The growth of dendrites (top) is impaired by blocks in the secretory pathway (right), but axons (bottom) fare just fine.


The Feng shui of Golgi arrangement helps give neurons their distinctive shapes, say Bing Ye, Ye Zhang, Yuh Nung Jan, and colleagues (University of California, San Francisco, CA). Dendrites, the group finds, rely on the secretory pathway for growth more than axons do.

A young neuron is a mass of tiny neurites, one of which becomes an axon while the others form dendrites. To find the blueprints for these very different architectures, Jan's group identified mutations that hampered the growth of dendrites but not axons. Several of their mutants impaired the transport of secretory membrane vesicles from the ER to the Golgi.

The secretory block reorganized the small, isolated stacks of Golgi —called outposts—that are found predominately in dendrites. When outposts inched forward, dendrites survived and extended. When outposts retreated, branches followed. Both extension and retraction required the outposts. “If we damage the outposts, we find that branches become stationary,” says Jan.

Axons, by contrast, were much more indifferent to secretion blocks. “Golgi outposts are not the only thing,” says Jan. “Vesicles can go directly from Golgi in the soma to the [axonal] growth cone.” And material can also be taken in from one part of the plasma membrane and inserted directly into another. These pathways might be enough for growing axons, which require less morphological remodeling. “In axons,” says Jan, “changes happen mostly at the growth cone. In dendrites, it happens all over.”

The final sum of these changes is different for each class of neurons. “Neuronal cell types are distinguished by their arboreal morphology,” says Jan. “There are hundreds, maybe thousands of different shapes. So how do we generate these? It's not random.” The authors now plan to compare outpost distribution in these different classes of neurons.

Only a few other dendrite shapers are known, including the NeuroD transcription factor and a neural activity–dependent kinase. Jan and colleagues will now examine whether these proteins control outpost distribution.


Ye, B., et al.