The effects of social loss and isolation on motivational behavior and dopamine and oxytocin receptor expression in female prairie voles

Abstract

In humans, grief is characterized by intense sadness, intrusive thoughts of the deceased, and intense longing for reunion with the deceased. Human fMRI studies show hyperactivity in emotional pain and motivational centers of the brain when an individual is reminded of a deceased attachment figure, but the molecular underpinnings of these changes in activity are unknown. Prairie voles (Microtus ochrogaster), which establish lifelong social bonds between breeding pairs, also display distress and motivational shifts during periods of prolonged social loss, providing a model to investigate these behavioral and molecular changes at a mechanistic level. To study the behavioral changes associated with motivation, a novel odor preference test was used to assess social-seeking behavior, and a sucrose preference test was used to assess non-social, reward-driven motivation. Females that lost a male partner investigated partner- and stranger-associated cues significantly more than females that lost a female cagemate or remained intact with a female cagemate or a male partner. Western blotting revealed significant increases of dopamine receptor type 1 (DRD1) and oxytocin receptor protein content in specific brain regions in response to the loss of distinct social relationships. Such effects included an increase in DRD1 in the medial preoptic area of the hypothalamus (mPOA) in females experiencing loss of a male partner compared to all other conditions. Pharmacological antagonism of DRD1 in the mPOA blocked the loss-associated increase to partner odor but not stranger odors. This reveals a novel dopamine-mediated mechanism for partner-seeking behavior during periods of partner loss in female prairie voles.

1. Introduction

For many species, stable social relationships provide multiple beneficial effects, positively contributing to an individual’s well-being and quality of life. Likewise, the loss of significant social relationships may have deleterious effects on an organism. In humans, the biological and psychological effects of social loss have been researched for many years. Specifically, partner and spousal loss are associated with cardiovascular-associated mortality (1,2), increased stress responsivity (3,4), and many psychological comorbidities (5–7). Spousally bereaved individuals are particularly at risk for the development of complicated or prolonged grief disorder, compared to individuals who have lost other significant attachment relationships (8). Complicated grief is distinguished from other mental illnesses by prolonged yearning for the deceased (9,10). This shift in motivational state has been studied neurologically using fMRI studies in humans, indicating heightened activity within the ventral tegmental area, nucleus accumbens (NAc), anterior cingulate cortex (ACC), and insular cortex (IC) in response to reminders of a romantic partner or other deceased loved ones (11–14). However, little is known about these shifts at the molecular level.

The prairie vole (Microtus ochrogaster) provides a new animal model of social loss (15). Prairie voles are a socially monogamous rodent species which establish social attachments between male-female breeding pairs, known as pair bonds, a form of attachment present in approximately 5% of mammals (16–19). Pair bonded prairie voles display selective affiliative behaviors towards a social partner and stranger-directed aggression (20,21). These behaviors and processes are mediated by multiple molecular systems, such as those involving corticotropin-releasing hormone, oxytocin (Oxt) and dopamine (DA) (22–31). Male prairie voles display various behavioral outcomes following the loss of a pair-bonded female partner, which may also be regulated by these neurochemical systems, such as prolonged distress, altered stress coping responses, and increases in partner seeking behaviors (15,32–38). Few studies have investigated this form of social loss in females, but it has been shown that females also display increased stress-related behaviors following partner loss and social isolation (36,39). As such, the present study aims to identify the molecular underpinnings of shifts in partner-seeking behavior in female prairie voles behavior during social loss.

