Sex difference in the effects of ventral pallidum vasopressin 1a receptor partial knockdown on social behavior in mice

Abstract

The neuropeptide vasopressin (AVP) regulates a diverse array of social behaviors, often having different functions in males and females. This sex difference is due, in part, to the AVP cells in the bed nucleus of the stria terminalis (BNST), which are more numerous in males than in females. These AVP cells send stronger projections to several brain regions that express the vasopressin 1a receptor (V1aR), including the ventral pallidum (vPal), an area broadly implicated in reward-seeking behavior. Previous experiments manipulating V1aR in vPal have found that activation of V1aR in this area mediates AVP effects on social behavior with differential effects on male and female rats across different social contexts. Consequently, to better understand the role of V1aR in vPal, we reduced V1aR expression in the vPal using a viral-mediated RNA-interference approach in male and female mice and tested their social investigatory, aggressive, copulatory and communicative responses to male and female conspecifics as well as their responses to anxiogenic or rewarding stimuli. Partial knockdown of V1aR in vPal of males reduced their social investigation of other males, but not of females, whereas the same manipulation had no effect on social investigation in females. In addition, partial knockdown in males reduced latencies to ejaculate during copulation. Reduction in V1aR within vPal did not influence communicative and aggressive behaviors, urine investigation, anxiety-like behavior, or sucrose preference. These results suggest that V1aR activity in the vPal specifically facilitates intermale investigation in adult mice but normally dampens their ejaculatory behavior.

Introduction

The neuropeptide arginine-vasopressin (AVP) has been strongly implicated in the regulation of social behavior and communication across species (Caldwell et al., 2008; Goodson and Bass, 2001; Insel, 2010; Kelly and Goodson, 2014; Rigney et al., 2022a), including humans (Oztan et al., 2020; Parker et al., 2019), and often in sex-specific ways. Indeed, one of the largest and most evolutionarily conserved sex different systems in the vertebrate brain is the male-biased and steroid-dependent AVP innervation of limbic structures (de Vries, 2008; De Vries and Panzica, 2006; Goodson and Bass, 2001), which originates from neurons in the bed nucleus of the stria terminalis (BNST) and medial amygdala (MA) (De Vries and Buijs, 1983; Rigney et al., 2023a; Rood et al., 2013). Removing BNST AVP cells or knocking down AVP within BNST reduces male social behavior, indicating that this system is a powerful regulator of male social behavior (Kelly and Goodson, 2014; Rigney et al., 2023b). BNST AVP cells send denser projections to several brain regions in males than in females, including the ventral pallidum (vPal) (DiBenedictis et al., 2020; Rigney et al., 2023a), a key node in the mesolimbic reward system (Morais-Silva and Lobo, 2024; Root et al., 2015; Soares-Cunha and Heinsbroek, 2023) implicated in reward seeking (Ottenheimer et al., 2018), hedonic impact (Smith and Berridge, 2005), emotional regulation (Knowland et al., 2017), and social investigation (Wang et al., 2020).

The vPal also expresses high levels of the vasopressin 1a receptor (V1aR) (Dumais and Veenema, 2016; Freeman et al., 2020), the dominant AVP receptor in the nervous system (Koshimizu et al., 2012), and one that mediates AVP effects on social recognition (Bielsky et al., 2005; Landgraf et al., 1995), social interest (Rigney et al., 2024), play behavior (Bredewold et al., 2014; Veenema et al., 2013), aggression and territoriality (Albers, 2012; Gobrogge et al., 2009), parental care (Bayerl and Bosch, 2019; Wang et al., 1994), and anxiety-like behavior (Landgraf et al., 1995; Liebsch et al., 1996). Pharmacological blockade of V1aR or viral-mediated V1aR partial knockdown in the vPal impairs formation of pair-bond mate-preferences in male prairie voles (Barrett et al., 2013; Lim and Young, 2004), whereas upregulation of vPal V1aR facilitates it (Lim et al., 2004; Pitkow et al., 2001). In adult rats, blocking V1aR in the vPal reduces male attraction to conspecific females but, surprisingly, increases female interest in male conspecifics (DiBenedictis et al., 2020). In juvenile rats, V1aR activation in the vPal decreases play fighting in males while increasing it in females (Lee et al., 2024). Taken together, these results suggest that AVP/V1aR interactions in the vPal result in sex differences in the regulation of social behavior.

To test this hypothesis, we examined whether reducing V1aR expression in the vPal, using viral expression of a short hairpin RNA (shRNA) targeted against V1aR mRNA, would impact social investigation, adult ultrasonic vocalizations, and urine marking in male and female mice, all aspects of mouse communication that show greater levels of expression in males than in females (Kas et al., 2014; Lehmann et al., 2013; Wöhr, 2014). In addition, given the role of the vPal in motivation (Reichard et al., 2019; Soares-Cunha and Heinsbroek, 2023), we tested whether vPal V1aR knockdown would more generally alter social (copulatory and agonistic behavior) as well as non-social (anxiety-like behavior, hedonic processing) behavior.

Methods

Animals

Adult male and female subjects (C57BL6J, Jackson Laboratories; #000664) and stimulus (CD1, Charles River) mice were maintained at 22 °C on a 12/12 h reverse-light cycle with food and water available ad libitum. Starting one week prior to experiment onset, subjects were single-housed in individually ventilated cages (Allentown NexGen) containing phytoestrogen-free bedding (Alpha-dri) and feed (Teklad). All animal procedures were performed in accordance with the Georgia State University Animal Care and Use Committee regulations and the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Before testing, subjects were given 2–3 opportunities to interact with novel same- and opposite-sex stimulus animals (see below) so as to promote social and communication behaviors (Lumley et al., 1999; Roullet et al., 2011). Subjects first interacted with sexually-active, opposite-sex stimulus animals and then, on a separate day, interacted with a same-sex stimulus mouse. This sequence was repeated once. A third opposite-sex interaction was provided if a subject had not displayed sexual interest during the prior two opposite-sex interactions. Subject males that failed to mount and subject females that failed to receive mounts were excluded from further testing. Male subjects that failed to display aggression during same-sex encounters were also excluded from further testing. During the first sexual interaction, both male and female subjects were paired overnight with opposite-sex stimulus animals. Male subjects interacted with ovariectomized females that were hormonally induced into estrus. Estrous female subjects, as determined by prior vaginal lavage, interacted with gonadectomized and testosterone-implanted stimulus males. At least one day later, male subjects were paired with gonadectomized, group-housed stimulus males for 10 min, while female subjects were placed with a group-housed, non-estrous stimulus female for 10 min. At least 4 days later, subjects were allowed to interact with sexually-active, opposite-sex stimulus animals (as above) for 90 minutes. At least one day later, all subjects were provided with another opportunity to interact with a same-sex conspecific for 10 min.

