The Capability of Axons to Influence Their Own Myelination

The central nervous system contains both myelinated and unmyelinated axons. Myelination greatly increases the velocity of action potential propagation and is associated with Hebbian learning. It acts by essentially insulating the axons that it ensheaths allowing salutatory conduction to occur. Oligodendrocytes are glial cells that are responsible for the initial myelination and subsequent maintenance of CNS Myelin. Electrically active neurons have been shown to promote myelination.

It has been hypothesised that axons that are electrically active can act via oligodendrocytes to influence their myelination, however despite intense study there is still much contention about to what degree this occurs and whether axons are capable of influencing their own myelination. The author will discuss this area of study by evaluating research published in three high profile papers; using the evidence they provide to investigate this hypothesis. Across the field there are multiple hypotheses about how neuronal activity might influence oligodendrocytes to myelinate active axons with many different experimental approaches to investigate this. One such hypothesis is that neural activity regulates the proliferation and differentiation of oligodendrititc progenitor cells (OPCs) that in turn may influence axonal myelination. The three papers discussed in this essay all investigate this hypothesis to varying degrees and depths.

Gibson et al, 2014 set out to examine the myelination in the premotor cortex. To achieve this the authors used a mouse based model that was engineered to express excitatory opsin under the control of a Thy1 promoter. This allowed a subset of neurons to be stimulated with blue light at will. Mitew et al 2018 also used a mouse model, genetically adding a synthetic receptor and GFP marker to allow stimulation and visualisation of the axons.

Hines et al, 2015 also investigates the aforementioned hypothesis, using zebrafish to allow the identification and analysis of single defined axons, which previously has limited studies into the mechanism of axonal selection. Neuronal activity leads to the proliferation and differentiation of oligodendrocytes. Oligodendrocytes in the CNS synthesise and maintain myelin. In vitro cultured oligodendrocytes will randomly myelinate axons with of diameter of 0. 4μm or greater, suggesting that other mechanisms may be involved for myelination in vivo as not all large diameter axons are myelinated. Gibson and colleagues found that in both juvenile and adult mice, neuronal activity leads to proliferation of cells in the pre-motor cortex. Furthermore, when stained via immunofluorescence for cell identity markers, 54% of these dividing cells expressed the oligodendritic promoter Oligo2; suggesting that OPCs are directly influenced by neural activity. This is corroborated by Mitew et al 2015, with OPC proliferation in the corpus callosum being observed following of a small subset of axons. Using both a viral and pharmacogenetic approach to measure OPC proliferation in the corpus callosum, they found that OPC proliferation was observed from both methods. In addition, the density of mature oligodendrocytes increased suggesting that OPC differentiation was amplified, this was also observed in adult mice abet with a delayed response.

Importantly, only oligodrendrititc lineage cells proliferated in response to stimulation, which signified that OPC proliferation was a result of neural activity and not a pathophysiological response. This is expanded upon by Hines Neuronal activity leads to changes in axon myelination. In vivo, the dynamic nature of myelination in response to cues such as neural activity, is linked with normal brain development, behavioural changes and learning. The impact of changes in quantity myelin ensheathing axons can have great implications on the conduction velocity of action potentials and thus on the aforementioned processes.

Mitew and collagues found that stimulating neural activity in P14-19 (peak myelin formation period in CC) mice resulted in active axons being surround with myelin basic protein (MBP) with a high ratio than non-stimulated axons. This infers that active axons are more likely to be myelinated, however, both stimulated and control axons showed an increase in myelination, suggesting that activity-associated myelination is not completely specific to the active subset of axons. However, MBP levels may not be an accurate marker for the presence of myelin as premylinating oligodendrocytes also express this protein ( Gibson). The authors also attenuated neuronal activity by overexpressing which reduced myelination in an axon specific manner, however this did not have any effect on OPC or oligodendrocyte number signifying that this effect involved selection at the axonal level. This method was also used by Hines and colleagues.

The approach by Gibson et al using Transmission electron microscopy (TEM) and MBP expression to measure myelin levels avoid the potential issue of mis-quantifying myelin expression. This study found that the diameter of myelin sheaths was greater in stimulated axons compared to the wild-type (WT) controls, thus supporting Mitew findings. The work by Hines et al demonstrates how neural activity may influence oligodendrocytes via the release of vesicles. By using time lapsed microscopy, the authors demonstrated that electrically active axons may secrete vesicles on to the processes of oligodendrocytes, thus instructing them to myelinate that specific axon. However, neuronal activity might not influence myelination by oligodendrocytes. Hines demonstrated using TTX toxin to supress activity in axons that are fated to be myelinated has surprisingly no effect on oligodendrocyte number, differentiation or myelination of phoz2b+ axons, which if oligodendrites are involved in myelinating neurons infers that electrical activity does not act to promote myelination, but simply biases which axons are myelinated. Furthermore, this implies that neural activity can act as a selection mechanism to bias axon choice, independent of axon diameter.

There is still debate if electrically active axons act via oligodendrocytes to regulate myelin. There are multiple reasons that this question remains open; it’s very difficult to model and observe this process in vivo, due to genetic manipulation being required, which may interfere with the WT function; the context of a study, i. e. in the CNS may impact on the what is observed and there are often closely related studies often provide conflicting results.

Finally, the three studies evaluated in this essay add to the evidence that electrically active axons can act via oligodendrocytes to influence their own myelination. This process may play a huge role in normal brain development and hebbian learning. In addition, the effects of this activity on OPC proliferation and development may have important clinical implications.