Oxytocin–Oxytocin Receptor Systems Facilitate Social Defeat Posture in Male Mice

Social stress has deteriorating effects on various psychiatric diseases. In animal models, exposure to socially dominant conspecifics (i.e., social defeat stress) evokes a species-specific defeat posture via unknown mechanisms. Oxytocin neurons have been shown to be activated by stressful stimuli and to have prosocial and anxiolytic actions. The roles of oxytocin during social defeat stress remain unclear. Expression of c-Fos, a marker of neuronal activation, in oxytocin neurons and in oxytocin receptor‒expressing neurons was investigated in mice. The projection of oxytocin neurons was examined with an anterograde viral tracer, which induces selective expression of membrane-targeted palmitoylated green fluorescent protein in oxytocin neurons. Defensive behaviors during double exposure to social defeat stress in oxytocin receptor‒deficient mice were analyzed. After social defeat stress, expression of c-Fos protein was increased in oxytocin neurons of the bed nucleus of the stria terminalis, supraoptic nucleus, and paraventricular hypothalamic nucleus. Expression of c-Fos protein was also increased in oxytocin receptor‒expressing neurons of brain regions, including the ventrolateral part of the ventromedial hypothalamus and ventrolateral periaqueductal gray. Projecting fibers from paraventricular hypothalamic oxytocin neurons were found in the ventrolateral part of the ventromedial hypothalamus and in the ventrolateral periaqueductal gray. Oxytocin receptor‒deficient mice showed reduced defeat posture during the second social defeat stress. These findings suggest that social defeat stress activates oxytocin-oxytocin receptor systems, and the findings are consistent with the view that activation of the oxytocin receptor in brain regions, including the ventrolateral part of the ventromedial hypothalamus and the ventrolateral periaqueductal gray, facilitates social defeat posture.

Social stress has deteriorating effects on various psychiatric diseases. In animal models, exposure to socially dominant conspecifics (i.e., social defeat stress) evokes a species-specific defeat posture via unknown mechanisms. Oxytocin neurons have been shown to be activated by stressful stimuli and to have prosocial and anxiolytic actions. The roles of oxytocin during social defeat stress remain unclear. Expression of c-Fos, a marker of neuronal activation, in oxytocin neurons and in oxytocin receptor-expressing neurons was investigated in mice. The projection of oxytocin neurons was examined with an anterograde viral tracer, which induces selective expression of membranetargeted palmitoylated green fluorescent protein in oxytocin neurons. Defensive behaviors during double exposure to social defeat stress in oxytocin receptor-deficient mice were analyzed. After social defeat stress, expression of c-Fos protein was increased in oxytocin neurons of the bed nucleus of the stria terminalis, supraoptic nucleus, and paraventricular hypothalamic nucleus. Expression of c-Fos protein was also increased in oxytocin receptor-expressing neurons of brain regions, including the ventrolateral part of the ventromedial hypothalamus and ventrolateral periaqueductal gray. Projecting fibers from paraventricular hypothalamic oxytocin neurons were found in the ventrolateral part of the ventromedial hypothalamus and in the ventrolateral periaqueductal gray. Oxytocin receptor-deficient mice showed reduced defeat posture during the second social defeat stress. These findings suggest that social defeat stress activates oxytocin-oxytocin receptor systems, and the findings are consistent with the view that activation of the oxytocin receptor in brain regions, including the ventrolateral part of the ventromedial hypothalamus and the ventrolateral periaqueductal gray, facilitates social defeat posture. (Endocrinology 159: [763][764][765][766][767][768][769][770][771][772][773][774][775]2018) S ocial stress can cause a variety of disorders in humans (1). In animal models, social stress induces anhedonia and social avoidance (2) and evokes a variety of responses in the neuroendocrine and autonomic nervous systems (3)(4)(5)(6). Among social stress-related stimuli, social defeat, exposure to a socially dominant conspecific, is an ethologically relevant form of stressful stimulus (7)(8)(9) and is thought to play an important role in various psychiatric diseases (10)(11)(12). However, the brain circuitry for social defeat is not well understood.