Stimulus animals: Adult male and female CD1 mice (Charles Rivers Laboratories) were used as stimuli for generating social experience in subjects as well as for social behavior tests, because exposure to different strain mice increases social communication (Arakawa et al., 2007; Nevison et al., 2000; Nyby et al., 1983). Male stimulus animals serving as intruders in the resident-intruder test were sexually naive mice that were gonadectomized and group-housed to reduce the initiation of aggression by stimulus animals (Beeman, 1947). Stimulus males used in the three-chamber social tests were gonadectomized and given a testosterone implant. To enhance behaviors resembling those of competitor/dominant males, these stimulus males were single-housed, given two copulatory interactions with estrous females, as well as two opportunities to aggress against group-housed males. Female stimulus animals used in 3-chamber, copulatory behavior, and same-sex interactions were ovariectomized and group-housed following surgery. After surgical recovery, they were given two sexual encounters with sexually-experienced males as well as two interactions with non-estrous females. In all cases, stimulus females were brought into estrus via subcutaneous injection of estradiol followed 48 hours later by a subcutaneous injection of progesterone 3–4 hours prior to testing (see below).

Viral Vectors

We used an adeno-associated virus (AAV) expressing a short hairpin RNA (shRNA) to target the mRNA sequence for V1aR (AVPR1; AAV8-GFP-U6-m-AVPR1A-shRNA, titer: 5.2 × 10¹² GC/mL, lot: 190826#49, sense target sequence: 5’-CCGGGTCGCCTTCTTCCAAGTGTTA-CTCGAG-TAACACTTGGAAGAAGGCGAC-TTTTTG-3’ and a targeting sequence of GTCGCCTTCTTCCAAGTGTTA; Vector Biolabs, Malvern, PA) and an AAV-expressing a scrambled shRNA (AAV8-GFP-U6-scramble-shRNA; titer: 4 × 10¹³ GC/ml; lot: 19044-190820; Vector Biolabs, Malvern, PA), which does not target any known sequence as a control. The AVPR1 shRNA target sequence demonstrated ~90% knockdown in vitro as tested by the manufacturer. BLAST searches did not reveal significant target alignment between the AVPR1 shRNA sequence and other coding mRNAs, including oxytocin receptor, indicating that the shRNA targeted AVPR1 mRNA specifically. Both scrambled (control) and AVPR1 shRNA vectors also expressed green fluorescent protein (GFP) to allow for visualization of the infected neurons.

Surgical procedures

All surgeries were conducted using 1.5–3% isoflurane gas anesthesia in 100% oxygen; 3 mg/kg of carprofen was given prior to surgery to reduce post-operative pain.

Stereotaxic surgery

Mice were positioned in a stereotaxic frame (David Kopf Instruments) with ear and incisor bars that held bregma and lambda level. After a midline scalp incision, a hand operated drill was used to make holes in the skull exposing the dura. For all subjects, 500 nl of V1aR-shRNA or scrambled control was delivered bilaterally to the ventral pallidum (VP) at a rate of 100 nl/min using a 5-μl Hamilton syringe with a 30-gauge beveled needle mounted on a stereotaxic injector. Injections were angled (males: 16° degrees; females: 18°) and 0.62 mm anterior to bregma; 0.1 mm lateral to midline, and 4.5 mm from pial surface to avoid the overlying lateral septum, an area of high V1aR expression (Urban, 1998). Following virus delivery, the syringe was left in place for 10 min and then slowly withdrawn from the brain.

Gonadectomy and hormone treatment

Testes were cauterized and removed at the ductus deferens via a midline abdominal incision. Silastic capsules (1.5 cm active length; 1.02 mm inner diameter, 2.16 mm outer diameter; Dow Corning Corporation) were filled with crystalline T (Sigma) and inserted subcutaneously between the scapulae after gonadectomy (GDX); this procedure leads to physiologic levels of T (Barkley and Goldman, 1977; Matochik et al., 1994). The ovaries of stimulus female mice were removed via an abdominal incision by cauterization at the uterine horn and attendant blood vessels. To induce sexual receptivity, ovariectomized stimulus females were injected subcutaneously with 0.1 ml of estradiol benzoate (500 μg dissolved in sesame oil, Sigma) approximately 48 hours prior to an injection of 0.1 ml of progesterone (250 μg dissolved in sesame oil). Progesterone was administered 3–4 hours prior to behavioral testing.