Only a single study has assessed behavioral and neuroendocrine outcomes in female prairie voles experiencing partner loss in a peripartum period, highlighting a role for the corticotropin-releasing hormone system in the brain to regulate the increased stress state observed in females during this period (39). Here, we used the novel odor preference test (OPT) to determine motivational states associated with non-social and social reward (15,40,41). This behavioral test presents the animal with two naturally rewarding odorants, social- and food-scented bedding. This allows for the study of the motivation for socially associated cues without a physical reunion, as well as the motivation for naturally rewarding, non-social stimuli. We also used a sucrose preference test to assess anhedonic state independent of social information following social loss. We also assessed the concentration of Oxt receptor (Oxtr), DA receptor type 1 (DRD1), and DA receptor type 2 (DRD2) in the medial preoptic area (mPOA), hippocampal area CA2 (dorsolateral portion; CA2), paraventricular nucleus of the hypothalamus (PVN), NAc, ACC, and IC. These regions have been implicated in various aspects of social attachment and loss, such as motivation and reward, emotional pain processing, social memory, social behavior, and human grief symptomology (for review, see (42)). Our immunological and behavioral studies revealed an association between partner odor investigation behavior and DRD1 expression in the mPOA during periods of partner loss. Therefore, we manipulated DRD1 signaling in the mPOA directly using a DRD1-selective antagonist to determine whether this diminished the loss-induced social seeking behavior.

2. Methods

2.1. Animals

Subjects were female captive-bred, descended from wild-caught Illinois prairie voles. Voles were weaned on postnatal day 21(±3) and housed in same-sex non-sibling pairs in microisolator cages (29.2 L x 19.1 W x 12.7 H) containing corn cob bedding and nesting enrichment packs of crinkle cut paper material with food (Tekland global rabbit diet 2030) and water access ad libitum. Colony rooms were maintained at 21°C and on a 14L:10D photoperiod, with lights on at 0600 h. All female subjects were between 60 and 130 days of age at the start of each experiment. Male partners were vasectomized (see Supplemental Methods for surgical procedures). All procedures were conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and the Institutional Animal Care and Use Committee at the University of Kansas.

2.2. Experimental Design

Figure 1 provides an overview of the timeline for each experiment.

Figure 1.

Experimental timelines of each experiment. (OPT: Odor Preference Testing; SPT: Sucrose Preference Testing & Training).

2.2.1. Experiment 1: An exploration of motivational and biological shifts following 1 week of social loss in female prairie voles.

Female subjects were placed into one of the following four experimental groups for the duration of the experiment: (1) female subject continuously housed with a non-sibling female conspecific (same-sex intact, SSI, n = 9), (2) female subject separated from a non-sibling female conspecific following one week of cohabitation (same-sex loss, SSL, n = 7), (3) female subject continuously housed with a male partner (opposite-sex intact, OSI, n = 10), (4) female subject separated from male partner following one week of cohabitation (opposite-sex loss, OSL, n = 8). Subjects underwent odor preference testing (OPT; partner vs food odor) on day 15 and sucrose preference training and testing on days 15–17, as detailed below. On day 17, subjects were sacrificed via rapid decapitation, brains and trunk blood were collected and frozen at −80°C for later immunological processing.

2.2.2. Experiment 2: The motivation for stranger-associated (i.e., non-partner social) cues following 1 week of social loss.

Female subjects were separated into the experimental conditions described in experiment 1 (SSI: n = 7, SSL: n = 6, OSI: n = 7, OSL: n = 6). Subjects underwent OPT (stranger vs food odor) on day 15. The stranger odor used was that of a cage within the same experimental condition as the subject (i.e., same-sex female subjects received bedding from a cage of same-sex females).

2.2.3. Experiment 3: The role of the mPOA DRD1 signaling in partner-seeking behavior following 1 week of partner loss.

One week prior to pairing, female subjects underwent stereotaxic surgery to bilaterally implant cannulae targeting the mPOA for the site-specific infusion of drug and vehicle treatments, further detailed below. All female subjects followed the experimental conditions described above for the OSL group. On the day of OPT, females were bilaterally infused with 200 nL of either artificial cerebrospinal fluid (aCSF; 150 mM Na+, 3.0 mM K+, 1.4 mM Ca2+, 0.8 mM Mg2+, 1.0 mM P3−, 155 mM Cl−) as controls (n = 9), 4 ng SCH 23390 (low dose, n = 10), or 40 ng SCH 23390 (high dose, n = 8) at an infusion rate of 200 nL/min. Females were placed into the behavioral arena for a 20 min habituation period prior to the OPT (partner vs food odor).

