Background

Verbal and nonverbal communications enhance an organism’s ability to establish and maintain a social hierarchy, and thus survive and reproduce. These forms of communication rely on the motivation to interact with fellow conspecifics and the ability to recognize and distinguish between different individuals. In humans, disorders that are commonly characterized by altered social behavior include depression, anxiety, and autism (Saunders and Roy, 1999; White et al., 2009). Although there is a general understanding of symptoms, the etiologies of these disorders, particularly the underlying mechanisms of social impairments and differences, remain unclear (Goldstein and Rosselli, 2003).

Mice offer advantageous models for studying the neurobiology of behavior through similarities to certain aspects of human behavior and physiology (Austin et al., 2004; Cryan and Holmes, 2005). Although a mouse model will not perfectly replicate human disease, overall symptoms can apply when investigating theories regarding genetic or environmental modulation (Crawley, 2007). For example, much of the current preclinical autism research revolves around the diagnostic criteria of abnormal social interaction, impaired social communication and interaction, and repetitive behaviors (Silverman et al., 2010).

A social habituation/dishabituation test enables measurement of two components of rodent social behavior: 1) the interest in a familiar social stimulus shown repeatedly over a short period of time; and 2) the ability to recognize a new social stimulus (Tejada and Rissman, 2012; Ujjainwala et al., 2018). In this paradigm, the mouse used to provide a social stimulus is placed under a wire cup to confine it to a fixed location and to physically separate it from the mouse whose behavior is being examined (“test mouse”). Interest is measured by the amount of time the test mouse spends exploring the stimulus mouse under the wire cup, typically quantified through sniffing. This paradigm can be used to investigate biological and environmental influences on social interest and/or social recognition. For example, our group recently reported that mice genetically deficient for the diazepam binding inhibitor (DBI) protein show a significant reduction in social interest compared to wild-type littermates, with no deficit in social recognition (Ujjainwala et al., 2018). Furthermore, rodents often show sex differences in social interest levels (Tejada and Rissman, 2012; Karlsson et al., 2015), and sex steroid hormones, such as testosterone and estradiol, affect social exploratory behavior. Variations of this protocol have also been used to examine the loss of the oxytocin gene as well as lesions to the CA2 region of the hippocampus and their effect on social recognition (Ferguson et al., 2000; Stevenson and Caldwell, 2014). In addition, this test could be applied to further evaluate the impacts of other environmental and behavioral factors, such as social isolation or gestational exposure to bisphenol A (Wolstenholme et al., 2012; Lander et al., 2017).

Restricted repetitive behaviors are most often associated with autism in humans, but these behaviors are also present in other disorders such as Rett, Fragile X, and Prader-Willi syndromes (Crawley, 2007; McFarlane et al., 2008; Lewis and Kim, 2009). In mice, this phenotype can present in numerous forms, including repetitive grooming, circling, jumping, or other stereotypical movements (Crawley, 2007). Given the overlap between impaired social interaction and repetitive behaviors in these and other neuropsychiatric and neurodevelopmental disorders, a single protocol that enables ready and integrated assessment of multiple phenotypes is advantageous.

Another paradigm commonly used to study social interest and recognition in mice is the 3-chamber test. This test requires a specialized apparatus containing multiple chambers separated by removable doors. This configuration is of higher novelty to a mouse than an environment that mirrors the home cage, potentially inducing higher levels of general exploratory behavior rather than social interaction specifically (Yang et al., 2011). By contrast, the test described in this protocol can be performed in a standard mouse cage used for daily housing in many laboratories. The other materials required can be obtained readily at a low cost.

Blinded analyses significantly enhance experimental rigor, especially regarding manual scoring of behavior, where pre-conceptions may lead to biased evaluation. For any given study, requiring separate researchers to either conduct experiments or perform analyses is not always feasible and can lead to differences in interpretation. Here, we provide a method to randomly code, name, and sort files using Excel. This method allows for one person to conduct the behavioral test and then score the videos in a blinded fashion. This process could be used to anonymize other file types for blinded analysis as well.