Movement of mitochondria (blue and green) does not correlate with axonal advance (yellow).
Numerous models have been proposed for how axons grow, with a general consensus that the axon cytoskeleton is stationary and growth occurs at the distal tip. In the current study, Miller and Sheetz labeled mitochondria of chicken dorsal root ganglion neurons and followed their movement along the length of the axon. As expected, the mitochondria moved along the cytoskeleton at a rate consistent with kinesin-based transport. Additionally, a subset of mitochondria appeared to dock on the cytoskeleton and then move distally at a substantially slower rate. Kymographs showed that these slow-moving mitochondria moved in a correlated manner, but that the distance between them increased.
Miller and Sheetz conclude that the cytoskeleton, with its docked mitochondria, was being stretched and new material added along its length to prevent thinning. This is similar to viscoelastic stretching that has been reported in Xenopus axons, but in those studies there was no evidence for the addition of extra material.
The team thinks the stretching is a motor-dependent process because it continues even when the growth cone is not moving, indicating that it is not a passive response to growth cone advance. The researchers imagine that a winch-like motor pulls the cytoskeletal fibers toward the distal tip of the axon, creating tension and pulling the cytoskeletal polymers apart. New cytoskeletal subunits would then fill in the gaps, along with docked mitochondria.
Why did the group detect this low velocity transport when others have not? It might be because they looked at organelle movement along the whole length of the axon and analyzed the proximal, middle, and distal regions separately. By contrast, other groups focused on the proximal region where low-velocity movement is absent and everything seems to move by fast transport. 