2.2.4. Experiment 4: The role of the mPOA DRD1 signaling in stranger-seeking behavior following 1 week of partner loss.

As in experiment 3, female subjects underwent OSL experimental conditions and surgical procedures. On the day of the OPT (stranger vs food odor), subjects received infusions of either aCSF (n = 12) or 40 ng SCH 23390 (n = 10), as no effect was seen in the previous experiment between the aCSF- and 4 ng-administered groups.

2.3. Pharmacological Manipulation

2.3.1. Stereotaxic Surgery

Female voles in experiments 3 and 4 underwent stereotaxic surgery for the bilateral implantation of cannulae to target the mPOA one week before cohabitation. Surgical preparation and analgesic/anesthetic protocols were followed as previously described (43). Guide cannula (RWD, 26-gauge, 5 mm in length) were bilaterally implanted to target the mPOA (15° angle; A-P 0.43, M-L 1.85, D-V −4.70).

2.3.2. Drug Preparation

In experiments 3 and 4, the DRD1-selective antagonist, SCH 23390 (Millipore, Catalog #505723) was used. The drug was prepared in ddH20 at a stock concentration of 2 mg/mL and frozen at −20 °C in 130 uL aliquots. On the day of testing, drug stock was diluted in aCSF to 4 ng or 40 ng per 200 nL. Drug doses and preparation were based on previous literature in prairie voles (43,44).

2.3.3. Site-specific Drug Administration

Thirty min prior to the OPT, 200 nL of either aCSF, 4 ng SCH 23390, or 40 ng SCH 23390 was bilaterally infused in freely moving animals using 30-gauge needles inserted into the guide cannulae (RWD, cut to 6 mm in length). The rate of infusion was controlled at 200 nL/min using an infusion pump (KDS 220/220P Legacy Syringe Pump). Following infusion, subjects were immediately placed into the OPT testing arena for a 20 min habituation period prior to the start of the OPT. This schedule was based on previous literature in prairie voles (43,44).

2.4. Behavioral Testing

2.4.1. Odor Preference Testing (OPT)

The OPT was developed to assess motivation for social and non-social cues without physical social interaction (15,40,41). The OPT was performed between 0800 h and 1200 h. The testing arena is composed of three chambers (75 L x 20 W x 25 H cm), with the outer chambers containing a petri dish (60 × 15 mm) taped to the center of the floor (Figure 2A). The testing procedures were adapted from previous work in our laboratory (15). In brief, animals were allowed a short period of habituation to the testing arena before being corralled into the center chamber while an experimenter added odor stimuli into the petri dishes. The odor stimuli were lightly soiled bedding (social stimulus), or clean bedding mixed with food particulates (food stimulus). Animals in experiments 1 and 2 were given a 10 min habituation period in an empty arena, and those in experiments 3 and 4 were given a 20 min habituation period to account for local action of the drug infusion. Following habituation, animals were allowed free exploration of the arena for 10 min. Behavioral testing was recorded using Logitech web cameras, and the duration of time in proximity (i.e., 1 body length) to the odorants and active olfactory investigation (i.e., sniffing and digging through the bedding) of the odorants were scored using Solomon Coder by an experimenter blind to experimental conditions.

Figure 2.

The investigation of partner- and stranger-associated cues is increased in females that lose a male partner. (A) Diagram depicting the testing arena, with food-scented bedding in a petri dish on one side and social-scented bedding on the other. (B) Experiment 1 odor investigation times in seconds for the partner- and food-scented cues in the odor preference test (OPT) (n = 7–10 voles/group). (C) Experiment 2 odor investigation times in seconds for the stranger- and food-scented cues (n = 6–7 voles/group). (SSI: same-sex intact, SSL: same-sex loss, OSI: opposite-sex intact, OSL: opposite-sex loss). *p < 0.05 **p < 0.005

2.4.2. Sucrose Preference Training and Testing

To measure anhedonia-like states, a sucrose preference test (SPT) and training protocol was adapted from previous literature in prairie voles (43,45,46). Here, we allowed for a 24 h period of bottle training that began between the hours of 0800 h and 1200 h. Two 50 mL tubes with sipper lids, one containing ddH20 and the other containing 0.25% sucrose in ddH20, were placed into the homecage. The concentration of sucrose was chosen based on a previous optimization experiment (43), which revealed that a 1% concentration, as previously used for prairie voles (45,46), led to over-consumption of the sucrose solution. Solution levels were checked throughout the training period, and water, but not the sucrose solution, was replenished if levels reached below 10 mL. Animals then underwent a 16 h food and water deprivation period, immediately followed by a 2 h, 2-bottle choice test between 0.25% sucrose solution and ddH20 in a clean cage. Sucrose and ddH20 consumptions were calculated using pre- and post-test bottle weights.

