Abstracts

Rationale: Absence seizures (AS) are characterized by the abrupt onset of synchronous spike wave discharges (SWD) across the brain. Efforts to assign AS generation to a localized brain region have yielded conflicting results over the necessary and/or sufficient structures. Such approaches may however neglect the roles of multifaceted interactions between regions. Therefore, a large gap exists in our understanding of the region-specific rules for involvement in AS.

Methods: We used surface electrocorticogram (ECoG) recordings (16 sites) in Scn8a+/- mice to measure oscillations during AS. We then used sparse identification of nonlinear dynamics (SINDy) to identify governing equations for AS generation and used high density silicon probe recordings (1152 sites) to identify the neuronal spiking correlates in the thalamus and cortex.

Results: In ECoG recordings, AS onset was in frontal (Fr), motor (M), and somatosensory facial and whisker (SSf/w) regions while other regions were delayed by 200-1000 ms. Phase synchrony between all regions increased during AS in frequency bands at 15, 22.5, 20 and 37.5 Hz. At 7.5 Hz, Fr synchrony increased with all regions and synchrony increased between Fr, M, and SSf/w while other regions decreased or were unchanged. We then used SINDy to identify governing equations for regional/frequency band specific oscillation phases. Of 791 candidate terms for each frequency in each region, sparse regression solutions identified an average of 12.1 active terms per frequency per region. Solutions followed the general form that phase velocity equals the regional mean phase velocity plus the sum of the sine of the phase differences with other specific regions and/or frequencies (modulated by phase shifts and oscillation frequency ratios). In differential equations of this form, convergences in oscillation frequency promote synchronization. We thus compared interregional phase velocities, finding recruitment to AS is correlated with a region’s convergence to the common phase velocity. Silicon probe recordings also revealed convergences in neuronal spike rates between hundreds of neurons and alignment with specific SWD phases both within and between several regions of the cortex and thalamus, providing spiking correlates of the macroscopic ECoG oscillations.

Conclusions: Here we provide a physically interpretable model of AS generation derived from experimental data. These results show multiple interactions govern the phase of specific oscillations within and between brain regions generating AS. Such inseparability of components is the hallmark of nonlinear systems, indicating AS are not governed by a single structure. Local defects may promote AS generation, but these may be one of many parameters biasing interregional communication towards AS. These results identify multiple targets for treatment and suggest exclusive focus on any single site would greatly diminish development of alternative treatment options targeting different regions or their interactions.

Funding: Please list any funding that was received in support of this abstract.: This work was supported by (Epilepsy training grant #NS07288) to JMH and (R01 #NS34774) to JRH.