Our results suggest that OTR activation in anteromedial BNST induces a vigilance response in which individuals avoid, yet attend to, unfamiliar social contexts. Our results suggest that OTR antagonists may have unappreciated therapeutic potential for stress-induced psychiatric disorders.

Background Oxytocin (OT) is considered to be a stress buffering hormone, dampening the physiological effects of stress. However, OT can also be anxiogenic. We examined acute and long lasting effects of social defeat on OT neurons in male and female California mice. Methods We used immunohistochemistry for OT and c-fos to examine OT neuron activity immediately after defeat (n = 6-9) as well as two (n = 6-9) and ten weeks (n = 4-5) later. We quantified Oxt mRNA with qPCR (n = 5-9). Intranasal OT was administered to naïve and stressed mice tested in social interaction and resident-intruder tests (n = 8-14). Results Acute exposure to a third episode of defeat increased OT/c-fos colocalizations in the paraventricular nucleus (PVN) of both sexes. In the medioventral bed nucleus of the stria terminalis (BNSTmv), defeat increased Oxt mRNA, total OT neurons, and OT/c-fos colocalizations in females but not males. Intranasal OT failed to reverse stress-induced social withdrawal in females, and reduced social interaction behavior in females naïve to defeat. In contrast, intranasal OT increased social interaction in stressed males and reduced freezing in the resident-intruder test. Conclusions Social defeat induces long lasting increases in OT production and OT/c-fos cells in the BNSTmv of females but not males. Intranasal OT largely reversed effects of stress on behavior in males but effects were mixed in females. These results suggest changes in OT sensitive networks contribute to sex differences in behavioral responses to stress.

There is compelling evidence for important sex differences in behavioral and hormonal responses to psychosocial stress. Here we examined the effects of gonadal hormones on behavioral responses to social defeat stress in monogamous California mice (Peromyscus californicus). Three episodes of social defeat induced social withdrawal in intact females but not males. Gonadectomy blocked corticosterone responses to defeat in females and sensitized male corticosterone responses. However, gonadectomy had no effects on social interaction behavior, suggesting that social withdrawal is not dependent on gonadal hormones in the adult California mouse. In contrast, defeat reduced exploratory behavior in the open field test for intact but not castrated males. We also examined the effects of social defeat on social interaction behavior when California mice were raised on corncob bedding, which has estrogenic properties. In this dataset of over 300 mice, we observed that social defeat did not induce social withdrawal when females were raised on corncob bedding. This finding suggests that the use of corncob in rodent studies could mask important sex differences in the effects of stress on brain and behavior. Although gonadal hormones do not affect social withdrawal behavior in adults, our data suggest that hormones may act earlier in development to induce a more resilient social phenotype.

Vasopressin V1a receptors (V1aR) are thought to contribute to the pathophysiology of psychiatric disorders such as anxiety and depression, sparking interest in V1aR as a therapeutic target. Although the global effects of V1aR have been documented, less is known about the specific neural circuits mediating these effects. Moreover, few studies have examined context-specific V1aR function in both males and females. By using the California mouse, we first studied the effects of sex and social defeat stress on V1aR binding in the forebrain. In females but not males, V1aR binding in the bed nucleus of the stria terminalis (BNST) was negatively correlated to social interaction behavior. In females, stress also increasesd V1aR binding in the nucleus accumbens (NAc). Infusions of V1aR antagonist in to the medioventral BNST (BNSTmv) had anxiogenic effects only in animals naïve to defeat. For males, inhibition of V1aR in BNSTmv had anxiogenic effects in social and nonsocial contexts, but for females, anxiogenic effects were limited to social contexts. In stressed females, inhibition of V1aR in the NAc shell had no effect on social interaction behavior, but had an anxiogenic effect in an open field test. These data suggest that V1aR in BNSTmv have anxiolytic and prosocial effects in males, and that in females, prosocial and anxiolytic effects of V1aR appear to be mediated independently by receptors in the BNSTmv and NAc shell, respectively. These findings suggest that males have more overlap in neural circuits modulating anxiety in social and nonsocial contexts than females.

Alterations in LHb ECB signaling may be relevant for development of stress-related pathologies in which LHb dysfunction and stress-coping impairments are hallmark symptoms.