2.5. Protein and Hormonal Assays

2.5.1. Western Blots

Brains were cryosectioned at −13 °C at 300 microns. Bilateral tissue punches were taken for the ACC, NAc, IC, PVN, mPOA, and the CA2. Tissue punches were lysed in RIPA buffer and total protein concentrations were determined using a BCA assay (Thermofisher). 12% Tris-Glycine gels (Invitrogen Novex) were loaded with protein extracts containing 15 μg of total protein cocktailed with β-mercaptoethanol dye for visualization. Following electrophoresis and transfer to a nitrocellulose membrane, membranes were stained with 5% Ponceau S and imaged with a BioRad ChemiDoc Imaging System as a measure of total protein content. Ponceau S was then washed away from membranes using ddH20. Membranes were blocked in 5% nonfat dry milk in 1X TTBS (Tris-buffered saline with 0.1% Tween 20) for 2 h at room temperature on an orbital shaker at 30–35 rpm (Corning LSE). Membranes were incubated at 4°C in primary antibodies anti-DRD2 (1:1000 concentration for 48 h; Millipore, Catalog #AB5084P), anti-DRD1 (1:1000 concentration for 48 h; MyBioSource, Catalog #MBS9402869), and anti-Oxtr (1:2000 concentration for 24 h; Santa Cruz, Catalog #SC515809) in 5% nonfat milk in 1X TTBS. Following each primary incubation, membranes were incubated in the appropriate HRP-conjugated secondary antibodies (anti-rabbit, Santa Cruz, Catalog #sc-2357; anti-mouse, Santa Cruz, Catalog #sc-516102) at a 1:10,000 concentration in 5% nonfat milk in 1X TTBS for 1 h. Following each secondary incubation, membranes were stained using Super Signal West Dura kits (Pierce, Catalog #34706), chemiluminescence was imaged with a BioRad ChemiDoc Imaging System, and membranes were quenched in 10% CH3COOH (Acetic Acid) at 37°C for 30 min and washed in 1X TTBS. Successful quenching was verified by imaging the chemiluminescence again. The optical density of all bands was then quantified using GelAnalyzer 19.1 (47).

2.5.2. ELISAs

At the time of sacrifice, trunk blood was collected into EDTA-coated tubes (Greiner Bio-One MiniCollect, Fisher Scientific, Catalog #22-040-200). Plasma was separated via two centrifugation cycles of 20 min and 10 min (2,000xG) at 4°C, transferring the plasma layer into a clean tube after each cycle and samples were then stored at −80°C until processing. Corticosterone concentrations were measured via ELISA using anti-corticosterone primary antibody (Millipore, Catalog #AB1297) in sodium bicarbonate buffer 0.05 M (sodium carbonate and sodium bicarbonate in ddH2O, pH 9.6) at a 1:10,000 dilution and secondary antibody corticosterone-HRP conjugate (Creative Diagnostics, Catalog #DAG2977) in phosphate buffer (sodium phosphate monobasic, anhydrous sodium phosphate dibasic, NaCl, and BSA in ddH2O, pH 7.0) at a 1:10,000 dilution. Absorbance was measured at 450 nm with a BioTek plate reader. Samples were run in duplicates, with inter-plate coefficients of variation (CVs) <10% and inter-assay CV <5%.

PVN oxytocin concentrations were measured using the Enzo Oxytocin ELISA kit (Fisher Scientific, Catalog #NC0556488) following the manufacturer instructions. Absorbance was measured at 450 nm with a BioTek plate reader. Samples were run in duplicates, with inter-plate CVs <10% and inter-assay CV <5%.

