The growth cones of developing axons navigate through a complex environment, and along the way they read molecular signals that tell them where and when to stop, go and turn. It is possible to simulate growth cone guidance in a dish, by culturing neurons and exposing their growth cones to gradients of attractive or repulsive molecules delivered by a micropipette. The growth cones respond by turning towards or away from the pipette.

Using this deceptively simple technique, Mu-ming Poo and colleagues have assembled a substantial body of work to show that the turning response to an attractive signal, such as brain-derived neurotrophic factor (BDNF) or netrin, requires a minimum level of intracellular cyclic nucleotides, and depends on the formation of calcium gradients across the growth cone. Indeed, the attractive response can be switched to repulsion if either cyclic AMP activity or calcium influx are low. Both attractive and repulsive responses require calcium influx. But how does calcium enter the growth cone? Earlier work has shown that L-type voltage-dependent calcium channels account for part of the influx, and two papers published in Nature now show that transient receptor potential canonical (TRPC) channels must be activated before the L-type channels can kick in.

Working with Xenopus spinal neurons, Wang and Poo found that the depolarization caused by the attractive cue netrin was completely abolished in the presence of a TRPC channel inhibitor, whereas other inhibitors were only partially effective. Knockdown of the Xenopus TRPC1 channel by morpholino antisense abrogated calcium influx and attraction in response to a netrin gradient. This finding, together with pharmacological inhibitor experiments, led the authors to conclude that TRPC1-mediated depolarization and calcium influx are prerequisites for the subsequent activation of L-type channels.

Li and colleagues concomitantly report their study of rat cerebellar granule neurons. A series of inhibitors of voltage-gated calcium and sodium channels did not affect growth cone attraction to gradients of BDNF or glutamate. Meanwhile, a TRPC channel inhibitor abrogated attraction to BDNF without affecting attraction to glutamate.

The authors found that cerebellar granule neurons express three TRPC channels. Inhibition of TRPC3 or TRPC6, by small interfering RNAs or dominant-interfering constructs, abolished growth cone attraction to BDNF (but not to glutamate), whereas knockdown of TRPC1 had no effect. Further pharmacological inhibitor experiments confirmed that, as in Xenopus neurons, attraction to BDNF also required the BDNF receptor TrkB and phospholipase Cγ (PLCγ). An agonist of PLCγ alone was sufficient to attract granule neuron growth cones, depending on extracellular calcium and TRPC channels. The authors propose a model whereby activation of TrkB by BDNF triggers, through PLCγ, the release of calcium from intracellular stores, which provides a sufficient increase in calcium concentrations to activate TRPC channels, allowing an influx of extracellular calcium, depolarization and turning.

TRPCs are part of a large family of TRP channels. TRPs are widely expressed, and have been implicated in many, especially sensory, functions. These two Nature reports indicate that TRPCs have an essential role in axon pathfinding. This exciting idea now awaits confirmation in vivo.