Effects of social isolation on 50-kHz ultrasonic vocalizations, affective state, cognition, and neurotransmitter concentrations in the ventral tegmental and locus coeruleus of adult rats

Abstract

Vocal communication, cognition, and affective state are key features of sustained health and wellness, and because vocalizations are often socially-motivated, social experience likely plays a role in these behaviors. The monoaminergic systems of the ventral tegmental area (VTA) and the locus coeruleus (LC) are associated with social and reward processing, vocalization production, and neurotransmitter changes in response to environmental stressors. The effect of social isolation on these complex behaviors and the underlying neural mechanisms is relatively unknown. To add to this body of literature, we randomized adult male Long-Evans rats to control (housed with a cagemate) or isolated (housed individually) conditions and assayed ultrasonic vocalizations, cognition (novel object recognition test), anxiety (elevated plus maze) and anhedonia (sucrose preference test) at 2, 4, 6, 8, and 10 months of age. At 10 months, VTA and LC samples were assayed for dopamine, norepinephrine, and serotonin using high performance liquid chromatography. We tested the hypotheses that isolation 1) diminishes vocalizations and cognition, 2) increases anxiety and depression, and 3) increases levels of dopamine, norepinephrine, and serotonin in the VTA and LC. Results showed isolation significantly reduced vocalization tonality (signal-to-noise ratio) and increased maximum frequency. There were no significant findings for cognition, anxiety, or anhedonia. Dopamine and serotonin and their respective metabolites were significantly increased in the VTA in isolated rats. These findings suggest chronic changes to social condition such as isolation affects vocalization production and levels of VTA neurotransmitters.

1. Introduction

Social isolation is a form of chronic stress that contributes to psychological disorders such as anxiety and depression [1], as well as diminished cognitive function [2,3], and increases the likelihood of early mortality [4]. While social isolation can occur in typical circumstances, such as in adults living in rural areas or with remote or secluded professions [5], the risks and consequences are greater for aging populations and individuals with degenerative conditions, where the prevalence of social isolation is even higher [6–8]. Globally, 50% of individuals over 60 years of age are at risk for social isolation [9]. Additionally, social isolation is linked to a 55% increased risk of dementia [10], as well as an increase in reported depression and anxiety symptoms [11]. Little is known, however, regarding how social isolation contributes to the development of these behaviors, how it impacts vocal communication, or how neural function may be disrupted. As such, addressing how social isolation effects neurotransmitters in regions that play a role in the modulation of vocal and non-motor behaviors is an important first-step.

The underlying and converging neural systems governing stress response to isolation, socially-mediated vocal communication, cognitive function, and affect are complex and not well-defined. The ventral tegmental area (VTA) is a region in the midbrain largely comprised of dopaminergic (DA) neurons that are involved in social reward and motivation pathways and vocal communication [12,13]. The VTA receives projections with both dopaminergic and non-dopaminergic neuronal activity that activate reward pathways and contribute to social behaviors such as reward learning, anticipation, and approach behaviors [14,15]. Additionally, DA neurons of the VTA are reciprocally connected to the nucleus accumbens, basolateral amygdala, and prefrontal cortex, and are likely involved in the distinct, but overlapping functions of vocal communication [16–19], cognitive function [20], and response to stress and fear [21,22]. Additionally, the VTA dopaminergic and serotonergic (5HT) systems are susceptible to changes in environmental stress. Studies in rats found an increase in VTA DA following socially-related stressors in adolescence and in social defeat models using resident-intruder paradigms [23–25]. A similar increase in VTA DA and metabolites was also found in adult rats following a brief, three-week social isolation paradigm [26,27]. Additionally, new electrophysiological data supports that 5HT mediates DA activation directly and indirectly through various 5HT receptor subtypes in the VTA [28].

The locus coeruleus (LC)-norepinephrine (NE) system is also critical in modulating stress-responses and has a powerful effect on the regulation of anxiety-related behavior. The LC is a small cluster of brainstem nuclei that has extensive projections to multiple central nervous system areas including the spinal cord, cortical areas such as the prefrontal cortex, subcortical areas including the amygdala and thalamus, and brainstem nuclei such as the VTA and substantia nigra [29,30]. Thus, the LC-NE system is responsible for a myriad of functions and dysfunction is linked to several neurodegenerative and neuropsychiatric disorders involving anxiety and cognition [31,32]. In rats, an increased activation of the LC-NE system has been linked to several types of stressful environmental stimuli, including social-intruder paradigms, pain, and noise, further validating its importance in stress response [33–36]. Additionally, the LC-NE system is positively correlated with socially-mediated vocal production, specifically in complexity of vocalizations produced [37–41]. As these mechanisms of modulation in the VTA and LC have not been fully-defined, especially in response to stress, first determining the neurochemical changes caused by chronic stress (isolation) in these regions, is warranted.