2.6. Statistical Analysis

Behavioral results in experiments 1 and 2 were obtained by performing mixed-model analysis of variance tests (ANOVAs) with loss condition (intact vs loss) and companion sex (female vs male) as between-subjects factors and odor type (OPT; social vs food) or solution type (SPT; water vs sucrose) as within-subjects factors. Significant results were followed by Bonferroni post hoc tests analyses. Immunological results in experiments 1 and 2 were obtained by performing two-way ANOVAs or Kruskal-Wallis nonparametric tests to assess the main effects of companion sex, loss condition, and their interaction. Significant results were followed by Scheffe (ANOVA) and Bonferroni (Kruskal-Wallis) post hoc analyses to reveal between-group differences. Behavioral results in experiments 3 and 4 were obtained by performing mixed-model ANOVAs with planned comparisons between drug conditions.

3. Results

3.1. Experiment 1: Partner loss increased partner-seeking behavior in the OPT and DRD2 and Oxtr expression in select brain regions.

Behavioral results of the OPT are summarized in Figure 2B.

In the OPT, the within-subjects factor of odor type was significant (F(1,30) = 10.059, p = 0.003, η² = 0.251), such that all females investigated the food-scented odor more than the partner-scented odor. There was a significant interaction between companion sex and loss condition (F(1,30) = 6.934, p = 0.013, η² = 0.188), such that females in the OSL group spent significantly more time investigating the partner-associated and food-associated cues compared to females that lost a female cagemate or remained intact with a male partner (Figure 2B). There were no significant main effects of loss condition (F(1,30) = 1.160, p = 0.29) or companion sex (F(1,30) = 3.631, p = 0.066).

In the SPT, there was a significant main effect of solution type (F(1, 40) = 87.862, p < 0.0001, η² = 0.687), such that all females displayed a preference for sucrose. Additionally, there was a significant interaction between solution type and companion sex (F(1,40) = 19.23, p < 0.0001, η² = 0.325), such that females paired with a male consumed more sucrose than sexually naïve females. There were no significant differences found between intact and social loss groups or the interaction of loss condition and companion sex (F’s(1,40) < 1.292, p’s > 0.212).

ELISA analyses for oxytocin content in the PVN revealed no significant differences between groups (F(1,37) = 1.119, p = 0.354). There were also no significant differences between groups in blood plasma corticosterone content (F(1,34) = 0.505, p = 0.681). Western blotting for DRD1, DRD2 and Oxtr expression revealed no main effects of companion sex and loss condition for any brain region (F’s(1,37) < 1.139, p’s > 0.293). There were significant interactions between companion sex and loss condition for DRD1 in the mPOA (F(1, 37) = 5.39, p = 0.026, η² = 0.437), DRD1 in the CA2 (F(1, 22) = 4.92, p = 0.037, η² = 0.671), and Oxtr in the CA2 (F(1, 22) = 4.37, p = 0.048, η² = 0.596). Specifically, DRD1 protein content was significantly increased within the mPOA in OSL females compared to all other groups (Figure 3G). Additionally, DRD1 protein content in the CA2 was significantly higher in SSL females compared to SSI females (Figure 3C), and Oxtr protein content within the CA2 was significantly elevated in SSL females as compared to SSI females (Figure 3D). No significant differences of DRD2 protein content within any region analyzed were found between groups (F’s(1,37) < 0.5, p’s > 0.8).

Figure 3.

DRD1 and Oxtr protein content is increased in select brain regions in females that experience partner or social loss. (A) Representative brain section depicting tissue punch sites for the dorsolateral CA2 region of the hippocampus (CA2). (B) Representative blots for CA2 DRD1 (top) and CA2 Oxtr (bottom) with Ponceau S stain for total protein content. (C) Normalized optical density of DRD1 protein content within the CA2 (n = 5–8 samples/group). *p = 0.037. (D) Normalized optical density of oxytocin receptor (Oxtr) protein content within the CA2 (n = 5–8 samples/group). *p = 0.048. (E) Representative brain section depicting tissue punch sites for the medial preoptic area (mPOA). (F) Representative blots for mPOA DRD1 with Ponceau S stain for total protein content. (G) Normalized optical density of dopamine receptor 1 (DRD1) protein content within the mPOA with representative blots (n = 8–12 samples/group). *p = 0.026. (SSI: same-sex intact, SSL: same-sex loss, OSI: opposite-sex intact, OSL: opposite-sex loss).

