By the time the first neurite forms, the group finds, it already marks the location of the axon. No one has seen this association before because they have been looking at stage 2 neurons, which have already formed many neurites. “To find the specific molecular events that are instructions for polarity,” says Dotti, “we should look at the immediate postmitotic stage, even before neurites form.”
At this stage, the centrosome and Golgi apparatus still sit on the opposite end of the cell from the site of cytokinesis. It was here, near the centrosome, where the first lamellipod (and thus the first neurite) formed. Polarized membrane trafficking was aimed at the plasma membrane near the centrosome. This polarization might be a result of mechanical deformation of the membrane by centrosome-organized microtubules (perhaps as a continuation of cytokinetic forces). Or it might be due to the local destabilization of actin filaments (which promotes membrane fusion events) by pericentrosomal proteins.
The first neurite is expected to receive the most external growth cues, given a uniform environment, and thus would be the first to reach any threshold that leads to axon commitment. Cells with two centrosomes—and therefore two early neurites—formed two axons. Loss of centrosomal function, in contrast, blocked axon formation and led only to dendrites.Not long after mitosis, the centrosome changes its position, so it is easy to imagine why scientists have missed the centrosome–axon correlation. Given the centrosome's ability to polarize membrane growth, Dotti wonders whether its rotation within the cell is actually required to make the later, dendrite-forming neurites.