Experimental procedures

Behavioral testing began 3–4 weeks following viral injections and occurred within the first 6 h of the dark cycle under red light illumination, except for the elevated plus maze (EPM) which was done under white light. On test days, subjects were adapted to the experimental room for 15 min prior to testing. All tests were scored by an experimenter blind to the viral condition of the subject and testing occurred across multiple cohorts of subjects. Three weeks following stereotaxic viral injections, subjects were first tested on an EPM to examine anxiety-related behavior (Lister, 1987). Subjects were then habituated to the three-chamber apparatus (see below) for two consecutive days, which was then followed by one day of baseline testing in an empty apparatus. After a one-day break, subjects were tested over two consecutive days with a social stimulus (urine, then caged conspecific; see below) from either a male or female stimulus animal (counterbalanced) and, after 5 days, tested for two more consecutive days with the other social stimuli. On the following day, each subject was placed with an opposite-sex stimulus animal to measure copulatory behavior and then, on the next day, placed with a same-sex conspecific to measure agonistic/competitive behavior. Finally, subjects were tested for sucrose preference over a three-day period to identify changes in their hedonic tone (Primo et al., 2023). A subset of subjects was tested for their ability to discriminate between social and non-social odorants on the day after the sucrose preference test. Female subjects were tested irrespective of estrous cycle day, except during copulation testing, when they were in behavioral estrus. Prior research indicates minimal effects of estrous cycle on female mouse communicative behavior (Coquelin, 1992; Maggio and Whitney, 1985; Moncho-Bogani et al., 2004). Following behavioral testing, subjects were sacrificed, and their brain tissue was processed for autoradiographic detection of V1aR binding in VP, diagonal band of Broca (DBB), and lateral septum (LS). All subjects were given at least a 72-hour gap between their final behavioral test and tissue collection to prevent any potential effects of the behavioral tests on binding. See Figure 1 for experimental testing sequence.

Fig. 1.

Timeline.

Elevated plus maze (EPM)

The EPM consisted of two open arms (30 cm (L), 5 cm (W)) and two closed arms (30 cm (L), 25 cm (H), 5 cm (W)) and two closed arms crossed perpendicular to each other and raised 60 cm above the floor. Subjects were placed at the arm intersection facing the open arm and were then allowed to habituate to the apparatus for 1 min; subjects were then observed for an additional 5 min. Time spent in the closed and open arms was scored manually (Lister, 1987).

Social investigation and communication

Ultrasonic vocalizations (USVs), urine marks (UMs), and investigation were recorded within an acrylic three-chamber apparatus (Harvard Apparatus; dimensions: (62 (L) x 40 (W) x 22 (H) cm; (Arakawa et al., 2008; Crawley, 2007; Moy et al., 2004). Instead of a solid floor, the apparatus was placed on absorbent paper (Nalgene Versi-dry, Thermo Fisher Scientific) to measure deposited UMs.

During habituation to the three-chamber apparatus, subjects were placed in the middle chamber and given 2 min to explore the apparatus which contained two empty cages placed at opposite corners of the outer chambers of the apparatus. On the third day, subjects were given a baseline test where USVs, UMs, and interaction with empty cages was recorded for 5 min. During each 5 min test with stimulus animals, subjects had access to both a caged, novel stimulus animal and an empty cage placed at opposite corners of the outermost chambers of the apparatus. For testing with social odors, subjects had access to 50 μl of fresh urine from a stimulus animal or 50 μl saline pipetted onto a clean piece of filter paper (3 cm²), that was taped on the outside of cages. Stimuli were presented in or on cylindrical cages (8 cm (D), 18 cm (H); 3 mm diameter steel bars, 7.4 mm spacing). The location of stimulus and the empty cages were counterbalanced across animals. We measured investigation of both empty and stimulus cages, as well as USVs and UMs. In addition, distance traveled and average velocity for a subset of subjects in each stimulus condition was measured off-line using AnyMaze software (Stoelting).

After each test, the apparatus and cages were thoroughly cleaned with 70% ethanol and allowed to dry before further testing. In all cases, urine stimulus from one sex was presented first, followed by a live stimulus of that same sex; this order was then repeated for the opposite sex (counterbalanced). In this fashion, mice first experienced a weak stimulus (urine) then a stronger social stimulus (caged conspecific).

Investigation and ultrasonic vocalizations

Investigation was defined as time a subject’s nose was oriented toward and within 2 cm of the stimulus or empty cage; climbing on the cage was not scored as investigation. Time spent in investigation of each stimulus was assessed live using a stopwatch (FisherBrand). USVs were detected using a condenser microphone connected to an amplifier (UltraSoundGate CM16/CMPA, 10–200 kHz, frequency range) placed 4 cm inside the apparatus and directly above the center compartment. USVs were sampled at 200 kHz (16-bit) with target frequency set to 70 kHz (UltraSoundGate 116Hb, Avisoft Bioacoustics). Recordings were then analyzed using MUPET, a MATLAB (MathWorks, RRID:SCR_001622) package that automates aspects of USV analysis (Van Segbroeck et al., 2017). Using this program, sonograms were generated by calculating the power spectrum on Hamming windowed data and then transformed into compact acoustic feature representations (Gammatone Filterbank). Each 200 ms window containing the maximum USV syllable duration was then clustered, via machine learning algorithms, into USV syllable types (repertoire units) based on time-frequency USV shape. Repertoire units that appeared as background noise were discarded. The total number of USVs produced during a behavioral test were analyzed. Due to technical issues with audio recording, USVs were only obtained from a subset of subjects from each condition.

Urine marking

Following testing, the substrate sheet was allowed to dry for 1 h and then sprayed with ninhydrin fixative (LC- NIN-16; Tritech Forensics Inc.) to visualize urine marks (Lehmann et al., 2013). After 24 h, sheets were imaged (Sony DSC-S700 camera), binarized and analyzed using a computer-aided imaging software (ImageJ, RRID:SCR_003070). UMs were measured as the total area (cm²) of visualized ninhydrin urine marks within the entire arena. Urine marks that were larger than 6 cm² and directed toward corners were rare and counted as elimination and were therefore excluded from all urine marking measurements (Bishop and Chevins, 1987).

Urine collection

Pooled urine samples were collected from ovariectomized females that were induced into estrus and from GDX+ T males that were single-housed and sexually-experienced (five to eight mice per sample). Estrous state was verified by vaginal lavage (Caligioni, 2009). To collect urine, mice were picked up by the tail base and held by dorsal neck skin; this method was often sufficient to induce urination. If the mouse did not urinate, stroking its belly from an anterior to posterior direction stimulated bladder voiding. Each mouse provided 15–50 μl of urine that was pooled into a 1.5 ml Eppendorf tube. Urine samples were used fresh within 1 h of collection to prevent breakdown of key chemosignals (Sipos et al., 1992).

