Abstracts

Rationale: The large neuronal depolarizations associated with epileptic events lead to a significant influx of Ca²⁺ into neurons. Ca²⁺ ions are involved in the regulation of many intracellular processes and therefore Ca²⁺ dynamics and clearance mechanisms may have a significant impact on seizure evolution. In this computational study we investigate how periods of long and recurrent neuronal bursting affect seizure dynamics. Methods: Calcium dynamics are simulated in a spherical neuron model incorporating Ca²⁺ influx, diffusion, buffering, and extrusion. The neuron model is composed of nine concentric shells, a core, an extrusion pump, and includes two buffer systems intended to represent mobile and immobile intrinsic Ca²⁺ buffers. We use a spatially distributed array of interconnected spherical neurons capable of reproducing recurrent bursting activity and the behaviours consistent with human seizures. Results: Simulated Ca²⁺ dynamics have two components: a fast component associated with rapid Ca²⁺ changes following APs and a slow component resulting from Ca²⁺ redistribution inside neurons. The first is determined by Ca²⁺ diffusion processes and fast buffering while the latter depends on characteristics of immobile buffers. Repetitive neuronal bursting, representing seizure-like activity in the network, alters the distribution of fast buffers and the concentration of slow buffers in neurons. Once periodic bursting is established in the network, the baseline of the free mobile buffer concentration in neurons decreases over time in a spatially nonuniform manner. This affects the fast buffer gradient concentration, which in turn impacts the Ca²⁺ diffusion process. The direction of Ca²⁺ transport inside neurons changes over the time course of the simulated neuronal bursting. The diffusion of Ca²⁺ ions from the upper into the deeper shells at the beginning is later reversed as Ca-bound fast buffers begin to flow in the opposite direction. The concentration of free slow immobile buffers shows a hyperbolic decrease in time with the fastest rate of decrease in the outermost shell. Typically after 200 seconds the concentration of slow and fast buffers stabilizes at new resting levels. Conclusions: This network model study throws light the on complex Ca²⁺ dynamics during simulated epileptiform activity. Throughout the duration of the simulation the concentration of fast and slow free buffers show a hyperbolic decreases in time. Inside a neuron, the rate of the buffer concentration decrease is not uniform. The process of Ca²⁺ buffers redistribution may produce gradients in buffer concentration affecting the direction of Ca²⁺ diffusion in neurons. The characteristic time of the Ca²⁺ buffer redistribution process in a neuron after the induction of recurrent bursting in the network is consistent with the duration of seizures in humans. Supported by NIH grand NS51382