Abstracts

Rationale: Surround inhibition has long been hypothesized as a defense mechanism against focal seizure propagation. However, how this inhibitory restraint is generated remains poorly understood. By applying novel microelectrode array recordings and wide-field calcium imaging in an acute pharmacological seizure model, we are now able to answer the following questions: What is the mediator that creates surround inhibition? How wide is the surround inhibition? How surround inhibition defect affects seizure dynamics?Methods: 500 nL 15mM 4-aminopyridine (4-AP) was injected into cerebral cortex in adult Sprague-Dawley rats. Seizure activity and its impact on the surrounding tissue was monitored by either a microelectrode array or the calcium indicator Oregon Green 488 BAPTA-1, AM. After confirming appearance of seizure-like events (SLE), bicuculine was applied either globally or locally at various distances in order to interrupt surround inhibition. Seizure dynamics were further analyzed by custom MATLAB functions.Results: We found that intra-parenchymal 4-AP injections induced repetitive non-propagating focal SLEs. Spike-triggered averaging showed unit activity in 4-AP injection sites was followed by positive local field potential deflections with 15 to 20 millisecond delays in the surrounding tissue. The signal was observed across the ipsilateral but not contralateral hemisphere. The short temporal delay and positive deflection suggested that the spike-triggered evoked potential was caused by feedforward GABA-A receptor-mediated inhibition. Globally impairing GABA-A receptor by applying bicuculine solution to cortical surface caused widespread SLE propagation. Focally disrupting the GABA-A receptor distally resulted in a new seizure focus that was non-contiguous to the 4-AP focus. Unit activity at the new non-contiguous focus exhibited opposed phase-locking and independent temporal dynamics from the 4AP focus. Calcium dynamics confirmed SLEs could jump and create two locally self-evolving ictal nodes at the opposite poles of the same hemisphere. All the non-contiguous SLEs were observed in the absence of pre-existing aberrant connectivity.Conclusions: 1. GABA-A receptor mediates surround inhibition. 2. Surround inhibition is hemisphere-wide. 3. Intact inhibition is required in both proximal and distal cortical area. 4. Two physically non-contiguous ictal nodes can be connected into a single epileptic network without pre-existing structural anomaly.