Ion channel gating is stochastic: channel opening involves randomly occurring step-like changes in conductance of the channel pore. While in the soma the number of channels is typically large and stochastic effects can often be ignored, in small structures the stochastic opening and closing of the ion channels can trigger spontaneous action potentials. Stochastic gating may constrain the diameter of unmyelinated axons to a minimum of 0.1um, as otherwise the rate of spontaneous action potentials would overwhelm the rate of synaptically triggered action potentials (Faisal & Laughlin, 2005). In the peripheral nervous system, unmyelinated C-fibers are often narrower than 1um diameter. Their spontaneous activity has also been implicated in some spontaneous and inflammatory neuropathic pain conditions. We hypothesized that axons of this size would be particularly sensitive to changes in the number or properties of their ion channels and that this could serve as a previously unexplored mechanism for pathological spontaneous activity that contributes to neuropathic pain.
To address this hypothesis we explored computer models of thin axons using the PSICS simulator (www.psics.org). We investigated a range of parameters to determine the sensitivity of spontaneous spiking to pathological ion channel regulation. We find that random closing of the K channels can depolarize the membrane such that Na spikes are initiated. Depending on the parameters, the fluctuations from stochastic K channels can even dominate spontaneous spike generation, as compared with Na channels. We also find a non-monotonic dependence of the spontaneous rate on the axon length. The highest spontaneous rates occur both for very short and very long cables. Spontaneous spikes are mainly triggered at the end of the cables, with a length scale that depends on the cell’s electrotonic properties. This implies that peripheral nerve endings are preferential zones for the generation of spontaneous activity.
We suggest that spontaneously generated action potentials could underly neuropathic pain. Deterministic models of spike generation typically lead to regular repetitive firing at high rates. In contrast, typical firing rates in neuropathic pain states are low (<1Hz) and irregular (Xiao & Bennett, ’07), consistent with our model.
Related Publications and Presentations
- Cian O’Donnell, Matthew Nolan, and Mark C W Van Rossum, “Stochastic action potentials in axons: a possible role in neuropathic pain”, Society for Neuroscience (SfN), 2009.