Vasopressin Signaling Buffers Synaptic Metaplasticity in a Sex-specific Manner

Social animals prefer to live in groups and transfer social signals to share their physical or psychological distress and seek comfort [1, 2]. During this process, social buffering is an essential concept in social networks. This refers to the phenomenon wherein the presence of social partners reduces the physiological stress response in a distressed animal [3]. This phenomenon exists in humans, non-human primates, rodents, birds, and even zebrafish [3]. Stress triggers a comprehensive biological response in the entire organism [4], however, research into social buffering has primarily focused on behavioral changes, endocrine changes, the activation of specific cell ensembles, changes in neural circuits, or changes in intrinsic activity [5, 6]. In recent years, stress-induced changes in glutamatergic synaptic plasticity, or metaplasticity, have garnered increasing attention.

In a previous study, Sterley et al. [6] found that stress can prime short-term potentiation at glutamate synapses onto corticotropin-releasing hormone (CRH) neurons in the hypothalamic paraventricular nucleus (PVN) but can be buffered by the presence of a naïve partner in female mice. In detail, short-term potentiation at glutamate synapses was recorded in brain slices prepared immediately after animals were exposed to a 5-min foot-shock, as well as in slices prepared 24 h after foot-shock. This indicates that acute stress induces robust and persistent changes in glutamate synapses on CRH^PVN neurons. Intriguingly, when the female animals were returned to their home with a same-sex littermate following a 5-min foot-shock, the changes of glutamate synapses in the stressed animals were buffered by their social partners. However, the buffering of synaptic changes after stress was not found in male animals. These findings provide evidence that stress-induced changes in glutamatergic synaptic plasticity, or metaplasticity, are socially buffered in female mice [4]. Even so, the mechanisms underlying social buffering of stress-dependent metaplasticity are still unclear.

Vasopressin neurons are adjacent to CRH neurons in the PVN and release vasopressin from their somatodendritic domain, allowing them to locally regulate PVN neurons and play a central role in numerous social functions [7]. Therefore, in a recent study, Loewen et al. [8] from the same laboratory hypothesized that local vasopressin signaling mediates the social buffering of behavior and metaplasticity. The behavioral and synaptic consequences of stress were evaluated by measuring stress-induced grooming behavior and short-term potentiation at glutamate synapses. They demonstrated vasopressin's critical role in regulating the social buffering of behavioral and synaptic metaplasticity and produced several fascinating results (Fig. 1).

Fig. 1.

To evaluate whether the presence of a naïve social partner attenuates foot-stock stress-induced grooming and synaptic changes in female mice, both single-housed and pair-housed female mice were subjected to foot-shock and returned to their home cage with or without a partner for 30 min. Loewen et al. [8] found that pair-housed mice spent significantly less time grooming, suggesting that the presence of a social partner buffers the foot-shock-induced stress. In addition, Loewen et al. [8] confirmed the social buffering of synaptic metaplasticity after stress and found that vasopressin signaling is a potential mediator of the synaptic changes of CRH^PVN neurons in social buffering in a precise and local manner. The role of another important neuropeptide released from the PVN, oxytocin, in the buffering of metaplasticity is still unknown. It would be interesting to investigate whether oxytocin buffers synaptic changes and grooming behavior in the future.

To evaluate whether the presence of a naïve social partner attenuates foot-stock stress-induced grooming and synaptic changes in female mice, both single-housed and pair-housed female mice were subjected to foot-shock and returned to their home cage with or without a partner for 30 min. Loewen et al. [8] found that pair-housed mice spent significantly less time grooming, suggesting that the presence of a social partner buffers the foot-shock-induced stress. Notably, they excluded the possibility that the difference was due to variations in locomotor activity or interruption by a naïve partner. Furthermore, using whole-cell recordings from CRH^PVN neurons, Loewen et al. [8] found that pair-housed mice exhibited significantly decreased short-term potentiation compared with single-housed mice. These results are consistent with their previous work on social buffering and synaptic changes after stress [6].

The authors next asked whether vasopressin signaling contributes to the social buffering of stress-induced grooming behavior and short-term potentiation at glutamate synapses on CRH^PVN neurons. To answer this question, they inhibited vasopressin receptor 1a on CRH^PVN neurons by injecting the V1aR antagonist SR-49059 prior to stimulating the mice via foot-shock. The mice were then returned to their home cage with an unstressed social partner. Notably, there were no significant differences in the time spent on grooming between the animals with or without SR, suggesting that V1aR signaling does not contribute to the social buffering of stress-induced grooming. However, when the authors tested the effect of SR on CRH^PVN neurons from vehicle- and SR-injected mice, they found that the short-term potentiation in SR-injected animals was significantly increased compared to vehicle-injected animals, indicating that V1aR signaling is involved in the social buffering of stress-induced short-term potentiation changes.

