James Rand and his colleagues at the Oklahoma Medical Research Foundation studied the functions of a post-synaptic adhesion protein called neuroligin and the consequences of mutations affecting this protein. There are four neuroligin-encoding (NLGN) genes in people, and mutations disrupting NLGN3 and NLGN4 are associated with a subset of autism cases. Rand and his group used the roundworm C. elegans because of its simple nervous system and its ease of genetic and molecular analysis. Many studies have demonstrated that C. elegans neuronal proteins are structural and functional homologs of the corresponding mammalian proteins.

Rand and his team showed that in wildtype animals, the extracellular domain of neuroligin (the ‘ectodomain’) can be removed (‘shed’) through the action of specific proteases (‘sheddases’). The ectodomain is not degraded, but is stable and can be detected and quantified in tissue extracts. The researchers engineered a version of neuroligin consisting of only a secretable ectodomain. They showed that such a construct is secreted and can diffuse throughout the animal. In addition, this isolated ectodomain can rescue neuroligin function in NLGN1 mutants. These preliminary findings demonstrate that the neuroligin ectodomain can function not only in tethered signaling, but also as a diffusible signaling ligand, thus mediating action at a distance.

In vertebrates, the predominant proteins that interact with neuroligins are neurexins (NRXN) — multi-domain presynaptic transmembrane proteins. C. elegans has one neurexin gene (vertebrates have three), and the NRXN1 protein is likely a major target of the shed neuroligin ectodomain. The researchers characterized the C. elegans NRXN1 gene and its two transcripts. They confirmed that both transcripts are expressed in most or all neurons, and not at all in other tissues. They also engineered a mutant strain that was deficient for both NRXN isoforms, and demonstrated specific deficits in its sensory behaviors.

Finally, C. elegans body muscles are innervated by excitatory motor neurons that release acetylcholine, as well as inhibitory motor neurons that release GABA. This neuromuscular geometry and physiology provided an opportunity to perform a relatively clean, controlled study of the interaction of NRXNs with specific neurotransmitter systems. Rand and his team inquired whether postsynaptic neuroligin in the muscle cells would assemble under both types of synapses, and they demonstrated that the neuroligin was only associated with GABA synapses, not with the acetylcholine synapses in the same cell.