When confronting dominant conspecifics, rodents show species-specific submissive behaviors such as defeat posture or freezing behaviors. Defeat posture in mice is an upright posture with the head angled upward toward the dominant conspecific, the ears retracted, and the forepaws extended (2).
Studies with stimulation, lesions, or mapping of neuronal activity have suggested that the ventromedial hypothalamus (VMH) and midbrain periaqueductal gray (PAG) play an important role in defensive behaviors in response to various threatening stimuli (13)(14)(15)(16)(17). However, the VMH and PAG contain heterogeneous neurons (18), and functions of each subtype of the VMH and PAG neurons remain poorly understood. Neural circuits underlying expression of defeat behaviors are not clear.
Here, we first examined whether social defeat stress activates oxytocin neurons in the hypothalamus and BNST in mice and then whether social defeat stress activates oxytocin receptor-expressing neurons in various brain areas including the VMH and PAG. We also used an anterograde viral tracer to investigate whether oxytocin neurons of the hypothalamic paraventricular nucleus (PVN) innervate the VMH and PAG. Finally, we examined the effects of deficiency in the oxytocin receptor on defensive behaviors during social defeat stress.
The animals were housed in rooms with controlled temperature (20°C to 24°C) and humidity (40% to 70%) under a 12-hour light/dark cycle (lights on at 7:30 AM to 7:30 PM). Food and water were available ad libitum. All animal procedures were approved by the Institutional Animal Experiment Committee of Jichi Medical University and were in accordance with the Institutional Regulations for Animal Experiments and Fundamental Guidelines for Proper Conduct of Animal Experiments and Related Activities in Academic Research Institutions under the jurisdiction of the Ministry of Education, Culture, Sports, Science and Technology.

Social defeat stress
A cage for social defeat stress consisted of two adjoining chambers of equal size (91 3 260 3 128 mm) separated by a perforated transparent partition-wall. Aggressive CD1 mice were selected as reported previously (40). An aggressive CD1 mouse was kept in one of the chambers for 2 days. C57BL/6J wild-type mice [10 to 14 weeks old; n = 8 (control) or 6 (social defeat)] and oxytocin receptor-Venus knock-in transgenic mice [29 to 31 weeks old; n = 5 or 6 (control) or 6 (social defeat)] were used as test animals. As the procedure for habituation to test cages, animals were placed singly in a new test chamber and kept there for 1 day with food and water available ad libitum. This procedure was repeated for 3 consecutive days. On the day after the habituation procedure, the mouse was placed as the intruder in the chamber of the aggressive resident male CD1 mouse and kept there for 10 minutes. The behavior of the intruder mouse during confrontation with the resident aggressor was videotaped. The duration of defeat posture (an upright posture with the belly exposed toward the resident) and duration of freezing behavior (no movements except for respiration with four paws on the ground) were measured. Immediately after exposure to the aggressor mouse, the intruder mouse was placed in the chamber next to the chamber of the aggressor mouse for another 100 minutes. The chambers between the intruder and aggressor were separated by a transparent perforated wall. A single nonstressed control mouse was placed in a test chamber, with another nonstressed control mouse placed in the adjoining chamber. Control mice were kept there for 110 minutes.
In experiments with oxytocin receptor-deficient mice, an oxytocin receptor-deficient mouse or its wild-type control sibling mouse [15 to 16 weeks old; n = 8 (wild-type) or 8 (oxytocin receptor-deficient)] was placed in the home cage of an aggressive CD1 mouse for 10 minutes. Two weeks after the exposure, the mouse was again exposed to a different aggressive CD1 mouse for 10 minutes. Behavior during the 10-minute exposures was recorded and analyzed by an observer who was blinded to the genotype.