The majority of studies reported in the literature focus on the immediate isolation of post-weaning and young rat pups or acute isolation in adult rats. For example, Seffer and colleagues found that brief isolation (four weeks) of adolescent rats disrupted approach behavior to social 50-kHz vocalizations, which was not replicated in isolated adult rats, highlighting the differential impact of isolation on social communication and behaviors based on rat age [42]. As such, the effect of long-term or chronic social isolation in adulthood that is analogous to human social isolation remains poorly understood. Rats are innately social mammals that produce ultrasonic vocalizations (USVs) in a variety of social contexts that carry communicative intent [43–46]. Specifically, 50-kilohertz (kHz) USVs are produced during positive affective states and social and mating encounters [47–51]. Similar to human speech, USVs vary in parameters such as intensity (loudness), frequency (pitch), and call complexity and are capable of changing the behavior of the recipient [51–53]. Additionally, rat USVs and vocal communication in humans share similar central and peripheral physiological mechanisms that govern sensorimotor control [54–60]. Thus, rat models afford us the necessary experimental control to systematically test the hypotheses that social isolation disrupts (a) socially-mediated ultrasonic vocal communication, (b) cognitive function, and (c) affective state in rats compared to non-isolated control rats [61–65]. Because male and female rats differentially respond to social stress in both behavior and neurological outcomes [66,67] and vocalizations produced by female rats are dependent upon estrus cycle [68–70], we used male rats as a necessary first step to determine the impact of isolation and the female sex should be investigated separately.

Thus, the purpose of this study was to investigate changes to vocalizations, cognitive function, affect, and neurochemical changes as a result of chronic social isolation in adult male rats. We hypothesized that (1) social isolation would alter USV acoustics, increase anxiety behaviors and anhedonia, and impair cognition to a greater extent over time and (2) DA and 5HT concentrations would increase in the VTA and NE in the LC would also increase in response to the chronic stress of social isolation. This is the first work to simultaneously investigate the impact of long-term, chronic, social isolation in adult male rats on vocalization production, cognitive function, affect, and alterations of neurochemicals in regions important for these complex, socially-motivated functions.

2. Material and methods

2.1. Experimental procedures

To determine the impact of social isolation on vocalization, cognition, anxiety, and anhedonia over time and neurotransmitter concentrations in the VTA and LC after a period of chronic stress, we randomized adult male wild-type (WT) rats into socially-isolated or control (paired) housing conditions at 2 months of age. Rats were assayed for vocalization, cognition (novel object recognition task), anxiety (elevated plus maze), and anhedonia (sucrose preference test) at 2 (baseline), 4, 6, 8, and 10 months of age. Following data collection at 10 months of age, rats were euthanized, and VTA and LC samples were analyzed, described below.

2.2. Animals

All procedures were approved by the University of Wisconsin-Madison School of Medicine and Public Health Animal Care and Use Committee (SMPH IACUC; Protocol M006390) and were conducted in accordance with the NIH Guide for the Care and Use of Laboratory Animals, Eight Edition [71]. Sixteen male Long-Evans rats arrived at 4 weeks of age and housed in pairs (Envigo™ Research Labs, Boyertown, PA, USA). At 2 months of age, rats were randomized into two experimental conditions: 1) control, housed with a cage-mate and 2) isolated, housed individually. An additional subset of female rats (n = 6) were used to elicit male vocalizations. All rats were housed in pairs in standard polycarbonate enclosures (290 mm × 533 mm x 210 mm) on a 12-hour reverse light cycle in accordance with our husbandry protocols. To test rats when they are in their most active state, testing occurred during the dark period in partial red-light illumination. Food and water were provided ad libitum, except during the sucrose preference test, described below.

2.3. Social isolation and control housing conditions

Rats underwent all behavioral assays described below at baseline (2 months) and then were randomized to social isolation or control housing conditions. In the social isolation condition, rats were housed singly in individual cages starting at two months of age until the completion of all behavioral assays at ten months of age. Rats in the control condition were housed with one same-sex cagemate. Health and growth were monitored by daily inspection for signs of severe stress and weighed weekly. No animal welfare concerns were noted throughout the study.

2.4. Ultrasonic vocalizations

Rats communicate using USVs in a variety of contexts such as social and mating interactions [43–46]. USVs were recorded using an established mating paradigm to elicit male calls, which is a reliable and consistent method to assess vocal communication [17,37,38,51,72–77]. This consisted of placing the male rat alone in his home cage with a sexually receptive female, allowing the male to show interest (mounting, chasing, sniffing) for up to five minutes and then removing the female to record male-only vocalizations. All rats were acclimated for ten days to this paradigm prior to testing. USVs were recorded with an CM16 microphone (Avisoft, Berlin, Germany), that has high directional properties, a flat frequency response of up to 150-kHz, working frequency response range of 10–180-kHz, with 16-bit resolution, and a sampling rate of 250-kHz. USVs were analyzed offline using Deepsqueak software in MATLAB, which automatically detected and analyzed the following selected USV acoustic parameters: tonality (signal-to-noise ratio), peak frequency (kHz), maximum frequency (kHz), and duration (ms) [78]. USV parameters were averaged per animal per timepoint. Density plots were used to visualize the USV parameters. Both frequency standard deviation, and bandwidth were bimodally distributed. Therefore, using finite mixture models, we determined a cut-off value to separate USVs into simple and complex vocalization categories (Fig. 1) and complex vocalizations, with greater frequency modulation, were selected as our investigative targets [79]. While some authors suggest that the further subcategorization of rodent vocalizations may provide additional information regarding social context [80], due to the exploratory nature of this study, a more simplistic dichotomy of vocalizations was implemented as these complex vocalizations have been associated with a variety of positive situations such as social interactions and rewards [81, 82]. Non-acoustic parameters included call type and call complexity (%).