Copulatory and aggressive behavior

To measure copulatory behavior, an opposite-sex stimulus mouse was placed in the subject’s home cage and then removed 90 min later. We measured the latency to engage in anogenital investigation as well as mounting, the total number of mounts performed/received, as well as the percentage of females that were mounted and males that ejaculated. To measure territorial aggression, gonadectomized, group-housed stimulus males were placed in the subject’s home cage and then removed after the subject’s first offensive attack (rolling fight) within a 10 min period; we recorded the latency to the first rolling fight. To elicit subject aggression, 50 μl of urine from sexually-experienced, single-housed stimulus males were collected and pipetted onto the stimulus male’s back immediately prior to testing. Female subjects (non-estrous) were placed with another non-estrous female for 10 min, but no aggression was observed. Both copulatory and aggressive/same-sex interactions were manually scored from digital video recordings.

Sucrose Preference

Subjects were acclimated to sucrose preference testing by placing two bottles filled with tap water in their home cage for two days prior to any sucrose intake measurement. These water bottles were then replaced with pre-weighed bottles, one containing sucrose solution (2.5% in tap water) and the other containing tap water. Animals had access to both bottles for the next 20 h. The bottles were then removed and any reduction in liquid was measured. A control water bottle was placed in a nearby empty cage to measure any drip from the water bottle. This drip rate (<1ml/250ml) was subtracted from all values prior to analysis. Preference for sucrose was calculated as the percentage of sucrose consumed compared to total consumption (Sucrose/(Sucrose + Water) * 100%).

Odor discrimination

We used a habituation-discrimination procedure to test for any manipulation-related changes in social or non-social odor discrimination on a subset of subjects (n = 17; 2 male shRNA, 5 male control, 4 female shRNA, 6 female control) (Baum and Keverne, 2002). Subjects were given five consecutive 2-min presentations of an odor stimulus with one min intervals between presentations. Subjects were first presented with deionized water (as a control), followed by two non-social odors (100% almond or 50% lemon extract, Frontier Natural Products), and then two urine samples (one from each sex; see above) within their home cage. The sequence of odor presentation was counterbalanced within non-social (almond or lemon) or social odor category (urine). For each odor stimulus, 30 μl was pipetted onto a clean piece of filter paper (3 cm²) that was taped to an empty food hopper in the subject’s home cage so that subjects could contact the urine samples. Time spent sniffing within 2 cm of the filter paper was recorded. After testing, the food hopper was removed and a new, clean hopper was then prepared and inserted for the next trial. All food hoppers were cleaned with 70% ethanol and allowed to dry before being reused.

Histological analysis

Receptor autoradiography

Subjects were euthanized by CO₂ exposure and brains were immediately extracted and flash-frozen onto dry ice. Tissue was allowed to sit on dry ice for 20 min and was then stored at −80°C. Prior to sectioning, tissue was allowed to adjust to cryostat temperature (Leica CM3050 S, Leica Biosystems) for 30 min and was then coated in a thin layer of OCT compound and cut at 20 μm (coronal) sections. Sections were mounted directly onto slides at room temperature and allowed to dry prior to storage at −80°C.

Frozen slides with brain sections containing the ventral pallidum (VP), diagonal band of Broca (DB), and lateral septum (LS) were allowed to dry at room temperature for 30 min. Once dry, sections were carefully placed into metal slide racks and then fixed with 1% paraformaldehyde in a 0.1M phosphate buffer solution, followed by two 10-minute rinses with a 1X Trizma buffer solution (using Trizma crystals with preset pH). Sections were then incubated for 1 hr in tracer buffer (15% bacitracin and 50% bovine serum albumin in 1X Trizma buffer with MgCl) with iodinated linear vasopressin V1aR antagonist (Phenylacetyl 1, 0-Me-D-Tyr2, [¹²⁵I-Arg6], (Perkin Elmer) to visualize V1aR binding, rinsed with 1X Trizma buffer with MgCl 3 times for 10 min each, and then briefly dipped in cold deionized water and dried with a hair dryer. Dried slides were exposed to 35 cm x 43 cm Kodak film (Carestream BIOMAX 856 7232) for 72 hours. A carbon-14 standard (American Radiolabeled Chemicals) was placed in the center of each film for calibration. Films were developed and then imaged on an Epson Expression 12000XL photo scanner at 2400 dpi (8-bit grayscale).

Receptor binding was quantified and converted to disintegrations per minute (DPM/0.4 mg) by comparing it to the rate of decay of the carbon-14 standard using a NIH-Rodbard corrected normal curve generated by ImageJ (NIH: https://imagej.net/nih-image/manual/menus/analyze.html). Image scale was set by measuring a scale box of known height (5 mm) and assigning the correct height in ImageJ. To control for background binding, density measurements were taken from cortical regions estimating nonspecific binding. For each region of interest, the nonspecific binding was subtracted from the specific binding for each measurement across three sequential sections. V1aR density measurements were taken from 3 sequential sections containing the VP, LS, and DB, which were identified by comparing landmarks to relevant anatomical plates (Bregma +0.74, +0.62, and +0.50 mm) from (Paxinos and Franklin, 2003) (Figure 2A).

Fig. 2.

Receptor binding density-revised.

Assessment of viral spread

One series of frozen tissue was allowed to dry at room temperature for 10 minutes and then cover-slipped with Prolong Gold with DAPI. Bilateral images were taken at 10x magnification using a Zeiss Axio Imager.M2 microscope (Carl Zeiss Microimaging), which transferred fluorescent images (FITC contrast reflector) to image analysis software (Stereo Investigator, MicroBrightField, RRID:SCR_002526). The presence/absence of fluorescently-labeled cells was noted across three sections through VP; subjects were excluded from behavioral analysis if viral spread was absent, unilateral, or outside of VP.