Immunocytochemical detection of c-Fos protein, oxytocin-immunoreactive neurons, and oxytocin receptor-expressing neurons For detection of neurons activated after social defeat stress, mice were exposed to aggressive CD1 mice for 10 minutes and placed in a chamber separated by a perforated transparent partition and next to the chamber of the aggressor mouse so that mice could smell and see the aggressive mouse. Because expression of c-Fos protein is known to peak between 90 and 120 minutes after stress, 100 minutes after the end of the social defeat stress, mice were anesthetized with tribromoethanol (200 mg/kg body weight, intraperitoneal injection; WAKO Pure Chemical Industries Ltd, Osaka, Japan) and were perfused transcardially with heparinized saline (20 U/mL) followed by 4% paraformaldehyde in 0.1 M phosphate buffer (pH, 7.4) for 15 minutes. Brains were removed from the skulls, postfixed in 4% paraformaldehyde solution overnight, and transferred to 30% sucrose solution in 0.1 M phosphate buffer until they sank.
The brains were then frozen on dry ice and stored at 280°C until processing for immunocytochemical examination.
Brain sections were cut coronally at an interval of 30 mm with a freezing sledge microtome and then processed for immunohistochemical detection of c-Fos and oxytocin or Venus as described previously (33,34,41).
Sections for immunocytochemical detection of oxytocin were incubated with guinea pig anti-oxytocin antibody (diluted 1:1,000,000; T-5021; Peninsula Laboratories LLC, San Carlos, CA) at 4°C for 2 days, biotinylated anti-guinea pig IgG (diluted 1:750; BA-7000; Vector Laboratories) at room temperature for 2 hours, and avidin biotinylated horseradish peroxidase complex (diluted 1:50; Elite ABC kit; Vector Laboratories) at room temperature for 30 minutes. Oxytocin immunoreactivity was visualized as a brown cytoplasmic precipitate with DAB or as a black precipitate with DAB and nickel sulfate.
For double immunocytochemical detection of oxytocin fibers and oxytocin receptorexpressing cells, oxytocin receptor-Venus knock-in mice (21 weeks old; n = 4) were anesthetized with tribromoethanol and perfused transcardially as described previously. Brain sections were incubated with a guinea pig anti-oxytocin antibody (diluted 1:200,000) and a rat monoclonal anti-GFP antibody (diluted 1:1000) in 0.1 M phosphate buffer containing 0.3% Triton X-100 and 5% normal goat serum at 4°C for 2 days, followed by incubation with fluorescent dye-conjugated secondary antibodies, Alexa Fluor ® 568 goat antiguinea pig IgG (diluted 1:500; A11075; Thermo Fisher Scientific Inc, Waltham, MA) and Alexa Fluor ® 488 donkey anti-rat IgG (diluted 1:500; A21208; Thermo Fisher Scientific Inc.) at 4°C for 1 day. Sections were observed with a fluorescent microscope (BZ-X700; Keyence Co. Ltd., Osaka, Japan) by making use of the Quick Full Focus function. Section images were captured at different focal planes with 1-mm intervals and were integrated into one image. The oxytocin-immunoreactive (oxytocin-ir) fiber area was measured in each brain area by using BZ-X Analyzer software (Keyence Co. Ltd.) and was expressed as micrometer squared per millimeter squared.
The numbers of sections for detection of c-Fos protein and oxytocin were four for the BNST, 11 for the supraoptic nucleus (SON), and 12 for the PVN in each animal. The intervals between sections were 90 mm.
