In many axonal populations, neurons first project exuberant branches and form excessive synapses with their target, before the surplus contacts are eliminated in an activity-dependent manner. The cessation of such developmental plasticity correlates with the onset of myelination, forming a precise, efficient and stable neuronal network. How the two processes – myelination and branch loss – relate to each other at the level of single axonal branches, however, remains largely unknown. The aim of my PhD project was to gain insights into this important interaction, by studying synapse elimination and myelination at the developing mouse neuromuscular junction (NMJ). Here excessive axonal branches are removed from multiply innervated NMJs in an activity-dependent manner until single innervation is established, while a simple and invariable myelination pattern forms, with individual glial (Schwann) cells being exclusively dedicated to sheathing one single axonal branch, which thereby becomes fully myelinated.
To investigate this phenomenon, I used pharmacological means, as well as genetic manipulations in knock-out and transgenic mice to screen for factors possibly influencing the remodeling process. Local pharmacological inhibition of cholinergic neurotransmission, which blocks the early stages of synapse elimination, also delayed myelination, suggesting that both processes depend either on each other, or on normal neuromuscular activity. I then used markers for labeling myelination progress, and developed a new assay to measure maturation of the nodes of Ranvier. A more detailed characterization of branch-specific myelination patterns demonstrated that indeed, axon branches that are still engaged in synaptic competition and hence remain plastic show delayed myelination and node maturation. However, when I directly correlated the progress of axon dismantling and myelination at single branches, and induced premature myelination by transgenic overexpression of the pro-myelinating factor, neuregulin (NRG1-III), I found surprisingly that myelination of competing branches does not significantly affect pruning, and that such myelinating branches can still be dismantled. On the other hand, genetic ablation of the microtubule severing enzyme Spastin, which results in a delay of axon branch removal after competition, led to increased myelination of competing branches, suggesting the possible effect of competition status via microtubule dynamics and axonal transport on myelination. Together the data points to a model of activity-dependent synapse elimination mediated by branch-specific microtubule dynamics, which in turn modulate transport of myelinating factors such as NRG1-III, leading to a coordinated succession of myelination shortly after competition resolves.
«
In many axonal populations, neurons first project exuberant branches and form excessive synapses with their target, before the surplus contacts are eliminated in an activity-dependent manner. The cessation of such developmental plasticity correlates with the onset of myelination, forming a precise, efficient and stable neuronal network. How the two processes – myelination and branch loss – relate to each other at the level of single axonal branches, however, remains largely unknown. The aim of m...
»