As reported in their previous study, investigative behavior contributes to the consequences of social buffering [6]. To exclude the possibility that injection of the V1aR antagonist SR affects investigative behavior, they measured the investigative behaviors and found that investigative behavior, locomotor activity, surveying, and rearing activity were not affected by V1aR blockade. Taken together, these results suggest that vasopressin signaling contributes to the social buffering of stress-induced short-term potentiation but is not required for stress-induced grooming or investigative behavior.

To further confirm the role of vasopressin signaling in the social buffering of stress-induced short-term potentiation, ex vivo experiments were performed. According to the findings from in vivo experiments, Loewen et al. [8] expected that vasopressin itself could buffer stress-induced short-term potentiation in CRH^PVN neurons in the absence of a naïve unstressed partner. To demonstrate this, brain slices from single-housed stressed animals were incubated in artificial cerebrospinal fluid (aCSF) or aCSF + vasopressin following short-term potentiation measurement. As expected, vasopressin-incubated slices exhibited significantly lower short-term potentiation than control aCSF-incubated slices, demonstrating the role of vasopressin in the social buffering of short-term potentiation in vivo. Having shown that inhibition of V1aRs prevents the social buffering of short-term potentiation in vivo, confirmatory ex vivo studies using a V1aR antagonist were performed. Recordings of short-term potentiation were made from CRH^PVN neurons in brain slices incubated with SR or vasopressin alone. Consistent with the findings from in vivo experiments, slices incubated with both SR and vasopressin presented significantly stronger short-term potentiation than vasopressin-incubated slices. As well, these experiments showed that vasopressin itself also affects the stress-induced short-term potentiation in female mice in a sex-specific manner, consistent with their previous findings [6]. Furthermore, to understanding the acute effect of vasopressin on CRH^PVN neurons, vasopressin was directly applied to CRH^PVN neurons using a puff pipette rather than incubation. No significant differences were found in the firing frequency of action potentials, membrane potential, or spontaneous excitatory postsynaptic currents after acute application vasopressin on CRH^PVN neurons from stressed female animals. This suggests that, instead of causing direct changes in plasticity, vasopressin specifically affects metaplasticity.

Because decreased NMDAR function is necessary for the induction of short-term potentiation in response to stress [9], the authors then asked whether the social buffering of stress-induced short-term potentiation via vasopressin depends on NMDAR- and AMPAR-mediated currents. Using brain slices from stressed female mice with vasopressin incubation or acute vasopressin application, they found that vasopressin treatment significantly reduced the AMPA/NMDA ratio, and that acute vasopressin application increased the NMDA current. To answer whether vasopressin increases NMDAR function and short-term potentiation directly through neuronal V1aR-mediated activity or indirectly via astrocytes, G-protein-mediated signaling in postsynaptic CRH^PVN neurons was disrupted by GDPβs. Acute vasopressin application after the disruption of G-protein signaling in postsynaptic CRH^PVN neurons did not decrease the AMPA/NMDA ratio. Taken together, these findings suggested that increased NMDA function in CRH^PVN neurons after vasopressin application contributes to reduced short-term potentiation in social buffering.

In summary, Loewen et al. [8] confirmed the social buffering of synaptic metaplasticity after stress and found that vasopressin signaling is a potential mediator in regulating synaptic changes of CRH^PVN neurons in social buffering in a precise and local manner. Their study fills a critical gap in our knowledge of the social buffering of stress-induced metaplasticity and provides a novel mechanism underlying synaptic buffering after social interaction. Although Loewen et al. [8] have made excellent progress in uncovering the mechanisms underlying synaptic buffering, several intriguing questions arise and need to be addressed in the future. The first such question is the mechanism underlying sex differences. In the current study and their previous work [6, 8], buffering of synaptic changes and the effects of vasopressin signaling are sex-specific. Therefore, future studies should first determine why synaptic buffering is more prominent in females than males. Future studies clarifying what induces the differences between males and females would help explain the sex-specific manner of the buffering effects of vasopressin on synaptic changes. Sex hormones have been highlighted for their role in sexual dimorphism in brain diseases and behavior [10]. Therefore, one possible direction may be to investigate the role of estrogen, a crucial hormone for female reproduction, in the sex differences of vasopressin-mediated buffering of metaplasticity. Another interesting question is what is upstream in mediating the vasopressin release in CRH^PVN neurons during social buffering and why the vasopressin-mediated buffering of synaptic changes does not affect grooming behavior. Besides, the investigation of vasopressin in social buffering also raises future questions regarding another important neuropeptide released from the PVN, oxytocin. Oxytocin is also closely related to social function and responses to stressors, playing a critical role in many social behaviors [10]. Therefore, it would be interesting to investigate whether oxytocin buffers synaptic changes and grooming behavior. Unraveling these mechanisms underlying social buffering may help bridge the gap of understanding linking social behaviors and neuronal activity.

Conflict of interest

The authors declare no competing financial interests.