Social reward among juvenile mice

Abstract

Mammalian social relationships, such as mother-offspring attachments and pair bonds, can directly affect reproductive output. However, conspecifics approach one another in a comparatively broad range of contexts, so conceivably there are motivations for social congregation other than those underlying reproduction, parental care or territoriality. Here, we show that reward mediated by social contact is a fundamental aspect of juvenile mouse sociality. Employing a novel social conditioned place preference (SCPP) procedure, we demonstrate that social proximity is rewarding for juvenile mice from three inbred strains (A/J, C57BL/6J and DBA/2J), while mice from a fourth strain (BALB/cJ) are much less responsive to social contact. Importantly, this strain-dependent difference was not related to phenotypic variability in exploratory behavior or contextual learning nor influenced by the genetic background associated with maternal care or social conditioning. Furthermore, the SCPP phenotype was expressed early in development (postnatal day 25) and did not require a specific sex composition within the conditioning group. Finally, SCPP responses resulted from an interaction between two specifiable processes: one component of the interaction facilitated approach toward environments that were associated with social salience, whereas a second component mediated avoidance of environmental cues that predicted social isolation. We have thus identified a genetically prescribed process that can attribute value onto conditions predicting a general form of social contact. To our knowledge, this is the first definitive evidence to show that genetic variation can influence a form of social valuation not directly related to a reproductive behavior.

Figure 1:

Conditioning procedure for evaluating the social preferences of juvenile mice. (a) Following 24 h of social housing in a distinct home cage environment, mice were socially isolated in a second novel home cage environment for the next 24 h. The conditioning context (social or isolate housing) was always counterbalanced relative to its pairing with the home cage environment (aspen or paper bedding). Mice were alternately housed in each environment every 24 h for a total of 10 days. Following the last day of conditioning, which entailed 24 h of social isolation, the spatial location and locomotor activity of individual mice were monitored during a 30-min test session. (b) Photograph of socially housed juvenile B6 mice in an aspen environment. (c) Photograph of an isolated juvenile B6 mouse housed in a paper environment.

Figure 3:

Spatial distributions of juvenile mice following social conditioning. Social preference scores of juvenile mice in Fig. 2b′ are replotted as the untransformed spatial data that was taken for each mouse. The continuous (black or white) horizontal line associated with each bar indicates the average time spent in the aspen/paper (peripheral) environments or the central environment by juvenile mice that were not conditioned (see Experiment 1 and Fig. 2a′). The dashed lines that bound each sample mean denote the 95% confidence interval for the respective measurement. The SCPP data illustrated in bar-form are presented as the mean ± SEM.

Figure 2:

Strain-dependent variation in the conditioning responses of juvenile mice. (a–c) Frequency distributions illustrate the number of mice (ordinate) expressing a particular preference score (abscissa) following (a) no conditioning, (b) social conditioning or (c) food conditioning. Mice from all of the strains were included in the distributions. (a′–c′) Juvenile mice did not differentially approach or explore the environments (a′) without conditioning. (b′) Social conditioning resulted in a CPP for mice from three strains but not BALB mice (Tukey’s HSD tests, *P < 0.05 for all post hoc tests comparing BALB with the other strains). (c′) Mice from all of the strains learned the conditioning contingency when food was used as a reward. Preference scores were calculated as the duration spent in the reward-paired (social or food) environment minus the duration spent in the reward-impoverished (isolation or food deprivation) environment. (a″–c″) There were strain-dependent differences in locomotor activity (a″) without conditioning, (b″) with social conditioning and (c″) with food conditioning. There was no difference in exploratory activity between BALB and B6 mice that were tested following the food conditioning procedure (Tukey’s HSD tests, *P < 0.05 compared with all other strains, ^#P < 0.05 compared with the A strain, ⁺P < 0.05 compared with the A and DBA strains). Data in panels a′–c′ and a″–c″ are presented as the mean ± SEM.

Figure 4:

Variation in social conditioning as a function of maternal care, social group characteristics and age. (a) When pups were cross-fostered to a mother of the alternate strain within 12 h of birth juvenile B6 mice, but not BALB mice, expressed SCPP (main effect of strain, *P= 0.02). (b) The strain difference in SCPP persisted when BALB and B6 mice (one per gender for each strain) were conditioned together in the same social group (main effect of strain, *P= 0.004). (c) Male and female B6 mice conditioned in same-sex groups also expressed robust SCPP responses. (d) At PD 25/25, SCPP was apparent for juvenile B6 mice following four conditioning sessions (*effect of conditioning, P= 0.003). Data in each panel are presented as the mean ± SEM.

Figure 5:

Social reward and isolation aversion in juvenile B6 mice. (a) B6 mice differentially approached socially conditioned cues relative to novel cues (*orthogonal contrast for social approach vs. novelty approach, P= 0.03, social approach = time in social environment minus time in novel environment). (b) Unconditioned B6 mice did not respond differentially to the presentation of novel cues in the testing arena (novelty approach = time in novel environment minus time in familiar environment). (c) B6 mice approached novel environments only when the other peripheral compartment of the testing arena contained cues that predicted social isolation (^#orthogonal contrast for isolation aversion vs. novelty approach, P < 0.001, isolation aversion = time in novel environment minus time in isolate environment). Data in each panel are presented as the mean ± SEM.