The numbers of sections for detection of c-Fos protein and Venus and for detection of oxytocin and Venus were three and three, respectively, for the anterior olfactory nucleus, eight and six to eight for the cingulate cortex, nine and eight for the insular cortex, eight and six to eight for the piriform cortex, six and six for the lateral septum, seven and six to seven for the cortical amygdala nucleus, seven and six to seven for the medial amygdala nucleus, 13 and 13 for the paraventricular thalamus, seven and seven for the posterior intralaminar thalamic nucleus, 11 and seven to nine for the arcuate nucleus, six and six for the ventrolateral part of the VMH, five and five for the posterior hypothalamic nucleus, 10 and nine to 10 for the dorsal raphe nucleus, 10 and 12 for the median raphe nucleus, and 11 and 10 or 11 for the ventrolateral PAG in each animal. The  intervals between sections were 360 mm for the insular cortex, 630 mm for the piriform cortex, and 90 mm for other areas. Some additional sections containing the VMH or PAG were processed for detection of c-Fos protein and oxytocin or for detection of oxytocin and Venus. Each brain region was identified according to the coordinates of the Paxinos mouse brain atlas (43).

Visualization of oxytocin fibers from the PVN by use of an anterograde viral tracer
Oxytocin-Ires-Cre knock-in mice were crossed with Rosa-CAG-LSL-tdTomato-WPRE::deltaNeo knock-in reporter mice (39), which have a loxP-flanked STOP cassette sequence and a CAG promoter-driven tdTomato sequence in the Gt(ROSA)26Sor locus. We confirmed that expression of tdTomato in the hypothalamic PVN was confined in oxytocin-ir neurons, suggesting that within the PVN of oxytocin-Ires-Cre mice, a Cre recombinase activity was expressed selectively in oxytocin-synthesizing neurons.
GFP tagged with a palmitoylation signal (palGFP) has been shown to be sorted to the plasma membrane and has been used to trace neuronal fibers efficiently (44,45). To label oxytocin fibers by selective expression of palGFP in oxytocin neurons of the PVN, the following adeno-associated virus (AAV) vectors were applied to oxytocin-Ires-Cre mice. The AAV vectors contained a sequence for expressing palGFP under the control of the CAG promoter selectively in oxytocin cells expressing Cre recombinase activity (AAV-CAG-FLEX-palGFP-WPRE).
AAV vectors were produced using the AAV Helper-Free System (Agilent Technologies, Inc., Santa Clara, CA) and purified on the basis of published methods (46,47). Briefly, HEK293 cells were transfected with an AAV vector plasmid that included a sequence for CAG-FLEX-palGFP-WPRE, pHelper, and pAAV-RC (serotype DJ; purchased from Cell Biolabs Inc, San Diego, CA) using a standard calcium phosphate method. Three days later, transfected cells were collected and suspended in artificial cerebrospinal fluid (124 mM NaCl, 3 mM KCl, 26 mM NaHCO 3 , 2 mM CaCl 2 , 1 mM MgSO 4 , 1.25 mM KH 2 PO 4 , and 10 mM D-glucose). After four freeze-thaw cycles, the cell lysate was treated with benzonase nuclease (Merck KGaA, Darmstadt, Germany) at 45°C for 15 minutes and centrifuged two times at 16,000g for 10 minutes. The supernatant was used as the virus-containing solution. Digital droplet polymerase chain reaction was performed to measure the viral titer using a QX200 droplet reader (Bio-Rad Laboratories, Inc., Hercules, CA). We used the following TaqMan MGB probe set for WPRE (Thermo Fisher Scientific Inc): 5 0 -TGCTCCTTTTACGCTATGTGGATA-3 0 for WPRE-Forward; 5 0 -CATAAAGAG-ACAGCAACCAGGATTT-3 0 for WPRE-Reverse; and 5 0 -VIC-CTGCTTTAATGC-CTTTGTAT-MGB-3 0 for the WPRE-TaqMan probe. The AAV vector was stored at 280°C in small aliquots until the day of the experiment.