Statistical Analysis

All data were analyzed in R (RStudio 2023.03.0+386 “Cherry Blossom”; macOS) and graphs were made using R (R Studio 2024.09.0+375). If data violated the assumption of homogeneity of variance, nonparametric tests were used. Packages used for data analysis were tidyverse, dlookr, ggplot, nlme, broom, car, stringr, emmeans, tibble, readr, rstatix, and effect size. Levels of V1aR expression, sucrose preference, and time in the open arm during EPM were assessed using a 2-way ANOVA, using subject sex and treatment (shRNA, scrambled vectors) as factors. Multiple tests were run for different response variables within the three-chamber apparatus. The response variables from three-chamber tests included: social investigation, distance traveled and average speed, USV production and urine marking. Social investigation was calculated as a social investigation index: the total time spent investigating a conspecific or urine stimulus divided by the total time spent investigating both the stimulus and the empty cage, multiplied by 100. For most response variables, 3-way ANOVAs in male and female stimulus conditions were first used to test the effect of treatment with subject sex and stimulus sex assigned as fixed factors. Separate 2-way ANOVAs were then run for each subject sex, with stimulus sex and treatment assigned as fixed factor effects. To directly compare the effect of sex, we also conducted separate 2-way ANOVAs for opposite- and same-sex stimuli in the three-chamber test, with stimulus sex and treatment assigned as fixed factor effects. Sex differences between subjects were not assessed for urine marking, USVs, or same- and opposite-sex interactions (aggression, sexual behavior) because the baseline expression of each of these behaviors was essentially binary by sex. Post-hoc Tukey HSD tests were performed to test for differences across treatments and sex with a p-value adjustment for multiple comparisons. Pearson’s chi-square tests were used to compare the proportion of subjects from each treatment group who urine marked to conspecific stimuli and urine stimuli. For response variables that significantly differed based on treatment, we also used a Spearman correlation test to compare the degree of behavioral change against V1aR binding density levels. All other behaviors were assessed using either Mann Whitney U or independent samples t-test, depending on the distribution of the data. Effect sizes were estimated using partial eta-squared (ηp²) for all ANOVA, Cliff’s delta (r) for the Mann-Whitney U, and Cramer’s V (V) for all Pearson’s chi-square tests.

Results

Viral-mediated shRNA decreased V1aR binding in vPal in both males and females

Injection of V1aR shRNA led to a significant decrease in V1aR binding in vPal for subjects receiving shRNA relative to those receiving the control vector (2-way ANOVA, main effect of treatment: F₁₍₄₅₎ = 10.358, p = .00239, ηp² = 0.19 (95% CI [0.05, 1.00])); Tukey post hoc, p = 0.0025; Figure 2A,B) with no sex differences in V1aR binding observed (F₁₍₄₅₎ = 2.853, p = 0.099, ηp² = 0.06 (95% CI [0.00, 1.00])). On average, this treatment resulted in a ~68% knockdown in males and a ~50% knockdown in females, relative to the respective control group. This partial knockdown of V1aR was specific to vPal as no significant reduction in V1aR binding in either the overlying lateral septum (F₁₍₄₅₎ = 0.001, p = 0.979, ηp² = 0.04 (95% CI [0.00, 1.00])); Figure 2C) or nearby diagonal band of Broca (F₁₍₄₅₎ = 1.66, p = 0.204, ηp² < 0.01 (95% CI [0.00, 1.00])); Figure 2C).

V1aR partial knockdown in the vPal decreased social investigation in males

Social Investigation

Both male and female subjects showed increased investigation of the caged stimulus animals (whether male or female) relative to the empty stimulus cage. Across all subjects, a 3-way ANOVA revealed a significant effect of stimulus sex (F₁₍₄₄₎ = 15.976, p < 0.001, ηp² = 0.27 (95% CI [0.10, 1.00])) and subject sex (F₁₍₄₄₎ = 37.602, p < 0.001, ηp² = 0.46 (95% CI [0.28, 1.00])), but no effect of treatment (F₁₍₄₄₎ = 0.260, p = 0.613, ηp² < 0.01 (95% CI [0.10, 1.00])). For male subjects exposed to social stimuli, a 2-way ANOVA indicated a significant effect of treatment (F₁₍₂₁₎ = 5.593, p = 0.028, ηp² =0.21 (95% CI [0.01, 1.00])) and stimulus sex (F₁₍₂₁₎ = 17.793, p < 0.001, ηp² = 0.46 (95% CI [0.19, 1.00])) on the investigation index. V1aR partial knockdown in male subjects decreased their social investigation toward male (Tukey post hoc, p < 0.012, Figure 3B), but not female (Tukey post hoc, p = 0.367, Figure 3A), stimuli. There was no correlation between social index and levels of V1aR binding density in the ventral pallidum (r_s = 0.171, p = 0.257). In contrast to males, there were no significant effects of treatment (F₁₍₂₃₎ = 0.453, p = 0.508, ηp² = 0.02 (95% CI [0.00, 1.00])) or stimulus sex (F₁₍₂₃₎ = 3.526, p = 0.073, ηp² = 0.13 (95% CI [0.00, 1.00])) on female social investigation (Figure 3). There was also no correlation between female social investigation and levels of V1aR binding density in the ventral pallidum (r_s = −0.151, p = 0.295). A 2-way ANOVA revealed a sex difference in social investigation toward opposite-sex (F₁₍₄₄₎ = 58.401, p < 0.001, η²p =0.57 (95% CI [0.41, 1.00])) and same-sex stimulus animals (F₁₍₄₄₎ = 7.268, p = 0.010, η²p =0.14 (95% CI [0.02, 1.00])) across all subjects. When exposed to opposite-sex stimuli, males exhibited higher social investigation than females in both the control (p<0.001) and shRNA groups (p<0.001, Figure 3A). In contrast, during same-sex trials, males exhibited higher social investigation than females in the control group (p=0.004), but not the shRNA group (p=0.483, Figure 3B). There was no treatment effect on investigation for male or female subjects with either opposite-sex (F₁₍₄₄₎ = 0.002, p=0.970, η²p < 0.001 (95% CI [0.00, 1.00])) or same-sex stimuli (F₁₍₄₄₎ = 0.713, p = 0.403, η²p =0.02 (95% CI [0.00, 1.00])).

Fig. 3.

Social investigation-revised.

