Pathfinding is fine without L1–L1 binding.

Brains of mice lacking the cell adhesion molecule L1 are a mess. Failures in neural migration, pathfinding, morphogenesis, and fasciculation result in shortages of cells in various regions and aberrant architecture. But now Itoh et al. (page 145) report that homophilic binding of L1 is not necessary for axonal guidance and neuronal migration in the central nervous system (CNS).

L1 was one of the first neural cell adhesion molecules to be discovered, and its binding partners have been proliferating ever since. Many of those binding partners contact several of L1's many domains, making individual contributions difficult to tease apart. Itoh et al., however, succeed in ablating only a subset of L1's binding interactions via a deletion of L1's sixth Ig domain.

The resulting protein does not bind either to itself or to α5β1 integrin. And yet mice expressing only this variant have brains and spinal cords that look normal. Thus, it seems that L1 homophilic binding, earlier found to mediate neurite outgrowth in vitro, is not needed for this function in the CNS.

When the L1 variant is backcrossed into another mouse strain, the progeny get hydrocephalus. This swelling of the brain results when cerebrospinal fluid is not correctly cleared from the brain ventricles. The link between L1 mutation and hydrocephalus has been noted before, in humans, but its mechanistic basis remains unclear. ▪