Two mutations in the gene encoding neuroligin 3 (NLGN3) — a postsynaptic adhesion molecule — have been associated with autism spectrum disorders (ASDs). How these mutations elicit behavioural changes remains unknown, not least because it has been difficult to ascribe common behavioural and synaptic deficits to these mutations in mice. Now, Rothwell, Fuccillo et al. reveal that both mutations impair striatal circuits in mice and that these impairments may promote repetitive behaviours, a feature of ASDs.

Credit: Jennie Vallis/NPG

The two mutations linked to ASDs are the deletion of NLGN3 and the R451C point mutation, which reduces NLGN3 levels. Nlgn3-knockout (Nlgn3-KO) mice and mice harbouring the point mutation (Nlgn3-R451C mice) exhibit social interaction deficits, typically found in ASDs, but these deficits differ between the two lines. Thus, to search for an ASD-relevant behavioural change that was common to both lines, the authors examined another ASD symptom domain, namely repetitive behaviours.

with training, the motor routines of Nlgn3-KO animals in the rotarod task became less variable than those of wild-type mice

The authors tested Nlgn3-KO and Nlgn3-R451C mice in the accelerating rotarod task, which requires mice to develop and maintain repetitive motor activity: the mutant mice learnt the task more quickly than wild-type animals and managed to complete trials involving greater 'terminal' speeds of rod rotation. Moreover, with training, the motor routines of Nlgn3-KO animals in the rotarod task became less variable than those of wild-type mice. Together, these data suggest that the Nlgn3 mutations promote the development of repetitive behaviours.

To examine which brain areas are involved in generating the repetitive behaviours, the authors created mice in which Nlgn3 could be conditionally knocked out (Nlgn3-cKO mice) with Cre recombinase expression and crossed these animals with mice expressing Cre in specific neuronal subpopulations. One brain area that was targeted using this approach was the striatum, which is known to influence the acquisition of repetitive and stereotyped behaviours. Indeed, knocking out Nlgn3 expression in striatal medium spiny projection neurons (MSNs) that express D1 dopamine receptors (D1-MSNs), but not in D2-MSNs, recapitulated the motor phenotypes of the Nlgn3-KO and Nlgn3-R451C mice, suggesting that the Nlgn3 mutations target D1-MSN function.

Quantitative RT-PCR showed that Nlgn3 mRNA was expressed at high levels in D1-MSNs compared with D2-MSNs in the nucleus accumbens (NAc) but was expressed in low levels in both cell types in the dorsal striatum. Furthermore, the injection of an adeno-associated virus expressing Cre into striatal subcompartments in Nlgn3-cKO mice revealed that deletion of Nlgn3 specifically in the NAc enhanced performances on the rotarod task. This suggests that D1-MSNs in the NAc — which is typically associated with reward-related behaviours rather than motor function — have an important role in the observed repetitive behaviour.

Finally, the authors examined the effect of NLGN3 on synaptic function in the NAc. They could not detect any changes in spontaneous miniature excitatory postsynaptic currents in D1-MSNs or D2-MSNs in Nlgn3-KO mice; however, they found that the frequency of spontaneous miniature inhibitory postsynaptic currents in NAc D1-MSNs was decreased in Nlgn3-KO and Nlgn3-R451C mice. Moreover, the ratio of the peak GABA receptor-mediated current to the peak AMPA receptor-mediated current was reduced in D1-MSNs in Nlgn3-KO mice. Thus, the loss of NLGN3 seems to be associated with an alteration in the balance between synaptic excitation and inhibition in this cell type.

This study reveals that ASD-associated NLGN3 mutations promote repetitive behaviour in mice and pinpoints a specific synaptic deficit in the striatal circuitry that underlies this phenotype.