Axon diameter determines conduction velocity, where increasing size confers an evolutionary advantage in prey-predator confrontations, the squid giant axon, which initiates the animal’s escape reflex, an obvious example. In mammals myelin increases axon conduction velocity (CV) without significantly increasing diameter; in the mammalian cortex only 20% of axons exceed 3 μm in diameter. However, the importance of increased CV with diameter diminishes in short central tracts, e.g., the optic nerve. Mitochondrial density in optic nerve axons is constant such that energy capacity increases as a squared function of diameter (Perge et al., 2009), whereas membrane bound ion channel and Na pump density increase linearly with axon size, bestowing larger axons with an increased capacity to re-equilibrate the ion perturbations that result from action potential firing. We describe differential sensitivity of mouse optic nerve (MON) axon sub-populations to sustain firing at high stimulus frequencies.

All procedures were carried out in accordance with the Animals (Scientific Procedures) Act 1986 under appropriate authority of project and personal licenses. Adult male CD-1 mice were killed by cervical dislocation and decapitated. Optic nerves were dissected, placed in a superfusion chamber, and bathed with aerated aCSF containing 10 mM glucose flowing at 2 ml min-1. The stimulus evoked compound action potential (CAP), whose profile displayed three distinct peaks, was evoked with supra-maximal stimuli. The MONs were stimulated at 1 Hz under baseline conditions.  
The areas of the three separate peaks recruited with increasing stimulus voltage were fit as cumulative histograms. The resulting estimates of mean ± SD values (n = 12) allowed creation of corresponding probability distributions, which yielded three distinct peaks whose profile matched that of the CAP, evidence that axon size underlies the separation of the CAP into three distinct peaks. In MONs in which high frequency stimulus was imposed for 5 minutes, the 3rd peak loss occurred at 10 Hz, the 2nd peak at 20 Hz and the 1st peak at 67 Hz. During stimulus axons release K+ thus the effect of increasing aCSF [K+] from the baseline value of 3 mM on the CAP was studied. The CAP started to fall when [K+] was increased to 9 mM and a sigmoidal fit showed an IC50 value of 12.5 mM with a slope of -9.38 (n = 8). Post-stimulus recovery revealed peak amplitude transiently exceeded baseline values in a frequency dependent manner in agreement with a decrease in [K+] below baseline (Ransom et al., 2000). Reducing the aCSF [Na+] from 153 mM in steps to 30 mM decreased rate of rise and amplitude of the individual peaks in a logarithmic manner (n = 10). From this data, based on the Nernstian relationship, stimulus induced elevations in [Na+]i were calculated for each peak.

Our data provides evidence that the ability of axons sub-populations to ‘follow’ at high frequency stimulus is determined by axon size, and that elevations in [Na+]i, not [K+]o, is the primary determinant that causes decrease of the CAP.