Setting up the brain's neural circuitry during development requires the correct construction and function of the synapses. New work by Frédérique Varoqueaux and colleagues reveals that neuroligins, a family of postsynaptic cell adhesion molecules, considered essential for building synapses, are in fact necessary only for subsequent synaptic maturation and function.

Neuroligins are crucial for proper brain function, shown by the fact that mutations to the human genes are associated with autism and mental retardation. Recent findings have indicated a functional role for neuroligins in synaptogenesis. For example, RNAi knockdown of neuroligins in cultured neurons has been shown to lead to the formation of fewer synapses, whereas their overexpression leads to more synapses. In addition, neuroligins are thought to be important for synapse maturation and function.

To investigate the in vivo role of these important molecules, Varoqueaux's team generated neuroligin-null mice. Of the four neuroligins in the mouse (NL1–4), NL4 is not detectable in newborn mice and is found only at extremely low levels in adult mice. Therefore, the team concentrated on NL1, 2 and 3, and made individual, dual and triple knockouts of the genes.

Mice lacking any one of the three genes seemed generally normal, as did the dual knockouts, although the latter showed reduced reproduction and drastically impaired parenting. However, mice lacking all three neuroligins died from breathing difficulties approximately 1 day after birth. NLs 1–3 are coexpressed in almost all neurons, and only the triple knockouts were neonatally lethal, indicating that there is a large degree of functional redundancy. The group therefore focused their studies on the triple knockout phenotype.

Histological analysis of the brains of NL1–3 triple knockout mice revealed that knocking out all three genes had apparently no effect on the gross cytoarchitecture of the brain. The density and organization of neurons in the olfactory bulb, hippocampus, cortex and brainstem was normal. Given the predicted role of neuroligins in synaptogenesis, this was rather surprising. However, closer inspection of the synapses of neurons of the brainstem respiratory network (which controls breathing) confirmed their normal ultrastructure. Furthermore, expression levels of integral components of the synaptic contacts were comparable with wild-type mice.

However, expression levels of synaptic communication proteins such as vesicle markers were reduced, indicating aberrant synaptic function. Indeed, electrophysiological analysis of the brainstem respiratory network neurons revealed that GABA (γ-aminobutyric acid)-mediated/glycinergic and glutamatergic synaptic transmission was markedly decreased. Therefore, the loss of synaptic maturation and function rather than loss of synapses per se is likely to be the cause of neonatal lethality in these mice.

The reason for the discrepancies between this work and previous in vitro studies is not yet clear. However, the creation of these mice, including the individual and dual knockouts, provides an important and valuable resource for continued in vivo investigations into neuroligin biology and synapse function.