Oxytocin-Ires-Cre knock-in mice (19 to 25 weeks old; n = 4) were anesthetized with tribromoethanol. AAV-CAG-FLEX-palGFP-WPRE (serotype: DJ; 1.3 3 10 9 virus genome per microliter, 0.3 to 0.5 mL) was injected into the left side of the PVN (stereotaxic coordinates: 0.8 mm caudal to the bregma, 0.25 mm lateral to the middle, and 4.7 mm below the skull) through a cannula (100 mm in inner diameter) attached to a microsyringe over a period of 5 minutes. The cannula was left in situ for another l0 minutes after completion of the infusion. Two to four weeks after the infusion, mice were anesthetized with tribromoethanol and perfused with 4% paraformaldehyde, and the brains were processed for immunocytochemical detection of palGFP and oxytocin. Immunoreactivity of oxytocin and palGFP was visualized with guinea pig anti-oxytocin antibody (diluted 1:200,000; 2-day incubation at 4°C), rabbit anti-GFP antibody (diluted 1:1000; 2-day incubation at 4°C; code598; Medical & Biological Laboratories Co. Ltd., Nagoya, Japan), Alexa Fluor ® 568 goat anti-guinea pig IgG (diluted 1:500; 1-day incubation at 4°C), and Alexa Fluor ® 488 goat anti-rabbit IgG (diluted 1:500; 1-day incubation at 4°C; A11034; Thermo Fisher Scientific Inc). Sections were observed using a confocal microscope (TCS SP5; Leica Microsystems, Wetzlar, Germany). The numbers of sections examined in each animal were two for

Specificity of antibodies
The specificity of antibodies for oxytocin and c-Fos protein was described previously (33). In brief, oxytocin antisera have minimal cross-reactivity with Arg 8 -vasopressin (manufacturer's data sheets). No immunostaining was observed in sections in which the antisera were preabsorbed with oxytocin. Preabsorption with Arg 8 -vasopressin did not impair immunostaining. Preabsorption of the c-Fos antibody, which was raised in a rabbit against a synthetic peptide (amino acid residues four to 17 of human c-Fos), with the synthetic peptide human c-Fos (4-17) blocked immunostaining. No immunoreactive staining with the GFP monoclonal and rabbit polyclonal (34) antibodies was observed in brain sections of wild-type mice.

Data analysis
Data are expressed as means 6 standard error of the mean. Data were analyzed by the Mann-Whitney U test, repeatedmeasures two-way analysis of variance (ANOVA), or t test. P , 0.05 was considered statistically significant.

Results
We first examined whether oxytocin neurons in the BNST and the hypothalamus were activated after exposure to aggressive CD1 mice. During the 10-minute acute exposure to aggressors, the intruder mice received attacks (88.3 6 15.0 seconds) and showed social defeat posture (95.0 6 35.9 seconds) and freezing behavior (280.6 6 60.0 seconds), suggesting that exposure to the aggressor caused social defeat stress in the present experimental conditions.
After social defeat stress, the percentages of oxytocinir neurons expressing c-Fos immunoreactivity were significantly increased in the BNST, hypothalamic PVN, and SON compared with those in the nonstressed control group, suggesting that social defeat stress activated oxytocin neurons in the BNST, PVN, and SON ( Oxytocin acts on the oxytocin receptor to induce activation of neurons. We therefore examined whether oxytocin receptor-expressing neurons are activated after social defeat stress. The oxytocin receptor is expressed in various brain regions including social brain areas (34).
We examined the expression of c-Fos protein in oxytocin receptor-expressing neurons. Oxytocin receptorexpressing neurons were identified by immunocytochemical detection of Venus protein in the oxytocin receptor-Venus knock-in mouse. We investigated the anterior olfactory cortex, cingulate cortex, insular cortex, piriform cortex, lateral septum, cortical amygdala, medial amygdala, paraventricular thalamus, posterior intralaminar thalamic nucleus, arcuate nucleus, ventrolateral part of the VMH, posterior hypothalamic nucleus, dorsal raphe nucleus, median raphe nucleus, and ventrolateral part of the PAG, where the oxytocin receptor is expressed abundantly (34).
