To Myelinate or Not to Myelinate

Myelinated and non-myelinated axons co-exist in the brain. What determines whether an axon should be myelinated or not? Rushton proposed that among myelinated and non-myelinated alternatives with the same conduction velocity (CV), a thinner fiber is chosen due to its smaller volume cost. If the required CV is below a critical value, a non-myelinated axon is thinner and used. Otherwise, a myelinated axon should be used. This argument leads to the prediction of a critical axon diameter, which segregates myelinated and non-myelinated axons. We show that minimizing axonal cost does not necessarily lead to a critical diameter. By considering two optimization scenarios, we show that the appearance of an axon diameter spectrum depends on the relative unit volume cost between a myelinated and a non-myelinated axon. Both scenarios suggest that if unit volume cost of a myelinated axon were significantly less than that of a non-myelinated axon, there would be a gap between the distributions of non-myelinated and myelinated axon diameters. Otherwise, there would be an overlap between the distributions of non-myelinated and myelinated axon diameters. In addition, the second optimization scenario predicts a gap in the CV spectrum. By comparing our predictions with data from the corpus callosum, which shows overlap of the diameter distributions, we conclude that myelinated axons are costlier than non-myelinated ones. Prediction of a gap in the CV spectrum needs to be tested experimentally.