Slide 17

Figure 17. The original Mainen and Sejnowski model of a cortical neuron replicates the kink in the action potential at spike onset and the increased variability in the voltage at spike onset (in comparison to the axon initial segment). a-c are from the soma and d-f are from the axon. Note the kink in the somatic action potential, but the spike in the axon is smooth, as predicted by Hodgkin and Huxley models. However, the Mainen and Sejnowski model is based upon a high density of Na+ channels in the axon intial segment - a finding that is supported by immunocytochemical studies (Inda, M. C., DeFelipe, J. & Munoz, A. Proc Natl Acad Sci U S A 103, 2920-5 (2006); Komada, M. & Soriano, P. J Cell Biol 156, 337-48; 2002), but not patch clamp studies (Colbert, C. M. & Pan, E. Nat Neurosci 5, 533-8 (2002).

However, we have recorded action potentials at a high resolution (83-166 kHz, no filtering) in the axon initial segment of layer 5 pyramidal cells and found that the peak dV/dt of these spikes are routinely between 1200 and 1600 V/s (in contrast spikes in the soma have a peak dV/dt of between 300 and 700 V/s). This high peak dV/dt of spikes in the axonal initial segment is consistent a previous study by Stuart et al. (J. Physiol. 505: 617) who measured a peak dV/dt of approximately 800-1000 V/sec. Our results are consistent with the presence of a relatively moderate to high density of Na+ current in the axon initial segment, in agreement with the immunocytochemical results quoted above.