The oxytocin receptor-Venus knock-in mice received attacks from the aggressive CD1 residents (50.0 6 6.1 seconds) and showed defeat posture (66.7 6 14.2 seconds) and freezing behavior (252.6 6 48.0 seconds) during exposure to the aggressive CD1 mice, suggesting that the exposure to aggressive residents caused social defeat stress. The percentages of c-Fos-positive oxytocin receptor-expressing cells were significantly higher after social defeat stress in the insular cortex, lateral septum, cortical amygdala, medial amygdala, paraventricular thalamus, posterior intralaminar thalamic nucleus, ventrolateral part of the VMH, and ventrolateral part of the PAG than in the nonstressed control group We then examined whether these brain regions containing oxytocin receptor-expressing cells receive innervation of oxytocin-ir fibers. We found oxytocin-ir fibers in oxytocin receptor-expressing regions, consistent with reported data for rats (48). We also quantified oxytocin-ir fiber areas and found considerable amounts of oxytocin-ir fibers in the ventrolateral part of the PAG and paraventricular thalamus and moderate amounts of oxytocin-ir fibers in the lateral septum, medial amygdala, posterior intralaminar thalamic nucleus, arcuate nucleus, ventrolateral VMH, posterior hypothalamic nucleus, and dorsal raphe (Fig. 4). Among brain regions in which oxytocin receptorexpressing neurons were activated, the VMH and PAG are thought to be involved in defensive behaviors during social defeat stress (15). We then examined whether oxytocin neurons innervate the ventrolateral VMH and the ventrolateral PAG, where activated oxytocin receptor-expressing neurons were localized.
Oxytocin-ir fibers were closely apposed to oxytocin receptor-expressing cells in the ventrolateral VMH and ventrolateral PAG [ Fig. 5(a) and 5(b)]. In mice that received social defeat stress, oxytocin-ir fibers were found close to c-Fos-positive cells in the ventrolateral VMH and PAG [ Fig. 5(c) and 5(d)].
We then examined whether oxytocin-synthesizing neurons in the hypothalamic PVN innervate the ventrolateral VMH and ventrolateral PAG. To label oxytocin fibers originating from PVN oxytocin neurons, an anterograde viral tracer, AAV-CAG-FLEX-palGFP-WPRE, was injected into the hypothalamic PVN of oxytocin-Ires-Cre knock-in mice. palGFP is sorted into the plasma membrane and thus is able to label neuronal fibers effectively (44). We found coexistence of palGFP and oxytocin immunoreactivity in the hypothalamic PVN [ Fig. 6(a)]. Almost all of the palGFP-positive cells showed oxytocin immunoreactivity (93.8% 6 2.1%; n = 3), whereas the percentage of oxytocin-ir cells expressing palGFP immunoreactivity was 73.2% 6 1.2% (n = 3). palGFP-positive fibers coexpressed oxytocin immunoreactivity in the BNST. The findings suggest that palGFP immunoreactivity can be used as a marker of oxytocin neurons.
Moderate amounts of palGFP-positive fibers were found in the ventrolateral part of the VMH, where oxytocin receptor-expressing cells were located, whereas oxytocin-ir fibers were very sparse in the dorsomedial part of the VMH [ Fig. 6(b) and 6(c)]. Considerable amounts of palGFP-positive fibers were found in the ventrolateral part of the PAG, where oxytocin receptorexpressing cells were located [ Fig. 6(b) and 6(c)], suggesting that PVN oxytocin neurons innervate the ventrolateral VMH and ventrolateral PAG.
We lastly examined behaviors of oxytocin receptordeficient mice during social defeat stress to clarify physiological roles of the oxytocin receptor in the control of defensive behaviors during social defeat stress. Social defeat stress was applied twice because it has been shown that experience of social defeat induces social memory and facilitates defensive behavior in the second social interaction (49)(50)(51)(52) and that the oxytocin receptor plays a role in the acquisition of social memory (33).
