Abstracts

Rationale: The cross-frequency coupling (CFC) of brain electrical activity at different frequencies represents interactions between different elements of brain network circuitry. This coupling phenomenon was examined in a rat model of temporal lobe epilepsy (TLE) to see if any changes to CFC as seizures progress could reveal details of network changes during TLE. Low-frequency phases have been proven to modulate the amplitudes of high-frequency oscillators. In rodent hippocampal and cortical circuits, it is known that gamma power is modulated by theta waves and represents the activity of inhibitory interneurons critical for network synchrony. Analyzing theta-gamma (θ-γ) CFC as spontaneous seizures evolve could provide insight into underlying ictogenesis brain dynamics.

Methods: Depth electrodes were implanted in eight locations of the brain in young adult male Sprague-Dawley rats (bilateral hippocampus, thalamus, subiculum, and entorhinal cortex). Chronic TLE was induced by intraperitoneal (i.p.) injections of pilocarpine (10mg/kg) following pretreatment with Lithium Chloride (LiCl) and allowed to undergo status epilepticus for 90 minutes before diazepam i.p. injection (10mg/kg). Following a 4-6 week latency period, intracranial seizure activity was recorded in awake, freely-moving animals. Noise-assisted multivariate empirical mode decomposition (NA-MEMD) was utilized to extract a set of finite orthogonal neuronal oscillators or intrinsic mode functions (IMFs). IMFs in the theta (4-8 Hz) and gamma (30-80 Hz) frequency ranges were analyzed for CFC by computing mutual information (MI) of the pairwise comparisons across all recording locations. MI assessed CFC of θ-γ by measuring the strength of the phase-to-amplitude modulation of these frequencies. An MI value of zero indicated no shared information, whereas a high MI value indicated greater coupling between sites.

Results: A significant decrease in θ-γ CFC was detected in rats up to one hour before seizure onset suggesting a decrease in the activity of inhibitory interneurons important for synchrony. This finding is also consistent with our measures of decreased multisite network synchrony. Eight of the twenty eight possible pairs that had the highest MI showed the greatest decrease in CFC suggesting a loss of coupling occurred across several locations of the brain network prior to seizure onset.

Conclusions: There is a significant need to understand the network brain dynamics as seizures evolve. In a rodent model of chronic TLE, we found the brain exhibited desynchronization across multiple sites preceding onset. Fast-spiking inhibitory interneurons, whose activity resides in the gamma frequency band, may be a critical element in ictogenesis in TLE. Hence a reduction in θ-γ coupling may help to explain how preictal desynchronization develops prior to ictal onset.

Funding: Please list any funding that was received in support of this abstract.: Supported by NIH R01 NS092760 to DJM.