Altered Behavior of Epileptic Interneurons and Principal Cells during Cognitive Processing

Gert Dehnen1, Jan Boström2, Rainer Surges1, Florian Mormann1

1 Dept. of Epileptology
2 Dept. of Neurosurgery

The human medial temporal lobe (MTL) plays a pivotal role in the perception of our environment and the encoding of experienced situations into episodic, autobiographic memory. Due to its recurrent neural architecture, it is also the most frequent region from which seizures originate in focal epilepsy. Previous studies have shown that principal cells in the human MTL can encode sparse, selective and invariant representations of the environment. In contrast, interneurons tend to respond less selectively to the presented stimuli. As it is assumed that interneurons, and consequently principal cells, show an altered recruitment in the epileptic MTL, we tested for differences of their dynamics in MTL regions during object recognition. To this aim, patients were presented with pictures of persons, animals, landscapes and objects on a laptop screen. After spike detection and sorting, interneurons and principal cells that responded to one or more of the presented stimuli were detected. By comparing response characteristics of single-neurons and firing relationships to ongoing oscillations in the local field potential (LFP) between the epileptic and the contralateral hemispheres, we looked for differences in neuronal processing specifically related to the epileptic process.

The most striking difference between the epileptic and contralateral hemisphere was the fraction of responsive units (Fishers-exact-test: p = 2.5∙10-16). Analyzed dynamic properties such as firing rate, peak and onset latency in response to stimulus presentation largely showed the same behavior. Phase locking of action potentials to ongoing oscillations, determined via Hilbert transform and Rayleigh test, showed a significant difference between the epileptic and contralateral hemisphere, particularly in the theta and delta band.

We hypothesize that reduced spike timing variability may cause the reduced neuronal responsiveness in the epileptic hemisphere.