Sensory processing relies on the formation of specific synapses between sensory neurons and projection neurons that carry coded information to the cortex. The development of these connections involves multiple mechanisms that are not fully understood. Three new papers demonstrate a role for the axon guidance molecule semaphorin 1A (SEMA1A) in axon and dendrite targeting in the olfactory system.

The researchers made use of a common model of sensory development — the Drosophila melanogaster olfactory system — together with some elegant genetic manipulations. This allowed them to pick apart how olfactory receptor neuron (ORN) axons from the antennae and from the maxillary palps that express the same olfactory receptors converge in a single distinct glomerulus in the antennal lobe, where they synapse with projection neuron (PN) dendrites.

Two papers published in Neuron employed genetic mosaic systems, generated by inducing mutations in small groups of neurons within an otherwise wild-type fly, to investigate these events. Luo and colleagues used flies with a mutation in smoothened, which causes occasional loss of the antennae and/or maxillary palps, to show that antennal ORN axons, which arrive early at the antennal lobe, constrain the targeting of late-arriving maxillary palp axons. This indicates that signalling between neighbouring axons is likely to be important for targeting. Both this study and a second paper by Hummel and colleagues examined the role of Sema1a in this process.

Hummel's group showed that Sema1a mutations result in targeting defects: axons of the same type of ORN projected to multiple glomeruli rather than converging on a single glomerulus. Furthermore, overexpression of Sema1a in ORNs disrupted glomerular formation. The requirement for Sema1a was specific to particular classes of ORN, and the targeting defects observed varied between classes. Similarly, Luo and colleagues showed that Sema1a is required for targeting of all maxillary palp, and some antennal, ORN axons. Some Sema1a−/− ORNs developed normally, whereas wild-type ORNs on a mutant background showed targeting defects, indicating that Sema1a acts non-cell-autonomously. Both studies suggest that SEMA1A exerts its targeting effects through the repulsive receptor plexin A.

In a third paper, published in Cell, Luo's group showed that Sema1a is also required for the targeting of axons and dendrites of olfactory PNs to the cortex and antennal lobe, respectively. However, in this case, Sema1a acts cell-autonomously. The cytoplasmic domain of SEMA1A was crucial for this activity, providing a link to downstream signalling mechanisms. The authors observed a gradient of Sema1a expression in PN dendrites, and showed that the levels of SEMA1A direct PN dendritic targeting. The use of molecular gradients in this discrete map is similar to continuous maps formed, for example, in the visual system.

These studies highlight the importance of a single molecule, SEMA1A, in multiple stages of olfactory system development. Future research might concentrate on integrating these findings with other mechanisms known to be important for ORN and PN targeting as well as extending these findings to mammalian systems.