Investigation of urine cues (social investigation index) differed by subject sex (F₁₍₄₁₎ = 9.575, p = 0.004, ηp² =0.19 (95% CI [0.04, 1.00])), but there was no effect of stimulus sex (F₁₍₄₁₎ = 0.071, p = 0.791, ηp² < 0.01 (95% CI [0.00, 1.00])) or viral vector treatment (F₁₍₄₁₎ = 0.002, p = 0.963, ηp² < 0.01 (95% CI [0.00, 1.00])); data not shown). There was no effect of viral vector treatment for either male (F₁₍₁₉₎ = 0.164, p = 0.690, ηp² < 0.01 (95% CI [0.00, 1.00])) or female subjects (F₁₍₂₂₎ = 0.104, p = 0.750, ηp² < 0.01 (95% CI [0.00, 1.00])), and there was no effect of stimulus sex for either male (F₁₍₁₉₎ = 0.857, p = 0.366, ηp² =0.04 (95% CI [0.00, 1.00])) or female subjects (F₁₍₂₂₎= 0.038, p = 0.847, ηp² < 0.01 (95% CI [0.00, 1.00])). A 2-way ANOVA revealed a sex difference in social investigation toward opposite-sex (F₁₍₄₄₎ = 58.401, p < 0.001, η²p =0.57 (95% CI [0.41, 1.00])) and same-sex urine stimulus (F₁₍₄₄₎ = 7.268, p = 0.010, η²p =0.14 (95% CI [0.02, 1.00])) across all subjects. There was no effect of viral vector treatment on either sex (F₁₍₄₃₎ = 0.697, p = 0.408, η²p = 0.02 (95% CI [0.00, 1.00])) in response to an opposite-sex urine stimulus, but males tended to exhibit higher social investigation than females (F₁₍₄₃₎ = 5.775, p = 0.021, η²p =0.12 (95% CI [0.01, 1.00]); post hoc analyses revealed that this trend was driven by control groups (Tukey post hoc, p = 0.051) and not shRNA groups (Tukey post hoc, p = 0.171). In response to same-sex urine stimulus, there was no effect of sex (F₁₍₄₃₎ = 2.774, p = 0.103, η²p =0.06 (95% CI [0.00, 1.00])) or treatment (F₁₍₄₃₎ = 0.238, p = 0.628, η²p < 0.01 (95% CI [0.00, 1.00])).

Activity

Activity, indexed by total distance traveled and average speed, was measured in the 3-chamber apparatus during exposure to social stimuli (Figure S1). Partial knockdown of V1aR had no effect on distance traveled (F₁₍₂₈₎ = 2.949, p = 0.097, ηp² = 0.10 (95% CI [0.00, 1.00])), with no effect of subject sex (F₁₍₂₈₎ < 0.001, p = 0.986, ηp² < 0.01 (95% CI [0.00, 1.00])) or stimulus sex (F₁₍₂₈₎ < 0.001, p = 0.990, ηp² < 0.01 (95% CI [0.00, 1.00])). Similarly, V1aR reduction in vPal did not alter the average speed of subjects during exposure to caged stimulus animals (F₁₍₂₈₎ = 0.447, p = 0.509, ηp² = 0.02 (95% CI [0.00, 1.00])) with no effect of subject sex (F₁₍₂₈₎ = 0.468, p = 0.500, ηp² = 0.02 (95% CI [0.00, 1.00])), or stimulus sex (F₁₍₂₈₎ = 0.474, p = 0.497, ηp² = 0.02 (95% CI [0.00, 1.00]); data not shown).

Partial knockdown of V1aR in the vPal did not alter urine marking behavior or USV production

V1aR partial knockdown in the vPal did not impact the proportion of males who urine marked in the presence of male stimuli compared to controls (X²_{(1, 23)} = 2.561, p = 0.110, r = 0.14 (95% CI [0.00, 1.00])), or the proportion that urine marked to female stimuli (X²_{(1, 23)} = 0.524, p = 0.469, r = 0 (95% CI [0.00, 1.00])). There was also no difference between the proportion of knockdown and control male subjects who urine marked in the presence of male urine (X²_{(1, 23)} = 1.495, p = 0.221, r = 0.14 (95% CI [0.00, 1.00])) or female urine (X²_{(1, 23)} = 0.100, p = 0.752, r = 0 (95% CI [0.00, 1.00])); Table S1). Reducing V1aR in vPal in males did not alter either the number of marks (U = 92, p = 0.109, r = 0.34, Figure 4A) or the total area marked (U = 86, p = 0.220, r = 0.26, Figure 4B) in response to another male. Similarly, V1aR partial knockdown in males had no effect on the number (U = 73.5, p = 0.6599, r = 0.10) or area (U = 73.5, p = 0.6598, r = 10) of urine marks in response to females (Figure 4A and B). There was also no effect of knockdown on number (U = 75.5, p = 0.5742, r= 0.12) or area (U = 77, p = 0.5122, r = 0.14) of urine marks in response to male urine or female urine (Number: U = 62, p = 0.827, r = 0.05; Area: U = 68, p = 0.9254, r = 0.02).

Fig. 4.

Urine mark.

Similar to the effects of V1aR knockdown in males, vPal V1aR partial knockdown did not impact the proportion of female subjects who urine marked in response to same-sex (female) conspecifics (X²_{(1, 26)} = 3.391, p = 0.066, r = 0.31 (95% CI [0.00, 1.00])), or the proportion of females marking toward males (X²_{(1, 26)} = 0.619, p = 0.431, r = 0 (95% CI [0.00, 1.00])). The portion of females which urine marked to male urine (X²_(1,26) = 0.170, p = 0.680, r = 0 (95% CI [0.00, 1.00])) and female urine (X²_(1,26) = 0.248, p = 0.619, r = 0 (95% CI [0.00, 1.00])) was not impacted by shRNA treatment (Table S1). V1aR partial knockdown did not affect either the number of marks (U = 62, p = 0.219, r = 0.25) or area marked (U = 61, p = 0.199, r = 0.26) in response to male stimuli. Similarly, V1aR partial knockdown had no effect on the number (U = 104, p = 0.078, r = 0.35) or area (U = 104, p = 0.079, r = 0.35) of urine marks in response to other females (Figure 4C and D). V1aR partial knockdown in female subjects also had no effect on number and area of urine marks in the presence of male urine (Number: U = 90, p = 0.7626, r = 0.07; Area: U = 90, p = 0.7627, r = 0.07) or female urine (Number: U = 92.5, p = 0.5761, r = 0.12; Area: U = 92, p = 0.6019, r = 0.11).

