King's College London

Our understanding of the cell behaviours and cytoskeletal requirements of axon formation is largely derived from in vitro models but how these relate to axon formation in vivo is not clear. In vitro, neurons progress through a well-defined multineurite stage to form an axon and both actin and microtubules cooperate to drive the first steps in neurite and axon morphogenesis. However, these steps are not recapitulated in vivo, and it is not clear whether the underlying cell biological mechanisms may differ also. Here, we investigate the mechanisms that regulate axon formation in embryonic zebrafish spinal neurons in vivo. We find microtubule organising centres are located distant from the site of axon initiation, and microtubule plus-ends are not enriched in the axon during axon initiation. Focal F-actin accumulation precedes axon formation, and we find that nocodazole-treated neurons with no detectable microtubules are still able to form nascent axonal protrusions that are approximately 10-μm long, dilated and relatively long-lived. We suggest spinal axon formation in vivo is fundamentally different from axon formation in in vitro models.