Purkinje cells (PCs) are the main and only output neurons in the cerebellar cortex and have been implicated in motor coordination, learning and cognitive functions. There are two types of action potential output occurring in PCs: simple spikes (SSs) and complex spikes (CSs); both being generated in the axon (Stuart & Häusser, 1994). CSs are activated from a single excitatory input from a climbing fibre, originating at the inferior olive. They occur at a low frequency and produce a prolonged depolarization. Sensory stimuli, movement and feedback control are involved in originating CSs. SSs receive various inputs from mossy fibres and generate a brief, single-action, high-frequency, excitatory potential (Miall et al. 1998). Propioception, sensory and motor association area inputs are the main signals that trigger SSs. However, until now elucidation of the site of initiation of SSs and CSs has been unclear.
Previous published data proposed that SSs and CSs originate at the axon (Stuart & Häusser, 1994; Davie et al. 2008), but leave unanswered the question of exactly where in the axons it is. Recently, Palmer and colleagues (2010) conducted research to determine just where in PCs that SSs and CSs are generated. Two experimental methods (one invasive and one non-invasive) were used to resolve this issue. A multi-site extracellular recording device (non-invasive) was used in rat PCs to visualise action potential across the axon. The averages of axonal and somatic spikes were aligned and the axosomatic delays were calculated by locating the time when the extracellular spike at each axonal location reached 10% of its maximum height, because it is dominated by sodium influx and not disturbed by other ionic currents, in the initial axon segment and first Ranvier node. Samples were taken at five or more locations within the first 200 μm of the axon. The axonal site with the longest axosomatic delay is by definition the site of SS generation. They observed that the first segment had the longest delay (approximately 20 μm). The second approach used voltage-sensitive dye imaging to directly measure membrane potential changes simultaneously at multiple locations along PC axons. The amplitude of the optical signal was measured, but here the delay was half the amplitude of axonal fluorescence to minimize interference. The longest delay for SSs in the axon occurred at 144 ± 47 μs at a distance of 15 ± 4 μm from the axon hillock. A comparison between cell-attached and extracellular measurements was made because the ionic currents could influence recordings. The difference between these two methods lies probably in the difficulty of showing if there is really an ionic and capacitive contamination in cell-attached recordings. The initial segment of the axon contains a high density of voltage-activated channels. However, when they fitted the data on a V-plot, the moving average demonstrated two minima, one close to the initial segment and the other close to the first Ranvier node. For CSs they applied the same techniques and saw that the results did not differ from SSs.
Using the two methods, they showed that the initiation of SSs and CSs is in the axon and presented conclusive results that this happens in the first segment (15–20 μm from the soma); also the PC axon in a mouse measures up to 2 mm and in a rat up to 3 mm. Although SSs and CSs are initiated at the same location, CSs retrogradely propagate to the soma; maybe by pre-charging of the membrane capacitance of the perikaryon and dendrites by the climbing fibre synaptic conductance. However, back propagation in the dendrites is not supported because of their low Na+ channel density (Stuart & Häusser, 1994; Golding et al. 2001). The waveform of the first spike in CSs preceded the soma signal by 42 ± 16 μs, meaning that the first spike of the CS to the soma is faster than the propagation of the SS (three times faster). It is worth mentioning that not all the neurons exhibited initation at the first segment; two cells showed initiation either at the first node or simultaneously at the first node and at the first segment.
Why is this important? These new data corroborate, with reliable and independent methods, that the action potential is generated in the axon, specifically in the first segment. These results propose that this type of ‘physiological arrangement’ could be for synaptic efficacy and timing, therefore enhancing correct functioning. If we know how action potentials are generated, we will realize how plasticity takes place and how information is processed. This is critical, as the site of action potential initiation corresponds to the final site of synaptic integration and consequently provides a single target for modulation of both spikes.
This is the first time that the site of initiation of CSs and SSs has been determined. Results from this study showing where the action potential is generated, should enable this to be modulated and should also allow us to find how the equilibrium between long-term potentiation and depression is acting, and modulate them also. They envision a way to manipulate and obtain new information at this site of generation of action potentials. Finally, it has been demonstrated that in neocortical pyramidal neurons this occurs similarly (Golding et al. 2001), suggesting that this could be a general principle for all neurons.