The production of ultrasonic vocalizations was unaltered by V1aR partial knockdown in the vPal in either male subjects (male stimulus: U = 24.5, p=1, r = 0; female stimulus: U = 16, p = 0.318, r = 0.29; Figure 5A) or female subjects (male stimulus: U = 25, p = 0.309, r = 0.26; female stimulus: U = 35, p = 0.189, r = 0.32; Figure 5B).

Fig. 5.

Ultrasonic vocalizations.

Male copulatory behavior, but not aggression, is altered following V1aR partial knockdown in the vPal

V1aR partial knockdown in males decreased their latency to ejaculate (Figure 6A, U = 64, p = 0.0412, r =0.50) without significantly reducing either latency to mount (U = 21.5, p = 0.102, r = 0.40) or latency to intromit (U = 52, p = 0.331, r = 0.24) with a sexually-receptive female. In contrast, V1aR partial knockdown in females did not alter their latency to be mounted (U = 89, p = 0.815, r = 0.05), to receive intromissions (U = 86, p = 0.945, r = 0.02), or to receive ejaculations (U = 71.5, p = 0.166, r = 0.28). The latency to ejaculate (male) or receive ejaculations (female) was not correlated with levels of V1aR binding density in either male (r_s = −0.029, p = 0.867) or female subjects (r_s = - 0.146, p=0.300).

Fig. 6.

Sex and aggression behavior.

Reduction of V1aR in vPal did not influence aggressive behavior in males (female subjects did not aggress) as there were no significant treatment differences in latency to a rolling fight (U = 52, p = 0.621, r= 0.11; Figure 6C) during resident-intruder interactions.

Anxiety-like and hedonic behaviors are unaffected by V1aR partial knockdown in the vPal

Alterations in social behavior following V1aR partial knockdown were unlikely to be due to knockdown effects on anxiety because anxiety-like behavior in the EPM was unaffected by V1aR partial knockdown (F₁₍₄₄₎ = 1.448, p = 0.235, ηp² = 0.03 (95% CI [0.00, 1.00])); Figure S2A). Furthermore, these results are unlikely to be explained by alterations to fluid balance or anhedonia because there was no effect of V1aR partial knockdown on sucrose preference (F₁₍₄₂₎ = 0.041, p = 0.841, ηp² < 0.01 (95% CI [0.00, 1.00])); Figure S2B) or water consumption (F₁₍₄₂₎= 0.774, p = 0.384, ηp² < 0.01 (95% CI [0.00, 1.00])); Figure S2C).

Partial knockdown of V1aR in vPal did not alter odor habituation or discrimination

Investigation of odors by all subjects decreased across trials (habituated) to both nonsocial (first nonsocial odor test: F_(1,13) = 3.45, p = 0.09, ηp² = 0.21 (95% CI [0.00, 1.00]), second nonsocial odor test: F_(1,13) = 10.57, p < 0.01, ηp² = 0.45 (95% CI [0.11, 1.00])) and social odors (first social odor test: F_(1,13) = 36.50, p < 0.01, ηp² = 0.74 (95% CI [0.48, 1.00]), second social odor test: F_(1,13) = 85.48, p < 0.01, ηp² = 0.87 (95% CI [0.72, 1.00]))(Table S2). Subjects also discriminated between social and nonsocial odors (F_(1,13) = 84.06, p < 0.01, ηp² = 0.87 (95% CI [0.72, 1.00]) across all treatment and sex groups (Tukey post hoc, p < 0.001).

There was no effect of partial V1aR knockdown on habituation to nonsocial (first nonsocial odor test: F_(1,13) = 5.58, p = 0.18, ηp² = 0.13 (95% CI [0.00, 1.00]), second nonsocial odor test: F_(1,13) = 0.002, p = 0.96, ηp² < 0.01 (95% CI [0.00, 1.00])) or social odors (first social odor test: F_(1,13) = 0.815, p = 0.38, ηp² = 0.06 (95% CI [0.00, 1.00]), second social odor test: F_(1,13) = 0.29, p = 0.59, ηp² = 0.02 (95% CI [0.00, 1.00]))(Table S2). All subjects, irrespective of viral treatment, discriminated between social and nonsocial odors (F_(1,13) = 0.01, p = 0.93, ηp² < 0.01 (95% CI [0.00, 1.00])).

Subject sex influenced levels of investigation during odor habituation to the first nonsocial odor (F_(1,13) = 5.72, p = 0.03, ηp² = 0.31 (95% CI [0.02, 1.00])) but not the second nonsocial odor (F_(1,13) = 0.01, p = 0.94, ηp² < 0.01 (95% CI [0.00, 1.00])). There was also an effect of subject sex on olfactory investigation during habituation to the first social odor test (F_(1,13) = 6.775, p = 0.02, ηp² = 0.34 (95% CI [0.04, 1.00])), as well as the second social odor test (F_(1,13) = 6.20, p = 0.03, ηp² = 0.32 (95% CI [0.03, 1.00])).

This latter effect was driven by greater odor investigation by control males compared to control females during the initial trial of the second social odor (Tukey post hoc, p < 0.001; Table S2).

Discussion

Substantial reduction of vasopressin 1a receptor (V1aR) in the ventral pallidum (vPal) by viral-mediated RNA interference reduced investigation of other males in male mice, but not their interest in female conspecifics. In females, the same reduction did not alter social investigation. The reduction in social investigation in males does not result from changes in general social interest, social odor processing, territorial aggression, anxiety, non-social hedonic tone, or general activity. The decrease in male-male investigation specifically is identical to deficits observed after impairments of BNST AVP cells (Rigney et al., 2024, 2022b, 2019), the major AVP input to vPal (De Vries and Buijs, 1983; DiBenedictis et al., 2020; Rigney et al., 2023a). Consequently, our results suggest that V1aR-expressing cells in the vPal are a key target of BNST AVP cells for regulating male-male social approach and investigation.

