In most mammals, including humans, the postnatal acquisition of normal social and nonsocial behavior critically depends on interactions with peers. Here we explore the possibility that mixed-group housing of mice carrying a deletion of Nlgn3, a gene associated with autism spectrum disorders, and their wild-type littermates induces changes in each other’s behavior. We have found that, when raised together, male Nlgn3 knockout mice and their wild-type littermates displayed deficits in sociability. Moreover, social submission in adult male Nlgn3 knockout mice correlated with an increase in their anxiety. Re-expression of Nlgn3 in parvalbumin-expressing cells in transgenic animals rescued their social behavior and alleviated the phenotype of their wild-type littermates, further indicating that the social behavior of Nlgn3 knockout mice has a direct and measurable impact on wild-type animals’ behavior. Finally, we showed that, unlike male mice, female mice lacking Nlgn3 were insensitive to their peers’ behavior but modified the social behavior of their littermates. Altogether, our findings show that the environment is a critical factor in the development of behavioral phenotypes in transgenic and wild-type mice. In addition, these results reveal that the social environment has a sexually dimorphic effect on the behavior of mice lacking Nlgn3, being more influential in males than females.

Deletions in the 15q11.2 region of the human genome are associated with neurobehavioral deficits, and motor development delay, as well as in some cases, symptoms of autism or schizophrenia. The cytoplasmic FMRP-interacting protein 1 (CYFIP1) is one of the four genes contained within this locus and has been associated with other genetic forms of autism spectrum disorders (ASD). In mice, Cyfip1 haploinsufficiency leads to alteration of dendritic spine morphology and defects in synaptic plasticity, two pathophysiological hallmarks of mouse models of ASD. At the behavioral level, however, Cyfip1 haploinsufficiency leads to minor phenotypes, not directly relevant for 15q11.2 deletion syndrome or ASD. A fundamental question is whether neuronal phenotypes caused by the mutation of Cyfip1 are relevant for the human condition. Here, we describe a synaptic cluster of ASD-associated proteins centered on CYFIP1 and the adhesion protein Neuroligin-3. Cyfip1 haploinsufficiency in mice led to decreased dendritic spine density and stability associated with social behavior and motor learning phenotypes. Behavioral training early in development resulted in alleviating the motor learning deficits caused by Cyfip1 haploinsufficiency. Altogether, these data provide new insight into the neuronal and behavioral phenotypes caused by Cyfip1 mutation and proof-of-concept for the development of a behavioral therapy to treat phenotypes associated with 15q11.2 syndromes and ASD.

The etiology of autism spectrum disorders (ASDs) is a complex combination of genetic and environmental factors. Neuroligin3, a synaptic adhesion protein, and cytoplasmic FMR1 interacting protein 1 (CYFIP1), a regulator of protein translation and actin polymerization, are two proteins associated with ASDs that interact in neurons in vivo. Here, we investigated the role of the Neuroligin3/CYFIP1 pathway in behavioral functioning and synapse formation in mice and found that it contributes to motor learning and synapse formation in males. Similar investigation in female mice revealed an absence of such phenotypes, suggesting that females are protected against mutations affecting this pathway. Previously, we showed that the social environment influences the behavior of male mice. We extended this finding and found that the transcriptome of wild-type mice housed with their mutant littermates, lacking Neuroligin3, differed from the transcriptome of wild-type mice housed together. Altogether, these results identify the role of the Neuroligin3/CYFIP1 pathway in male mouse behavior and highlight its sensitivity to social environment.

11Social behavior in animals is an adaptive process influenced by environmental factors and direct and 12 indirect genetic effects. Indirect genetic effects (IGEs) include mechanisms by which individuals of 13 particular genotypes can influence the behavioral phenotypes and genotypes (via modulated patterns of 14 gene expression) of other individuals with different genotypes. In groups of adult mice, IGEs can be 15 unidirectional, from one genotype to the other, or bidirectional, resulting in a homogenization of the 16 behavioral phenotypes within the group. Critically, it has been theorized that IGEs constitute a large 17 fitness target on which deleterious mutations can have pleiotropic effects, meaning that individuals 18 carrying certain behavior-altering mutations can impose the fitness costs of those mutations on others 19 comprising the broader social genome. Experimental data involving a mouse model support the existence 20 of these IGE-amplified fitness losses; however, the underlying biological mechanisms that facilitate these 21 remain unknown. In a mouse model of IGEs, we demonstrate that the Major Urinary Protein 20 22 pheromone, also called Darcin, produced by mice lacking the adhesion protein Neuroligin-3 acts as a 23 vector to deleteriously modify the social behavior of wild-type mice. Additionally, we showed that lack of 24 social interest on the part of Neuroligin-3 knockout mice is independent of their environment. These 25 findings reveal a new role for mammalian pheromones in mediating the externalization of social deficits 26 from one individual to others comprising the population through IGEs. 27 Author Summary