Sticking to their routes

Abstract

The central nervous system contains axonal wiring systems that would send telephone engineers running for their lives. This complexity of nerve routing and connections is established during development through mechanisms that remain poorly understood. In tackling this problem, neurobiologists have gained some understanding through the analysis of axon-associated molecules, particularly membrane receptors. Traditional cell-adhesion molecules (CAMs) are known to control nerve fasciculation and growth, whereas an ability to control axon guidance directly has not been demonstrated. By contrast, several other receptor classes with essential axon guidance functions have recently been identified. Thus, we have had CAMs controlling the growth and bundling of nerves on the one hand, and distinct receptors governing actual nerve guidance on the other. However, things can never remain that simple! A recent paper by Castellani et al.1xAnalysis of the L1-deficient mouse phenotype reveals cross-talk between Sema3A and L1 signaling pathways in axonal guidance. Castellani, V. et al. Neuron. 2000; 27: 237–249Abstract | Full Text | Full Text PDF | PubMedSee all References1 has now thrown the proverbial spanner in the works by showing that L1, a classic immunoglobulin superfamily CAM, modulates the chemotactic guidance of murine corticospinal (CS) axons.CS tract axons grow from the cortex to the ventral pyramids of the medulla, whereupon they hit the rostral spinal cord, cross over the midline (decussate) and turn dorsally into the cord. It has been shown previously that this decussation is aberrant in L1-deficient mice, mimicking aspects of diseases found in humans with L1 mutations. Although L1 traditionally controls axon fasciculation and axon growth, the authors give a novel explanation for these decussation defects: L1 can also control axonal chemorepulsive responses. They show that the ventral spinal cord releases a chemorepellent for cultured CS axons, but that L1-deficient axons fail to respond to it. Repulsion is blocked by antibodies to neuropilin-1, the receptor for Sema3A, a soluble chemorepellent. A variety of experiments then confirm that the spinal cord repellent activity is almost certainly Sema3A, and that L1 ectodomains bind in a stable complex with neruropilin-1. Significantly, Sema3A is a likely physiological cue at the CS decussation point, as it is expressed in the ventral spinal cord, beginning precisely at the medulla–spinal cord junction. In a final, fascinating twist, soluble L1 ectodomains were shown not only to block the repulsive response of CS axons to Sema3A, but to turn it into attraction! This did not occur in L1-deficient axons, indicating that L1 is necessary (1) for the Sema3A repulsive response, and (2) for transduction of a contrary, attractive response to Sema3A.This is the first time that an Ig superfamily CAM has been shown to influence a chemotactic signalling pathway directly. How L1 does this remains to be clarified, as does its ability to ‘switch’ the response to Sema3A. Nevertheless, as suggested by Castellani et al., the ability of L1 to influence Sema3A at the CS decussation might be influenced by axon fasciculation, itself controlled in part by homophilic L1 interactions. Therefore, it is beginning to appear that CAMs such as L1 are true multitasking molecules, shaping axon adhesion as well as guidance. The implications are fascinating, to say the least.