Difference in axon diameter and myelin thickness between excitatory and inhibitory callosally projecting axons in mice

Abstract

Synchronization of network oscillation in spatially distant cortical areas is essential for normal brain activity. Precision in synchronization between hemispheres depends on the axonal conduction velocity, which is determined by physical parameters of the axons involved, including diameter, and extent of myelination. To compare these parameters in long-projecting excitatory and inhibitory axons in the corpus callosum, we used genetically modified mice and virus tracing to separately label CaMKIIα expressing excitatory and GABAergic inhibitory axons. Using electron microscopy analysis, we revealed that (i) the axon diameters of excitatory fibers (myelinated axons) are significantly larger than those of nonmyelinated excitatory axons; (ii) the diameters of bare axons of excitatory myelinated fibers are significantly larger than those of their inhibitory counterparts; and (iii) myelinated excitatory fibers are significantly larger than myelinated inhibitory fibers. Also, the thickness of myelin ensheathing inhibitory axons is significantly greater than for excitatory axons, with the ultrastructure of the myelin around excitatory and inhibitory fibers also differing. We generated a computational model to investigate the functional consequences of these parameter divergences. Our simulations indicate that impulses through inhibitory and excitatory myelinated fibers reach the target almost simultaneously, whereas action potentials conducted by nonmyelinated axons reach target cells with considerable delay.

Keywords: callosal axons; conduction velocity; impulse propagation; myelination; synchronization.

Fig. 1

Differential expression of GFP and RFP signals in excitatory and inhibitory axons. A) The tracer BDA labels both inhibitory and excitatory axons in the corpus callosum in the contralateral side of the injection. White arrows show inhibitory (GAD + BDA) long-projection axons, arrowheads depict GAD immunoreactive but unlabeled with BDA axons with unknown origin, whereas yellow arrows show long-projection GAD immunonegative (presumed excitatory) axons. B) GFP is expressed in non-GAD positive (excitatory), whereas RFP is in GAD immunoreactive axons. C) Line profiles indicate that RFP and GFP expression is in separate neuronal profiles (left panel). Middle graph shows that GFP-expressing axons (excitatory) devoid GAD. In contrast, RFP is co-localized with GAD (right panel). Corresponding colored bars on A depicts the measured areas for each graph. Scale: A: 15 μm; B: 10 μm.

Fig. 2

A) The electron microscope picture of sagittally cut sections from unlabeled control corpus callosum show variability in axon diameters and myelin thickness. B) DAB-Ni labeled (electron-dense) myelinated (arrows) and nonmyelinated (arrowheads) excitatory axons have different axon diameter and myelin thickness. C) High-magnification DAB-Ni labeled myelinated excitatory axon. D) DAB-Ni labeled inhibitory axons (arrows) are more scarce and almost always myelinated. E) High-magnification micrograph of myelinated inhibitory axons. F) High-magnification electron microscopy images show the fine structure of myelin sheets of excitatory F) and inhibitory G) projection axons. The number of laminae is more numerous around inhibitory (12 in this example) than excitatory axons (6 in this example). Scale: A, B, D: 500 nm; C: 100 nm; E: 250 nm; F: 40 nm; and G: 80 nm.

Fig. 3

A) Graph showing the relationship between axon diameter and myelin thickness of excitatory (black dots) and inhibitory (gray triangles) long-projection axons. B) The relationship between the myelin thickness and the number of myelin lamellae of inhibitory and excitatory myelinated axons is plotted. Regression lines (solid) with 95% confidence interval (dotted lines) clearly indicate differences in the myelination of excitatory and inhibitory long-projections axons.

Fig. 4

Representation of the models for myelinated and unmyelinated axons. Parameters for the model were sourced from literature and experimental studies. Lengths of all axons were maintained equal for each individual simulation. The sections corresponding to the soma and nodes were incorporated with active channels to enable the elicitation of action potentials. The internodes only contained passive channels and had their membrane properties adjusted corresponding to the extent of myelination.

Fig. 5

Effect of varying various biophysical parameters on the conduction velocity for axons of various lengths: A) axial resistivity, B) axonal diameter, C) number of layers of myelin, D) thickness of each layer of myelin, E) length of internodes, and F) overall membrane capacitance of internodes. L2 and L200 indicate axons of total length 2 and 200 mm, respectively. Note that for the internodal length a non-monotonic relation is observed, wherein for very short lengths a positive correlation is observed, whereas for larger lengths this trend reverses. In C–F) the unmyelinated fibers do not undergo any changes and are displayed only for ease of comparison with the myelinated fibers.

Fig. 6

Comparison of sensitivity to changes in various biophysical parameters for both excitatory and inhibitory myelinated axons evaluated for axonal length of 12 mm.

Fig. 7

Simulation outcomes from varying the axonal diameter across a much larger range of values for both excitatory and inhibitory myelinated axons of length 200 mm A) shows the conduction velocity with increasing axonal diameters, B) shows the conduction delay from the soma to the end of the axon.