3.2. Experiment 2: Partner loss does not increase stranger-seeking behavior in the OPT.

Behavioral results are summarized in Figure 2C. There was a three-way interaction between companion sex, loss condition, and odor type (F(1,24) = 11.140, p = 0.003, η² = 0.317), such that SSL and OSI females investigated the food-associated cue more than the stranger-associated cue but SSI and OSL females showed no preference. Further, there were no differences in the investigation of stranger or food odors for females across the four social conditions (F’s(1,24) < 0.679, p’s > 0.418).

3.3. Experiment 3: Infusion of a DRD1-selective antagonist into the mPOA decreased partner-seeking behavior in the OPT.

In the OPT, there was a significant interaction between odor type and drug condition (F(2, 24) = 6.751, p = 0.005, η² = 0.563), such that the administration of SCH 23390 at a high dose (40 ng), but not a low dose (4 ng), resulted in significantly lower investigation times of the partner-associated cue compared to aCSF-administered controls (Figure 4B). There were no significant main effects of odor type (F(1,24) = 0.328, p = 0.572) or drug condition (F(2,24) = 0.023, p = 0.977).

Figure 4.

Site-specific mPOA DRD1 antagonist administration decreases partner-associated cue investigation. (A) Brain sections depicting injection sites within the medial preoptic area (mPOA). (B) Olfactory investigation times in the odor preference test (OPT) of experiment 3, in which partner-scented and food-scented bedding were used (n = 8–10 voles/group). *p = 0.005. (C) Olfactory investigation times in the OPT of experiment 4, in which stranger-scented and food-scented bedding were used (n = 10–12 voles/group).

3.4. Experiment 4: Infusion of a DRD1-selective antagonist into the mPOA did not decrease stranger-seeking behavior in the OPT.

There was a significant within-subjects effect of odor type (F(1,20) = 7.608, p = 0.012, η² = 0.38), such that all females investigated the food-associated cue more than the stranger-associated cue. However, females in the aCSF-administered group did not differ when compared to the 40 ng drug treatment group in stranger or food odor investigation (Figure 4C; F’s(1,20) < 1.026, p > 0.323).

4. Discussion

Yearning for reunion and longing for the deceased are distinctive features of the human grieving process (9,10). Recently, we developed the novel OPT behavioral test in the socially monogamous prairie vole to model this aspect of social loss (15). Previously, we reported that male prairie voles experiencing the loss of a female partner display increased motivation towards partner-associated cues, which coincided with shifts in DA system activity within brain regions linked to motivation and emotional pain (15). Here, we demonstrate that females separated from a male partner, but not a female cagemate, display increased investigation of partner-associated cues as well. In addition, the sucrose preference test yielded nonsignificant results across social loss conditions, a finding that is different from previous literature in female voles, which report significant decreases in sucrose preference and consumption following prolonged social isolation from a same-sex cagemate or sibling (45,46,48). These contradictory results may be due to differences in the duration of isolation, as subjects in previous studies were isolated for much longer durations of 4 to 6 weeks, as compared to 1 week in the present study. This may be indicative of a deteriorating hedonic motivational state in response to isolation as time goes on. When isolated from a female cagemate, female voles also demonstrate increased stress-related behaviors (48–51) and physiology (38,39,45,46,52). While the increased stress state is not always observed during social isolation (36,52,53), it does typically require a much longer period of isolation than used in the current study, i.e., 4–6 weeks. As partner loss inevitably induces a form of isolation, the covariate of isolation adds difficulty to the interpretation of partner loss studies in females. The inclusion of sexually naive females that either remain or lose a female cagemate allows the effects of social isolation to more readily be compared to effects directly related to partner loss. That is, the comparison of females that lose a female cagemate and females that lose a male partner allows for a clearer distinction of the responses to isolation and general social loss and the responses to the loss of a pair-bonded partner. This distinction is crucial in understanding the significance of different social relationships within prairie vole sociality. Upon the investigation of peer relationships, females appear to have an increased capacity to form “peer preferences,” or preferences for same-sex cagemates over unfamiliar same-sex conspecifics (21,54). When given a lever-pressing task to access conspecifics, females work hardest to access pair-bonded mates and familiar same-sex conspecifics. Male prairie voles, in contrast, do not work harder for access based on familiarity. Rather, males work harder for access to an opposite-sex conspecific, regardless of their level of familiarity (55). This suggests that the motivation to gain access to distinct types of social interactions is dependent on biological sex. While this is critical to a deeper understanding of social motivation, this and many classic behavioral paradigms inevitably require reunion between subjects and conspecifics, making the study of social motivation during partner loss and social isolation exceedingly difficult. Notably, the OPT is one of the first behavioral paradigms to not require physical reunion. Thus, further studies and methods allowing for this distinction are necessary.