The durations of attacks received were not significantly different between wild-type mice and oxytocin receptor-deficient mice (45.7 6 10.0 and 57.4 6 7.3 seconds for the first stress and second stress, respectively, in wild-type mice; 54.8 6 7.2 and 66.1 6 10.3 seconds for the first stress and second stress, respectively, in oxytocin receptor-deficient mice). The duration of defeat posture was significantly increased during the second stress compared with that during the first stress in wildtype mice but not in oxytocin receptor-deficient mice [repeated-measures two-way ANOVA (F 1,31 = 1.960; P = 0.183; no significant effects of genotype), (F 1,31 = 11.371; P = 0.0046; significant effects of number of times for stress), (F 1,31 = 5.750; P = 0.031; significant interaction)] [ Fig. 7(a)]. As a result, the duration of defeat posture during the second defeat stress was significantly shorter in oxytocin receptor-deficient mice than in wild-type mice [P , 0.05 vs wild-type mice (t test: t 14 = 2.22; P = 0.043)], suggesting facilitative roles of oxytocin-oxytocin systems in social defeat posture.
On the other hand, the duration of freezing behavior during the second stress was increased compared with that during the first stress, consistent with the results of a previous study (49), and there was no significant difference between wild-type mice and oxytocin receptordeficient mice [repeated-measures two-way ANOVA (F 1,31 = 1.780; P = 0.203; no significant effects of genotype), (F 1,31 = 5.736; P = 0.0312; significant effects of number of times for stress), (F 1,31 = 0.0739; P = 0.790; no significant interaction)] [ Fig. 7(b)]. Considering this augmentation of freezing behavior in both wild-type mice and oxytocin receptor-deficient mice, selective blockade of facilitation of social defeat posture in oxytocin receptor-deficient mice cannot be explained by a possible role of the oxytocin receptor in the control of general social arousal.

Discussion
Social defeat stress induces species-specific defeat posture in animals. Concerning neural mechanisms of social defeat, social defeat stress has been shown to activate the ventrolateral part of the VMH and PAG (15). On the other hand, oxytocin neurons in the hypothalamus are activated by a variety of stressful stimuli (26). Here, we suggest that PVN oxytocin neurons innervate the ventrolateral VMH and PAG and that social defeat stress activates the PVN oxytocin neurons and oxytocin receptor-expressing neurons in various brain regions, including the ventrolateral VMH and PAG, although a causal relationship between activated oxytocin and  oxytocin receptor-expressing cells remains to be determined. Oxytocin receptor-deficient mice showed a deficit in facilitation of social defeat posture during social defeat stress. Our results suggest that social defeat stress activates oxytocin-oxytocin receptor systems in the brain and that activation of oxytocin-oxytocin receptor systems facilitates social defeat posture during social defeat stress.
In the current study, oxytocin neurons in the hypothalamus and BNST were activated after social defeat stress. At present, neural mechanisms for activation of oxytocin neurons are unclear. Hypothalamic oxytocin neurons receive innervations from various neurons (53), including noradrenergic neurons that coexpress prolactin-releasing peptide (PrRP) in the medulla oblongata (26,54). Activation of oxytocin neurons has been shown to be mediated by PrRP/noradrenergic neurons (54,55). It is tempting to speculate that PrRP/ noradrenergic neurons in the medulla oblongata also play a role in activation of oxytocin neurons after social defeat stress. Further studies should be performed to test this hypothesis.
Social defeat has been shown to have long-lasting effects on behaviors, including anhedonia and social avoidance (2). On the other hand, application of oxytocin has been shown to have anxiolytic, prosocial, and social memory-enhancing actions and to reverse social defeatinduced reduction in social interactions (56,57). In the current study, oxytocin receptor-deficient mice showed no facilitation of social defeat posture but showed facilitation of freezing behavior to the second defeat stress similar to that in wild-type mice, suggesting that endogenous oxytocin-oxytocin receptor systems have facilitative actions selectively on species-specific defeat posture during social defeat stress. This action of oxytocin cannot be explained by the general anxiolytic, prosocial, and social memory-enhancing actions of oxytocin because there were no significant changes in freezing behavior. Application of oxytocin has been shown to augment or induce stress responses depending on experimental conditions, including sex difference (37,56). The results of the current study further show the importance of activation of endogenous oxytocin-oxytocin receptor systems for induction of social stress-related behavior.
