Dendrites in Epilepsy

Epilepsy is one of the most common neurological disorders affecting up to 1% of the world’s population.[1] It is characterised by periods of “abnormal excessive or synchronous neuronal activity in the brain”.[2]

Although many studies have investigated changes in the cell’s output in tissue from epilepsy patients or animal models, few have examined the contribution of dendites, the cell’s input structure, to seizures. In collaboration with others in Bonn, Cologne, Tubingen and Israel, we have found dendritic changes that consistently boost the input-output relationship in different hippocampal neurones and in different models of acquired and genetic epilepsies.

Spermine_schematic
Schematic of spermine effects in epilepsy. Normally spermine blocks sodium channels in the dendrites. In epilepsy, spermine levels decrease, relieving the block of sodium channels. The additional depolarising conductance in the dendrite converts the cell from regular-firing (blue trace) to burst-firing (red trace).

In CA1 neurones from chronically epileptic animals, the sodium current is up-regulated in the dendrites. Although others have also shown an increased persistent Na+ current, we showed that the loss of spermine, which usually blocks the sodium channels, was responsible for the enhanced dendritic current.  Further, we found that the enhanced dendritic sodium current boosts synaptic inputs and may contribute to burst-firing. A similar enhanced dendritic excitability was observed to boost synaptic input in a model of a human genetic epilepsy. Preliminary data from the laboratory suggests that supra-linear integration and the generation of dendritic spikes may also be enhanced in epilepsy.

In granule cells, the effects of epilepsy were a little different. Granule cell typically have only apical dendrites which receive input primarily from cortical regions though to encode different aspects of the environment (see Integration in the DG). In chronically epileptic animals, the integration of input on the apical dendrite was no different from control animals. Also the ability of the dendrite to support the action potential travelling back along the dendrite, a simple measure of the electrical properties of the dendrite, was the same.

2010_10_20b 3D stack
3D peusdo-coloured image of granule cell from epileptic animal. Granule cell soma (centre) was filled with dye via patch-pipette (seen coming from the left). Apical dendrites project to the top of the image. This granule cell also had basal dendrites projecting to the bottom of the image and the axon (thin process) emanated from this basal dendrite.

However, following epileptic seizures some granule cells (5-10%) develop a dendrite projection in the opposite direction, called a basal dendrite. I found that this basal dendrite differs dramatically from the apical dendrite. Granule cells with a basal dendrite received input from their neighbouring granule cells. Changing the type of network from one where active cells only send their output out, to a network where active cells also activate neighbouring cells, possibly creating a positive feedback loop.

Our results suggest that network, intrinsic and dendritic properties collude to enhance excitability and promote synchrony, and thereby would contribute to the seizures in epileptic patients. In epilepsy, the principle cells fire high-frequency bursts of action potentials which in turn drive postsynaptic targets. In the dentate gyrus, the post-synaptic targets also change in epilepsy; granule cells now also target the basal dendrites of neighbouring cells. Furthermore, these basal dendrites are capable of d-spikes which boost the input-output response and are resistant to inhibition. Thereby further promoting activity, and perhaps contributing to a positive feedback loop. In CA1, input onto pyramidal neurones is boosted and the spatio-temporal specificity of d-spikes may be lost. Resulting in increased excitability and the loss of input feature detection.

We are excited to investigate if these altered dendritic properties are observed in the living animals, and if the hyper-excitable dendrites contribute to seizure generation and memory deficits? Furthermore, we are interested in examining the specific modulation of dendritic conductances as a possible therapeutic target.