A primary goal of this study was to evaluate the SD and VSD procedures and, most importantly, the potential differences in the resulting defeated phenotype of male and female mice. SD is a valid paradigm to study the effects of social stress on vulnerability to develop mental disorders (Nasca et al. 2019). Its outcomes in mice resemble those of social isolation: depression, anxiety, and PPI disruption (Ago et al. 2014). Although results should be considered cautiously, the SD is a highly translatable experience that resembles in some aspects to social human conflict, which is one of the main sources of stress in modern society. A large body of evidence proves that mice exposed to SD, as humans exposed to stressful experiences, present an increased risk of developing anxiety-like behavior and behavioral alterations such as social withdrawal and anhedonia, which are core features of depression (Montagud-Romero et al. 2015, 2018; Reguilón et al. 2022).

In this study, we employed VSD, a procedure in which the female experimental mice witness an agonistic encounter between two unfamiliar conspecifics. After VSD, female mice were housed with the resident mice separated by a perforated transparent Plexiglas during 24 h. The VSD procedure has been shown to induce a neuroendocrine stress response in both male and female mice. VSD stimulates corticotropin-releasing hormone release and induces a behavioral phenotype similar to that of male defeated mice (Iñiguez et al. 2018; Ródenas-González et al. 2023; Sial et al. 2016). It is established that repeated exposure to SD induces social withdrawal and increased anxiety-like behavior (Giménez-Gómez et al. 2021; Macedo et al. 2018). Consistent with this, we observed that SD male mice exhibited social withdrawal as they spent less time in the interaction zone compared to EXP male mice. On the contrary, VSD did not induce social avoidance, as female mice across experimental conditions displayed similar SI ratios. This is consistent with previous works that show that social stress in females does not affect social interaction behavior (Ródenas-González et al. 2023). These results suggest that although witnessing an aggressive SD encounter is perceived as stressful, female mice do not find interaction with the resident mice aversive or threatening. A plausible explanation for the low percentage of stressed females showing social avoidance is a confounding increased sexual motivation. It has been reported that female mice exhibit a socio-sexual preference toward dominant mice (Parmigiani et al. 2009; Rich and Hurst 1998). It is possible that witnessing and being exposed to the chemosensory signals of a dominant male during cohabitation increases sexual motivation (Haga et al. 2010; Moncho-Bogani et al. 2002). It is important to remark that inter-male territorial aggression, as occurs in SD, is a testosterone-dependent behavior with an ethological relevance on mice mating behavior.

Furthermore, our results showed that SD induced an increase in anxiety-like behavior. As repeatedly reported, SD male mice decreased the time spent in the open arms of the EPM compared to non-stressed mice. A similar effect was observed in female mice, in agreement with our previous studies (Ródenas-González et al. 2023; Yohn et al. 2019). However, while stressed male mice made more total entries and decreased the percentage of open entries, VSD female made fewer number of total entries but increased the percentage of open entries. Therefore, the anxiogenic profile induced by social stress showed sex-dependent particularities. In this line, although we did not find any differences between control and stressed mice in the OF test, female mice showed increased distance traveled, velocity and time in the center, according to their baseline greater novelty-induced anxiety compared to male mice (Johnston and File 1991; Miller et al. 2021).

In agreement with previous studies, we observed a reduction in weight gain as a result of vicarious stress exposure (Sial et al. 2016; Iñiguez et al. 2018). Female stressed mice suffered a significant reduction of gain weight compared to EXP mice, especially at the end of the VSD procedure (days 3 and 4). Compared to SD male mice, the reduction in gain weight in VSD female mice may be more prominent as a result of the 24 h cohabitation with a male resident, in which the source of stress was maintained and thus could affect activity and food intake. In summary, our results highlight the importance of design elements and sex differences in the behavioral outcomes of SD and VSD.

It is important to assess the ecological validity of the social defeat experience. Territorial aggression is a naturally occurring interaction between male mice in proximity. A sexually mature male will attack an individual that enters his territory mainly by relying on chemosensory signals. However, it is important to remark that in naturalistic conditions, defeated mice would have the ability to avoid or escape from potential aggressive opponents. Also, defeated mice adopt a submissive posture that prevents further instigation and attacks from a dominant male. Forced repeated exposures to dyadic defeat with different resident mice in a closed time window does not contribute to the ecological validity of the model (Lyons et al. 2023).

The VSD has been proposed as a suitable stress model for studying stress in female mice. One of the strengths of the VSD model is that, since direct physical interaction is prevented by a physical barrier, the nature of stress is purely psychological. This variation improves the face validity of VSD with respect to the standard SD as a procedure to study the consequences of psychosocial stress. Nevertheless, we acknowledge that the continuous sensory but not physical contact during the 24-h-long cohabitation with the resident mice in VSD female mice is not a representative experience of naturalistic conditions (Lyons et al. 2023). In this sense, some of the design elements of the VSD could differentially affect the defeated phenotype.

The immune response has been shown to be an important contributor to the severity of depressive-like and anxiety behavior induced by SD (Montagud-Romero et al. 2021). Our results show that both SD and VSD have negative behavioral effects which correlate with increased pro-inflammatory signaling. Overall, we found increased striatal IL-6 and hippocampal CX3CL1 protein detection in males compared to females.

To our knowledge, no previous reports have observed these differences. However, the sexual disparity in the production of cytokines in response to stress has been previously reported in both humans and rodents (Bekhbat and Neigh 2018). It has been shown that although females exhibit decreased release of the cytokine in response to stress, they may be more vulnerable to their effects on behavior (Derry et al. 2015; Medina-Rodriguez et al. 2022). The increase in neuroinflammatory markers in non-stressed male mice could be due to the higher social conflict in their home cage compared to female mice.

In rodent models, IL-6 plays a crucial role in the affective and cognitive impairments resulting from social stress (Chourbaji et al. 2006; Niraula et al. 2019). In this study, biochemical analysis showed that both male and female stressed mice exhibited increased striatal IL-6 levels in the striatum. These results are consistent with previous studies showing an increase in IL-6 levels after SD in male mice (Ferrer-Pérez et al. 2019; Reguilón et al. 2022). In addition, this data is in agreement with the study of Takahashi et al. (2017), which also found stressed-induced increased levels of IL-6 in stressed female mice compared to control mice. On the contrary, we previously reported no differences in IL-6 levels in the striatum of female mice immediately after the last VSD (Ródenas-Gonzalez et al. 2023). Nevertheless, differences in the timing of the measure and the experimental design could explain these differences.

Our laboratory and others have previously described the involvement of the CX3CL1/CX3CR1 signal axis in controlling microglial activation after SD stress exposure (Liu et al. 2020; Montagud-Romero et al. 2020). As an important regulator of microglial activity, CX3CL1 function is crucial for the triggering of secondary cascades that regulate the neuroinflammatory response. We found increased CX3CL1 levels in the hippocampus of stressed female mice. The observed sex-specific effects in hippocampal CX3CL1 levels might be a result of a distinct impact of stress on the immune response that differs between sexes. It is important to remark that ovarian hormones affect stress-induced increases in pro-inflammatory cytokines and chemokines as well as sensitivity to their effects on behavior (Finnell et al. 2018). Another possibility is that the differences in CX3CL1 levels observed in stressed female mice are due to the qualitative difference between direct/physical vs. vicarious/emotional stress in male and female mice exposed to SD and VSD, respectively.