Most living organisms are organized by social structures that facilitate the development of vital functions of the species such as survival (protection from predators), nutrition (collecting and providing food) and the continuity of the species itself (facilitation of the reproduction). Social structures are dynamic and when situations of inter or intra-specific conflicts appear social structures change in order to find the balance. At the clinical level, it is described that the social pattern is significantly altered in many neurological and psychiatric diseases, and this affects the development of daily routines and the quality of the interaction with their counterparts.

From basic research, animal models for neurological and psychiatric diseases try to mimic these behavioural patterns in order to provide experimental models which, with more or less validity, can be used to study the underlying biological and psychological phenomenon. Furthermore, these models provide the opportunity to assess preventive and/or therapeutic strategies. Rat was earlier considered to be the excellent animal model for scientific studies. However, recent advancement in biotechnology has made mice to emerge as the rodent specie of choice in generation of a wide range of genetic and psychopharmacological models [1]. Therefore, this paper reviews the existing literature on social behaviour in mice, including historical evolution of behavioural testings. The paper also assesses the use of behavioural tools in the study of basic traits and social behaviour in different animal models for diseases of interest.

Searches were made in MEDLINE for published studies in the English language from the beginning of the data base (1930) to December 2012 using the keywords: “mice” “social interaction”, “barbering”, “tube test”, “nesting behaviour”, “maternal separation/deprivation” “isolation”, “resident intruder test”, “sexual behaviour”, “maternal behaviour”, “social recognition test”, “olfactory discrimination test” and “playing behaviour”. The number of scientific papers found was 821.

Since 1930 there has been a substantial increase in both basic and translational scientific research on social behaviours in mice. The first studies basically focused on the characterization of social behaviour per se. It was not until the early 60’s that biological approaches and the use of animal (rodents) models to study varied diseases including neurological and psychiatric disorders and their impact on social relationships, were established. In the 90’s the number of scientific publications in different areas (see figure 1) and using a diversity of behavioural tests (see figure 2) experienced an exponential increase. Among the most relevant issues is the use of social interaction test [2], the resident-intruder test to measure aggression and the model of maternal deprivation mainly used to assess the ontogenetic hypothesis of schizophrenia.

Social behaviour is a fundamental characteristic of living organisms and is defined as an interaction between members of the same specie [3]. Normal social structures implicitly involve behaviours with varying degrees of hierarchy or equality, depending on genetic factors

(mouse strain, gender, age, genetic mutations or neurochemical systems) and environmental (conditions of housing, feeding, temperature, isolation, dyads or social group membership, perinatal development.). Social behaviours, as stated before, are dynamic structures, whose deviations by default (i.e. apathy, anhedonia, isolation) or excess (i.e. aggressiveness, irritability, sexual aggression) can reach pathological range and therefore become diagnostic criteria for neurological and psychiatric diseases. In addition, alterations of social behaviours are also symptoms of other organic diseases (see Tables 1 and 2).

When studying social behaviours, several important concepts are taken into account. For instance, in any animal group each subject has social attributes that influence its social relationships with other animals. Related to this, is the concept of sociability which is defined as “the tendency to form cooperative interdependent relationships that allow two-way communication which transcends mere sexual activity” [40]. Nevertheless, the organization of a group is one of the most significant goals for many animal species, and is the basis of social organization [40]. Each species has developed patterns of behaviour and physiological mechanisms that are related to their own social organization and population dynamics [1]. It is, therefore, important to note that results obtained in behavioural paradigms in rats may be different from that obtained in mice [41]. In turn, there are also differences between mice depending on the strains a factor which is usually underestimated (the strain used is the one available rather than the one with the proper behavioural profile) [41].

It’s been told that the social structure of a group depends mainly on the dominance-subordination relationships and/or other attributes such as aggressiveness, competitiveness, individualism, etc. [40]. Thus, dominance is defined as a learned and predictable relationship established between a pair of animals (dyad) where an animal is subordinate by its partner. In this context, ranges, hierarchy and the order of dominance represent the assignment of a numerical value to an animal, in the attempt to describe the relative position of an animal in its social group [1]. On the contrary, there are social structures where agonistic behaviour is observed. These are the result of adaptive actions to solve conflicts arising between two members of the same specie through aggressive behaviour or threats, submissive or passive behaviours and playing behaviours involving physical contact [42]. Aggressive behaviour is a behaviour that causes

Table 1. Reference reviews (1930-2012) of scientific studies about social behaviour in mice. Principal test for measure social behaviour and representative literature.

harm or destruction to animals. In most animal species, males are more aggressive than females. Some authors classify the aggressiveness in aggressive competition, where two individuals within a group compete for the same resource and territorial aggression when aggression is directed to an animal that is considered has invaded a territory [40]. Based on these concepts, early studies in mice also used the social hierarchy testing (SHT) to assess each animal’s social rank within their group. That is, animals were classified in rank 1: dominant, rank 2: activesubordinated, rank 3: subordinated liabilities and rank 4, submissive [10].

