Neurons decide which of their many branches will become the axon through an internal cytoskeletal rhythm centered in the cell body, according to a study published in Nature. The finding challenges a decades-old assumption that this critical decision is guided mainly by growth cones reading external chemical cues.
Every neuron eventually settles on one branch, or neurite, to serve as its axon, the long fiber that sends signals onward, while the rest become dendrites that receive input. This asymmetry, known as neuronal polarity, underlies the one-way flow of information through brain circuits. Exactly how a young neuron — which starts out with several seemingly equivalent neurites — picks just one to become the axon has remained an open question.
The prevailing view held that growth cones, the sensory tips at the ends of neurites, detect guidance molecules such as the growth factors NT3 and TGFβ and relay that information through signaling pathways to determine polarity. But the researchers note this model struggles to explain why neurons in the brain, an environment saturated with growth-promoting factors, reliably produce only a single axon. It also does not explain why, even when two neurites of the same neuron are exposed to the same growth-promoting substrate simultaneously, only one becomes the axon.
Using live imaging in cortical pyramidal neurons in cultured cells and in living tissue, alongside genetic loss-of-function experiments, optogenetic tools and localized cytoskeletal manipulation, the researchers instead traced the decision to a soma-based oscillator. Before polarization, neurites take turns extending and retracting, with only one neurite growing at any given moment while the others pull back.
An actin wave that primes an axon
This rhythmic behavior, the study shows, is driven by periodic actin branching at the soma that depends on the actin-related protein 2/3 (ARP2/3) complex. That branching activity reshapes a network of actin and myosin spanning the cell, producing a wave that first causes neurites to retract before propagating into a single neurite tip.
When the wave reaches that tip, it locally relaxes the contractile actomyosin network, allowing a transient, microtubule-based protrusion to form. This event biases the chosen neurite toward becoming the axon. Once the neuron exits this oscillatory phase, the selected neurite can keep extending independently of ARP2/3 and becomes resistant to retraction, while ongoing actomyosin activity suppresses axon formation in the remaining neurites, steering them toward a dendritic fate instead.
The researchers describe this as a cell-intrinsic, cytoskeleton-driven mechanism that does not depend on the direction of external guidance signals. By generating a single axon reliably, regardless of the surrounding environment, the mechanism helps explain how the brain's neurons consistently establish the correct wiring architecture needed for unidirectional signal flow across neural circuits.