Understanding the pathophysiology of autism, and ultimately the development of treatments for impairments associated with the condition, is greatly dependent on reliable animal models. As described in a paper published in Cell, mice lacking contactin-associated protein-like 2 (Cntnap2) recapitulate the three core symptoms of autism, two of which are normalized by treatment with the antipsychotic drug risperidone.

Autism is currently described as a syndrome characterized by a triad of symptoms: impaired communication, impaired social interaction, and repetitive behaviours and restricted interests. Several common and rare mutations have been associated with the disorder, including variants of the gene encoding CNTNAP2. Peñagarikano et al. set out to investigate how alterations in this protein might contribute to the pathophysiology of autism, in mice lacking Cntnap2 (Cntnap2−/− mice).

The authors showed that juvenile Cntnap2−/− mice emitted fewer ultrasonic vocalizations in response to maternal separation and spent less time interacting with unfamiliar mice then did wild-type mice. As adults, Cntnap2−/− mice did not show the usual preference for another mouse over an inanimate object. These findings point to abnormal social behaviour and communication. The knockout mice also showed evidence of repetitive behaviour: they spent more time grooming and digging than wild-type littermates, and used more rigid behavioural strategies in a water maze test and in a T-maze test. In addition, Cntnap2−/− mice were hyperactive and developed stress-induced seizures after 6 months of age — an interesting finding considering that a human mutation in CNTNAP2 is associated with a syndrome that includes both epilepsy and autism.

How might an absence of CNTNAP2 contribute to this phenotype? In the embryonic brain of wild-type mice, Cntnap2 was preferentially expressed in migratory and post-migratory zones of the developing cortex and in regions that contain migrating interneurons, suggesting that CNTNAP2 may have a role in neuron development and migration. Indeed, 5-bromodeoxyuridine (BrdU) staining and immunohistochemistry experiments revealed an abnormal distribution of neurons in deep cortical layers in the brains of 1-week-old and adult Cntnap2−/− mice, and ectopic neurons in the corpus callosum from postnatal day 14 onwards — findings that are all indicative of deficits in the migration of cortical projection neurons in Cntnap2−/− animals.

A gene co-expression network analysis revealed that Cntnap2−/− is part of a module of functionally related genes, and these genes are more highly expressed in GABAergic than in glutamatergic neurons. The authors therefore also assessed how the absence of Cntnap2 affected GABAergic interneurons and found that Cntnap2−/− mice have fewer interneurons — particularly parvalbumin-positive ones — in cortex, striatum and hippocampus.

the finding that risperidone affects repetitive but not social behaviour suggests that new therapeutics for autism may involve different drugs for different symptoms.

Assessing the functional consequence of these alterations using two-photon calcium imaging, the authors showed that cortical neurons of Cntnap2−/− mice had a highly asynchronous firing pattern, unlike neurons from wild-type mice. As the average firing rate and amplitude did not differ between the two groups, the abnormal firing pattern was probably due to altered network properties.

Finally, the authors investigated whether the drug risperidone, which is used to alleviate hyperactivity, repetitive behaviour and self-injurious behaviour in people with autism, could reverse the behavioural abnormalities in Cntnap2−/− mice. The drug normalized the hyperactivity, excessive grooming and rigid behaviour in the T-maze test but had no effect on social interactions, resembling its effect in humans.

The results from this study are important in several respects. They inform us about possible roles of CNTNAP2 in neuronal development, suggest that neural asynchronization may be important in autism symptoms and indicate that Cntnap2−/− mice may be used as a model for autism. In addition, the finding that risperidone affects repetitive but not social behaviour suggests that new therapeutics for autism may involve different drugs for different symptoms.