Previous studies demonstrate conflicting results when reporting plasma corticosterone concentration changes as an effect of social isolation and loss. Some have reported increases in circulating corticosterone in response to long term social isolation (45,46) and partner loss (32,37,38,49). Conversely, others report no significant differences in corticosterone concentrations as an effect of isolation (36,53). Here, we found no differences in plasma corticosterone concentration between any experimental conditions. As in the differences we report in sucrose preference and consumption results, these previous studies compared longer periods of isolation (4 and 6 weeks) than the shorter term 1 week isolation and social loss seen here. This may be indicative of a time-sensitive response to isolation and partner loss within the corticosterone system of prairie voles. Encouragingly, in another study, female and male voles displayed increased corticosterone compared to intact controls when isolated from a same sex sibling for 1 hour and for 1 hour daily for 4 weeks. However, in the same experiments, animals isolated continuously for 4 weeks show no differences in corticosterone concentration from their control counterparts (52). These results suggest that the biological stress systems of prairie voles may be much more complex, and thus, more study is needed to unravel the effects of loss and isolation. The sole use of plasma corticosterone as a measure of the responses of this biological stress system may not be enough to understand the impacts of changes in social environments. Many studies have included reports of the corticotropin-releasing hormone and corticotropin-releasing factor system in select brain regions of male prairie voles following partner loss and social isolation (32,33,37,39,46,49,52). The study of these systems may be more telling of biological changes that occur in isolation and loss.

We also demonstrated that females that lost a female conspecific displayed significant increases in Oxtr and DRD1 protein content in the CA2 when compared to females that remain housed with a female conspecific. These results bring into question how these neurochemical systems may be influenced by social environment. The intertwining relationship of the Oxt and DA systems has been reviewed and studied as it relates to pair bonding behaviors in prairie voles (for review, see (56)). It has been shown that the concurrent DRD2 and Oxtr signaling in the NAc is required for pair bond formation in females (57). Individually, both DA and Oxt (or the analogous neuropeptide vasopressin) signaling have been shown to regulate male and female pair bonding (25–27,58). DRD1 and DRD2 seemingly work in tandem during the bonding process in male prairie voles; with DRD2 signaling playing a critical role in male and female partner preference formation, and DRD1 signaling in the maintenance of an established pair bond (25). Conversely, DA signaling is not necessary for the formation of peer preferences over novel conspecifics in either sex (59). This is interesting, as the differences seen in DRD1 and Oxtr within the CA2 in the present study were reflected in groups with differing peer-related social environments. In other words, social isolation, demonstrated here by our same-sex loss groups, elicited increased DRD1 and Oxtr protein content in the CA2 when compared to females whose female cagemate remain intact. Inherently, social attachments, such as pair bonding and peer relationships, require various social memory processes. The CA2 has been shown to play a role in social memory and social recognition in mice and rats (60–64). Oxt signaling in the CA2 has long been studied for its role in social memory in mice and rats (for review, see (65). Less is known about DA within the CA2, especially in consideration of social memory processes. Some report that systemic, direct NAc, and direct frontal cortex manipulations of the DA system in mice and rats facilitate social recognition and learning (66–68). The prairie vole CA2 has only recently been explicitly characterized (69), but it has been shown that social recognition behavior in males is influenced by pair bond status. Namely, bonded males exhibit social recognition of females, but single males do not (70). Furthermore, single males demonstrate social recognition for other males (71). In combination, these results contribute to the selective nature of pair bonds and their maintenance in male prairie voles. Studies that further characterize the involvement of the CA2 in the unique social structure of the prairie vole are needed, especially as the study of partner loss and social isolation gain interest.