In the decade of the 60’-70’s a clear scientific interest for social behaviour is observed and new methods, such as the social interaction test (SIT), are used. The test evaluate both the social and non-social behaviours and distinguish a great variety of behavioural elements such as social investigation (sniffing the anogenital region, the head, or

Table 2. Representative examples of the use of mouse models in studies of social behaviour and their psychological bases, neurochemistry, modeling of neurological disorder, psychiatric and other diseases.

the snout of the partner), follow (following the partner around the cage), squire (following the moving partner while maintaining a constant nose contact with its fur), push under (pushing the snout or the whole anterior part of the body under the partner’s body, and then resting), crawl over (crawling over the partner’s back, crossing it transversally from one side to the other), mutual circle (partners mutually sniffing each other’s anogenital region, while describing tight circles), vibrating tail and aggressive behaviour (including fighting accompanied by biting and blows to the head) as well as social inactivity (lying flat or standing still while maintaining close physical contact with the partner). Among non-social behaviours the SIT evaluates several actions such as exploring (moving around the cage, rearing, sniffing the air, the walls or the sawdust), digging (digging in the sawdust, pushing and kicking it around using the snout and/or both the forepaws and hind paws) and self-grooming (wiping, licking, combing, or scratching any part of own body) [6].

The component of learning and memory in social behaviour can also be studied and evaluated using two behavioural tests: the Social Recognition test (SR) in which the animal must be able to recognize the resident of a cage after having been previously exposed to its smell [9] and the Olfactory discrimination test (ODT) based on the ability of animals to discriminate two different smells (i.e. the smell of almonds versus that of lemons) [10].

Dominance relationships often have an implicit element of aggression that can be evaluated with the social dominance tube test (DTT). In this test two mice of the same genotype and gender are placed at opposite ends of an acrylic tube and released. A subject is declared a “winner” when its opponent backs out of the tube [12].

Home cage sleeping behaviour (SB) evaluates the percentage of animals sleeping huddled in the same quadrant in each cage, and it is known to be impaired in animal models of psychiatric illness such as schizophrenia [12].

In the Dalila effect or barbering, animals show shaved whiskers and hair loss which can be automatically generated by the animal itself or by a cagemate, usually the most dominant, namely the “Dalila mouse”. Scarce research is devoted to this phenomenon and it is speculated whether it is due to dominance, if it involves some level of aggression and, therefore, suffering or pain in the animal that receives it, while other authors consider that at least we can speak of a behaviour derived from social anxiety. To classify the Dalila effect the following scale is used: 0: no barbering, 1: whisker removal or shortening, 2: snout/ face denuding, 3: individual bald patches on head and body, 4: multiple alopecic areas on head and/or body; 5: severe alopecia including complete snout denuding and large pronounced alopecic areas on head and body [16].

On the other hand, there’s a group of social behaviours directly linked to reproductive functions, upbringing and ontogenetic maturation of the litters. They can be measured through a variety of successive events: mating, nesting, maternal care and games. The sexual Behaviour (SB) of the male is quantified by the latency and frequency of mounts, intromissions and ejaculations, while the female measures the level of lordosis [22]. The nesting behaviour (NB) evaluates the ability of the animal to make its nest construction [25] while the parental behaviour (PB) includes measures of protective behaviours, cleaning and food supply from mother to offspring [4]. In the social play behaviour (SPB) the elements under consideration are those of social interaction but in this case the range of age of the individuals is postnatal [39].

In other behavioural paradigms, the experimenter alters the normal conditions of housing of animals in order to induce changes or disrupt social behaviour (induced social behaviour). For instance, in the social isolation-induced (II) the animal is isolated for a month in order to increase its aggressiveness. Thereafter, the territorial aggression towards an intruder mice can be measured by the resident-intruder test (RIT). The latency of first attack, number of attacks and the time of persecution by the aggressive resident are measured [32]. A variant of this test is the Social Disruption stress (SDR) in which, using a similar procedure, the aggressive behaviour of a group of animals living in the same cage against a single attacker is being measured in a 2 minutes test [39]. In studies of ontogeny, maternal deprivation or temporary isolation rearing maternal deprivation (MD) (usually 24 hours, during the ninth postnatal day) is used to model emotional disorders and psychiatric field that allow the hypothesis on the basis ontogenetic diseases such as schizophrenia [36].

Although the first reports on social behaviour in mice were done in the 30’s, it was in the 70’s that clear scientific interest was raised. Soon after, due to the use of genetically engineered mice, interest in this field witnessed an exponential growth. Disorders in social behaviour are characteristic of many mental disorders such as autism, schizophrenia, depression and Alzheimer’s disease. These diseases have been mimicked in animal models of mice. At the moment, basic research in social behaviour is related to gender, aggression and parental relationships. The understanding of the biological and psychological basis of social behaviour is becoming increasingly relevant. Importantly, the consideration of social behaviour in the experimental design of basic and translational research works using murine models may improve the predictive validity of new preventive and/or therapeutic strategies.

This work was supported by Instituto de Salud Carlos III ISC3-PI10/00283.