia, eFSE pups were immersed in cool water to aid in the return of core temperature to normothermia and to promote seizure cessation. Control animals (N=5) were separated from their dames for an equal amount of time as eFSE animals.At approximately 2 months old, control and FSE rats were trained on an active avoidance task on a rotating arena. Rats would run counter the direction of arena rotation to avoid being moved into a sector of the arena where they would receive a mild electrical shock. Rats that reached criterion of five or fewer shocks within 2 consecutive sessions before a 16-session cutoff were labeled as ‘learners’ (FSE-L) while those that did not meet criterion were labeled ‘non-learners’ (FSE-NL).At approximately 3 months old, all rats were implanted with an array of hippocampal electrodes for the recording of CA1 local field potentials. During performance of the task, we analyzed the properties of theta, slow and medium gamma oscillations in relation to avoidance. In the case of non-learners, we analyzed these oscillations in relation to running away from the shock zone post shock. Peri-stimulus time histograms of spectral properties of each bandwidth of interest were analyzed for time epochs -6 and +6 seconds from each acceleration peak. The normalized amplitude properties for each bandwidth as a function of time from peak acceleration were then analyzed using General Estimating Equations (GEE). Results: GEE interactions between time from peak acceleration and normalized signal amplitude reveal that controls and FSE-L exhibited similar patterns in the transition between epochs approximately 3-4 seconds before avoidance and 4 seconds post avoidance that were dominated by slow gamma. In contrast, epochs closer to peak acceleration were dominated by theta and mid-frequency gamma oscillations. FSE-NL animals did not exhibit a separation between theta, slow and medium gamma properties -4 seconds before avoidance or 4 seconds post avoidance. However, oscillations patterns did exhibit the same features as FSE-L and control animals where slow gamma was reduced and theta and medium gamma oscillations were dominant in time epochs closer to peak acceleration. Conclusions: Our results indicate that post-FSE cognitive outcomes on the active avoidance task may be a consequence of alterations to CA1 synaptic inputs, as determined by alterations to slow and medium gamma oscillations in relation to avoidance behavior. In line with prior reports, slow gamma may reflect the efficacy of CA3-CA1 inputs as well as active recall prior to avoidance from the shock zone. As the FSE-NL did not exhibit effective learning of the task, it stands to reason this is why they did not show electrophysiological or behavioral evidence of recall. Yet, this result is further evidence that poor CA1-CA3 communication may lie at the root of cognitive dysfunction post-FSE. This is reinforced by the maintained relationship between medium gamma and increased movement speed in epochs following peak acceleration. This may imply that entorhinal cortex-CA1 connectivity is maintained in FSE-NL animals. The results may also point out important differences in the functional aspects of slow and medium gamma during moments of increased cognitive or sensorimotor demands. Funding: NS108296 to GLH, JMB and TZB