Abstracts

Rationale: There is often a stereotyped sequence of seizure propagation from one brain area to another during secondary generalization of focal seizures. How such patterns of propagation are formed and consolidated is the focus of our study. In order to simplify this complex problem, we have created a model system where there are only two cortical areas (left and right cingulate cortices) bridged by just one axonal tract, the corpus callosum. Cingulate cortex is an area especially amenable to studies of callosal circuits in an in vitro preparation; this is an area closest to the corpus callosum, and so the distance between neurons synaptically coupled by callosal axons is relatively short and less likely to be severed in a slice preparation. Bathing these slices in epileptogenic solutions results in spontaneous seizures that propagate from one hemisphere to the other, and the corpus callosum is the only physical structure available in these slices through which activity can propagate (Walker et al., PLoS ONE 7:e31415, 2012). Methods: 350 micron thick coronal slices from 18-25 day old mouse anterior cingulate cortices were cut, preserving axonal connections across the corpus callosum. We produced epileptiform events via application of bicuculline or low magnesium ACSF to both hemispheres simultaneously. We measured neuronal activity electrophysiological using extracellular electrodes.Results: In one of our models of seizure generation (BIC: blockade of inhibition via GABA-A receptor antagonists), we see a strongly unidirectional mode of propagation. That is, once seizures emerge after about 10 minutes of incubation, they soon evolve towards a very reliable pattern, propagating from a leading hemisphere to lagging hemisphere, and rarely in the reverse direction. This finding in BIC is significantly contrasted with experiments using zero millimolar magnesium solutions to evoke seizures (0Mg model), as very little directionality is found in these recordings: BIC vs. 0Mg is 0.86 0.04 (n = 17) vs. 0.66 0.04 (n = 12), respectively (p<0.02, rank sum test, where 1.0 = 100% directionality and 0.5 = no directionality). Some hint as to why this could occur was found when we were able to successfully bisect the callosum after seizures had propagated for about an hour in BIC. After the cut, we observed that the leading hemisphere had a much higher rate of independent seizure generation. This finding is consistent with a coupled oscillator hypothesis for propagation, and we ve developed a simple computational model using IGOR macros to demonstrate how that mechanism could work. Furthermore, we have shown that very small temperature differences between hemispheres have a strong effect on which direction seizures propagate in the slice.Conclusions: The patterns of seizure propagation and direction can be significantly influenced by different modes of inducing those seizures, and we find that our experimental results are consistent with results from a computational model of seizure propagation based on coupled oscillators.