Too much shootin1 (green) results in a neuron with a surplus of axons (red; arrowheads).
Although it sports an axon at one end and a fringe of dendrites at the other, a neuron starts out symmetrical. The imbalance develops because the branches, or neurites, that sprout from a youthful cell compete with each other. The fastest-growing extension typically morphs into the axon, and the stragglers become dendrites. Researchers have identified some of the molecular events that dictate which neurite transforms into an axon—the enzyme PI 3-kinase accrues in the winning branch, for example. But they do not understand what sets up the asymmetry.
Toriyama et al. identified one candidate, a new protein they dubbed shootin1, whose levels soar in axons and during cell polarization. Early in a neuron's development, shootin1 quantities fluctuate in neurites. But they shoot up and remain high in the neurite that will become the axon, while plummeting in the losers. When the team genetically modified cells to overexpress shootin1, multiple axons sprouted; slashing shootin1 production blocked cell polarization.
To determine how the protein interacts with PI 3-kinase, the team suppressed shootin1 levels. The kinase no longer accumulated in growing axon tips. The findings suggest that shootin1 spurs neurons to polarize by controlling the location of PI 3-kinase.
The researchers hypothesize that random fluctuations in shootin1 levels unleash a positive feedback loop that promotes polarization. A neuron actively transports shootin1 into the neurites, and the protein diffuses back to the cell body. If by chance one neurite gets a little more shootin1 than the other branches, it will outgrow them. As the lucky neurite extends, shootin1's diffusion time stretches. The protein would thus remain in the neurite longer, propelling even more growth. 