Brain regions of the oxytocin receptor responsible for induction of social defeat posture remain to be clarified. It has been shown that pharmacological inhibition of neuronal activity of the ventrolateral VMH by use of nonselective expression of an inhibitory artificial receptor increases locomotion during social defeat stress (15). These results suggest that the ventrolateral VMH plays an inhibitory role in motor activity during social defeat stress. In the current study, we found that the PVN oxytocin neurons projected to the ventrolateral VMH, that social defeat stress activated PVN oxytocin neurons and ventrolateral VMH oxytocin receptor-expressing neurons, and that deficiency in the oxytocin receptor suppressed facilitation of the social defeat posture, the upright standstill posture, during social defeat stress. Thus, it is interesting to speculate that the PVN oxytocin neuronventrolateral VMH oxytocin receptor-expressing neuron system contributes to species-specific defensive behavior during social defeat stress. Further studies are needed to clarify this point.
The ventrolateral VMH has also been shown to be involved in fighting and mating behaviors in male mice (18,(58)(59)(60)(61)(62)(63)(64). The ventrolateral VMH contains heterogeneous neurons. Some neurons in the ventrolateral VMH are activated during fighting. Selective activation of neurons expressing estrogen receptor 1 (ESR1) evokes attack or mating behavior, whereas inhibition of ESR1-expressing cells (63) or destruction of progesterone receptor-expressing cells (60) suppresses intermale aggression, suggesting involvement of neurons expressing ESR1 or progesterone receptor in the control of aggression or mating. Although relationships between neurons expressing ESR1 and/or progesterone receptor and those expressing the oxytocin receptor remain to be clarified, it is tempting to speculate that oxytocin receptor-expressing neurons form a subpopulation different from ESR1 or progesterone receptor-expressing neurons in the ventrolateral VMH to induce defeat posture. It is also possible that a single population of neurons that express ESR1, progesterone receptor, and oxytocin receptor mediates facilitation of aggression, mating, and defeat posture depending on the situation.
The results of the current study also suggest that social defeat stress activated oxytocin receptor-expressing neurons in the ventrolateral part of the PAG. The ventrolateral part of the PAG has been shown to mediate passive defensive behaviors in threatening situations (65)(66)(67). Stimulation of the ventrolateral PAG induces freezing (68) or immobility (69) behavior, whereas inhibition of PAG function has been shown to reduce defensive freezing behaviors (70). Inhibition of the ventrolateral PAG has also been thought to increase motor activity (66). In the current study, deficiency in the oxytocin receptor reduced facilitation of social defeat posture, which is a motionless posture. It is thus possible that the oxytocin receptor in the ventrolateral PAG plays a role in the social defeat posture. In the current study, we also found oxytocin-ir fibers and activation of oxytocin receptor-expressing cells in the insular cortex, cortical amygdala, medial amygdala, lateral septum, and thalamus. All of these areas have been shown to be involved in social interaction. Physiological roles of the oxytocin receptor in these areas during social defeat stress remain to be determined.
During acute social defeat stress, animals show species-specific submissive defeat postures. Expression of submissive posture can induce suppression of aggressive behaviors by the dominant conspecific animal (71). It is interesting to speculate that oxytocin facilitates submissive defeat posture in an imminent confrontation with dominant conspecifics to reduce aggression of the dominant conspecifics and, as a result, increases adaptive values in group-forming animals.
In conclusion, social defeat stress-activated oxytocin neurons and oxytocin receptor-expressing neurons in various brain regions, including the ventrolateral VMH and PAG, and blockade of the oxytocin receptor systems attenuated facilitation of the social defeat posture during social defeat stress. The newly defined PVN-ventrolateral VMH and PVN-ventrolateral PAG oxytocin pathways may lead to a new therapeutic target for stress-related disorders.