Social-seeking behaviors in the OPT and DRD1 protein content in the mPOA were significantly increased as an effect of partner loss. Thus, we examined the possible causal role of DRD1 signaling in the mPOA in increases of social-seeking behaviors. Indeed, investigation times of partner-associated cues were significantly decreased when a DRD1 antagonist was site-specifically infused into the mPOA of females that lost a male partner compared to CSF-administered controls. Notably, the administration of this DRD1 antagonist in the mPOA did not affect the investigation of stranger-associated cues. This suggests that DRD1 signaling in the mPOA may specifically modulate partner-seeking behavior. In other rodents, the mPOA plays an influential role in maternal behavior (72,73). Two studies have also examined the effects of site-specific infusion of the DRD1 antagonist used in the present study, finding that its infusion into the NAc shell, but not the mPOA, disrupted aspects of maternal care in rats (74,75). Thus, DRD1 signaling in the mPOA of rats may not be directly related to maternal behavior. The mPOA has also been investigated for its role in social and nonsocial reward in mice and rats. Pharmacological and optogenetic manipulation of dopaminergic receptors in the mPOA influence male and female mating behaviors (76–79). In mice, the activation and silencing of medial amygdala neurons projecting to the mPOA promotes and suppresses reinforcement behaviors in a social conditioned place preference test, respectively. These projections do not respond to food stimuli and their ablation does not change the behavioral response to food stimuli (80). In female mice, neurotensin-expressing mPOA neurons that project to the ventral tegmental area are preferentially excited by the presentation of a male odor as compared to a female odor and a nonsocial, food-related odor (81). These findings are similar to ours that DRD1 antagonism in the mPOA does not change seeking behaviors toward food-related cues, but these species do not exhibit the characteristic social structure of social monogamy that is seen in prairie voles. Thus, the role of the mPOA in the prairie vole brain may expand to influence behaviors and molecular processes vital to pair bonding. Few studies have investigated the mPOA in prairie voles. One study found that pair-bonded males exhibited increased co-expression of c-fos, a marker for neuronal activity, and tyrosine hydroxylase, the rate limiting factor in DA synthesis, in the mPOA following an interaction with an unfamiliar male as compared to males that did not interact with a conspecific (82). It has also been shown that female prairie voles that mated several times over a period of 6 hours show increased c-fos expression in the mPOA when compared to sexually naive females, females induced into estrous via estradiol benzoate injections then exposed to familiar and unfamiliar females, and females induced into estrous and exposed to a male but not allowed to mate (83). This suggests the role of mPOA, and in extension the role of DA signaling in the mPOA, could be conserved while also expanding to accommodate behaviors specific to pair bonding.

4.1. Conclusion

In summary, the present study is the first to delineate the effects of partner loss and social isolation on behaviors associated with seeking behaviors toward social and nonsocial cues and their underlining neurocircuitry in female prairie voles. Here, we show that DRD1 signaling within the mPOA drives increases in the investigation of partner-associated cues following the loss of a pair-bonded partner in female prairie voles. Previous experiments from our laboratory (15) show that male prairie voles that lost a female partner display significantly increased in partner odor investigation compared to males that either remained with or lost a male cagemate and males that remained with a female partner, and there were no significant differences in investigation of stranger-scented bedding or food-scented bedding, suggesting this increase is specific only to partner-related cues in both sexes. Additionally, males show increased DRD1 and DRD2 mRNA expression in both the ACC and IC (15), regions that demonstrated no such differences in the present investigation of the female prairie vole brain. Thus, the neurochemical outcomes of partner loss and social isolation are sexually dimorphic. Consequently, more research is needed to further elucidate the neurocircuitry related to motivational shifts following partner loss and social isolation.