One limitation of our study is that knockdown of V1aR in vPal was incomplete and so may have limited our ability to see additional effects that would be apparent after more substantial reduction of vPal V1aR. Therefore, we cannot conclude that V1aR in vPal is not involved in regulating other behavioral states, such as anxiety. However, the extent of partial knockdown of V1aR was significant, with half to ⅔ reduction in V1aR binding, a level of depletion that is considerably greater than similar manipulations in prairie voles that altered male mate preference (Barrett et al., 2013). Additional effects of V1aR partial knockdown may also be more apparent if they had extended into the nearby diagonal band of Broca, which also expresses high levels of V1aR in mice (Hammock and Levitt, 2012). In addition, although the BNST provides the majority of AVP input to vPal (Rigney et al., 2023a; Rood et al., 2013) AVP inputs from other, likely hypothalamic, sources are present near or in vPal (Rood et al., 2013), limiting our suggestion that BNST AVP inputs to vPal V1aR cells are the key drivers of our effects.

The specific decrease in social investigation following V1aR partial knockdown in vPal of males (a reduction of investigation of males but not females) is identical to reductions in social investigation following deletion (Rigney et al., 2019), optogenetic inhibition (Rigney et al., 2024), or AVP mRNA interference of BNST AVP cell bodies (Rigney et al., 2022b), which are known to provide substantially stronger AVP innervation to vPal in males than in females (De Vries and Buijs, 1983; DiBenedictis et al., 2020; Rigney et al., 2023a). Because AVP is generally thought to alter behavior through its action on V1aR (Raggenbass et al., 1998), the present results suggest that AVP, likely from the BNST, acts on V1aR in vPal to facilitate intermale social interactions.

The present findings differ significantly from the effects of blocking V1aR in the vPal of male and female rats (DiBenedictis et al., 2020). Injections of a V1aR antagonist in vPal reduced male rats’ investigation of females but not of other males, an effect opposite to what we observed in mice. Moreover, V1aR blockade increased female rats’ investigation of male conspecifics, whereas our reduction in vPal V1aR did not alter social investigation in females. One important difference between these studies in rats and mice is that our viral-based manipulation chronically (3+ weeks) depleted V1aR expression whereas the study in rats used acute pharmacological inhibition of V1aR. Consequently, any additional changes in behavior following V1aR partial knockdown may have faded as the vPal adapted to a chronic decrease in V1aR. This idea is supported by opposite behavioral effects of chronic versus acute V1aR manipulations in the lateral septum (Everts and Koolhaas, 1999; Liebsch et al., 1996). Species differences as well as differences in social experience (mice: sexually-experienced; rats: sexually-naive) may have altered behavioral responses following vPal V1aR inhibition in rats and mice. Nevertheless, both our study and DiBenedictis et al., 2020 found that impairment of V1aR in the vPal erased sex-differences in social investigation and so it appears that V1aR expression in vPal may be needed for appropriate and sex-specific social investigation in both species.

We observed a slight, but significant, decrease in latency to ejaculate, without concomitant reductions in latency to mount or intromit, indicating that males ejaculated faster during copulation. While this may indicate increased sexual arousal (Sachs, 2007), the fact that V1aR partial knockdown spared interest in females, did not alter latencies to mount and intromit, or change precopulatory ultrasonic vocalizations (Musolf et al., 2010; Nyby, 1983), militates against this interpretation. Instead, these results suggest specific effects of vPal V1aR in transitions from intromissions to ejaculation. Although manipulations of V1aR in VPal in prairie voles did not observe effects on male copulatory behavior (Barrett et al., 2013; Lim et al., 2004; Pitkow et al., 2001; Young et al., 1999), these experiments were not designed to detect changes in copulatory sequences. Notably, knockdown of AVP within BNST cells (Rigney et al., 2022b), which provide the majority of AVP input to vPal (De Vries and Buijs, 1983; DiBenedictis et al., 2020; Rigney et al., 2023a), reduced the number of male mice that copulated to ejaculation without significant changes in ejaculation latency. We cannot easily reconcile these different effects, but this may reflect an unbalancing of BNST AVP outputs as our manipulation targeted only one of these BNST AVP target zones.

As GABA cells predominate in the vPal (Root et al., 2015) and V1aR is mostly expressed in these cells (Lee et al., 2024), AVP may primarily excite GABA vPal circuits. While excitation of vPal GABA cells increases reward-seeking behavior in other contexts (Faget et al., 2018; Farrell et al., 2021; Heinsbroek et al., 2020; Tooley et al., 2018), the more specific effects of V1aR partial knockdown, limited primarily to male-male interest, suggests that activity was reduced in a limited set of V1aR+ GABA neurons in vPal. Mapping the outputs of these vPal V1aR-expressing neurons may illuminate this specificity of function.

Reductions in male-male social investigation similar to those observed after V1aR partial knockdown also occurred after blocking V1aR in the mouse lateral septum (LS; (Rigney et al., 2024)), a brain structure that receives strong AVP innervation from BNST cells as well (De Vries and Buijs, 1983; Rigney et al., 2023a). Similarly, pharmacological blockade of V1aR in both the LS and in the vPal have similar effects on rat play fighting (Lee et al., 2024; Veenema et al., 2013). This pattern of findings suggests, along with known connectivity of LS and vPal (Groenewegen et al., 1993; Risold and Swanson, 1997), that V1aR cells in the vPal and LS are interconnected and are part of an integrated V1aR-BNST AVP circuit that drives sex differences in social behavior.

Highlights:

• RNA interference reduces vasopressin 1a receptor (V1aR) in ventral pallidum (vPal)

• V1aR partial knockdown in vPal reduces male-male investigation and ejaculation latency

• V1aR partial knockdown in vPal does not influence female social investigation or sex behavior

• V1aR partial knockdown does not alter communication, aggression, or anxiety/reward responses