Abstracts

Rationale: We have recently shown that interictal spikes (IS) in the hippocampus can disrupt cognition in an animal model, which may have important implications for patients with temporal lobe epilepsy. We now assess which specific operations of hippocampal processing are most vulnerable to IS, and if other performance aspects such as reaction time are affected as well. We also examine endogenous hippocampal theta oscillations during behavior to assess their potential to naturally suppress IS. Methods: 20 Sprague-Dawley rats were infused with a pilocarpine solution (0.5 mg pilo/uL) into the ventral hippocampus, while monitored on depth-EEG for seizure activity in both hippocampi. Rats were then trained in an operant behavior task called delayed-match-to-sample, which requires hippocampal-dependent short-term memory. Some rats were trained on the task prior to pilocarpine infusion, to control for effects of brain lesions on task learning. Ultimately, 8 rats (5 epileptic rats with sufficient IS and 3 saline-treated without IS) with acceptable task performance were included in the analysis. Depth EEG was recorded from both hippocampi throughout all trials, and screened off-line to determine exactly where any IS occurred in the trials. Three epochs (stages) of short-term memory were analyzed - encoding, maintenance and retrieval. Power spectra data were also quantified to assess behavior-correlated theta power influences on the probability of IS to occur. Results: IS that occurred during the (memory) encoding epoch of trials did not affect performance in those trials (p=0.69). However, IS during the maintenance epoch tended to impair performance (p=0.053). Furthermore, IS during the retrieval epoch strongly impaired performance, with rats almost 3 times more likely to make an error (p<0.001). Hippocampal IS also affected reaction time, adding almost 0.5sec to response latency (p<0.001). Theta oscillations dramatically decreased the probability of an IS to occur (p<0.01), with IS approaching 0% likelihood during periods of maximal theta power during behaviors such as turning. Conclusions: We found that operation-specific interference in cognition processing, caused by IS, extends to structures in the limbic system. Hippocampal IS caused major disruptions in memory retrieval and slight impairments in memory maintenance. They did not affect short-term memory encoding, a finding distinct from the effects of neocortical IS described in earlier studies using scalp EEG. These findings can likely be attributed to specialized roles of the hippocampus in memory maintenance and retrieval. Hippocampal IS also increased reaction time. Lastly, we provide a quantification of the potential of theta oscillations to suppress IS during natural behaviors like turning and walking. These results fortify the argument that it may be critical to “treat the EEG” in patients with frequent IS, and put a spotlight onto the potential of endogenous oscillations to be utilized therapeutically for this purpose. This work was supported by NIH Grants F30NS064624 and R01NS056170.