Fig. 1.

Representative spectrograms of simple (left) and complex (right) ultrasonic vocalizations of one adult male rat in the control condition.

2.5. Novel object test of cognitive function

The novel object recognition test is often used as a measure of cognitive function and involves placing a familiar and novel object in a transparent enclosure (30 cm × 40 cm x 44 cm) and measuring exploratory behavior during familiarization and testing phases [83,84]. This test takes advantage of the innate exploratory nature of rats when in the presence of novel stimuli [84]. The novel object side was counter-balanced to decrease risk for side biases between trials. Behavior was video recorded for offline experimental rater analysis. Exploratory behavior was defined as the rat’s nose within 2 cm of the object [83]. The amount of time spent exploring objects during the test phase was calculated prior to the familiarization phase to ensure the rater is blinded to object assignment (novel vs. familiar). To account for potential confounds with exploration motivation, we used a recognized discrimination ratio (D), which calculates the difference between time the animals spent exploring the novel and familiar objects in the testing phase over the total amount of time the animal spent exploring objects in the testing phase [83]. A larger discrimination ratio is indicative of a more accurate ability to discriminate the novel from the familiar object, suggesting a greater cognitive function [83,85–87]. Outcome variables were averaged per rat per timepoint.

2.6. Elevated plus maze anxiety assay

An elevated plus maze was used to assay anxiety behaviors [88,89]. This assay indexes anxiety behavior based on exploration in closed and open spaces. Rats naturally explore both closed and open spaces, but explore open spaces less frequently when anxious [88]. The apparatus of the elevated plus maze is constructed of fiberboard with a black acrylic surface and consists of four arms (two open without walls and two enclosed by 40 cm high walls), is 50 cm long and 10 cm wide and is raised 50 cm off of the ground. During a two-day acclimation period, rats were handled by examiners for 5 min per day to decrease the risk for an unfamiliar environment to trigger anxiety behaviors for this test. During the testing day, the animals were placed into the center of the four-arm maze (facing an open arm opposite to where the experimental handler is) for a total of 5 min [90]. These trials were video recorded to allow for offline analysis. Dependent variables for this task included: number of open and closed arm entries (#), time spent in closed and open arms (sec), and averages of these variables were calculated for each rat per timepoint.

2.7. Sucrose preference test of anhedonia

To examine the presence of anhedonia, an indicator for depression, the sucrose preference testing paradigm was used [91–93], which applies the natural bias of rodents to prefer sweetened drinking water to indicate a positive affective state. No preference is an indicator of anhedonia, or lack of interest in rewarding behavior [94]. Rats were individually placed into a novel cage identical to their home cage and acclimated for 3 days to drink from two 50mLwater bottles both filled with normal filtered water for 2 h each day. On testing day, rats were offered two counterbalanced bottles: one 50 mL bottle filled with 8% sucrose solution water and one 50 mL bottle filled with normal filtered water. After two hours, the volumes of the water bottles were measured (mL) to determine sucrose and filtered water intake and the average consumption was calculated for each group.

2.8. Brain processing and high-performance liquid chromatography

Upon the completion of all behavioral testing at 10 months of age, rats were deeply anesthetized with isoflurane and rapidly decapitated. Brains were dissected and frozen at − 80 °C. Using a cryostat at − 15 °C, 250 µm brains were sliced in the coronal plane and mounted on gelatin-coated slides. Using the Brain Punch Set (FST 18035–02, Foster City, CA, USA), 2 mm micro punches of the VTA and LC were collected bilaterally and stored in 1.5 mL microcentrifuge tubes at − 80 °C. Samples were homogenized using 0.5 mL of 0.2 M perchloric acid and 100 mg isoproterenol in 0.1 M acetic acid including 2 mg/mL EDTA-2Na per 100 mg wet tissue weight. Homogenates were sonicated on ice, placed in an ice bath at 4 °C for 30 min, then centrifuged for 15 min at 13,000 rpm at 4 °C. Each sample’s supernatant was collected in a 1.5 mL microcentrifuge tube. Sample pH was modified to pH of 3.0 with 1 M sodium acetate. Samples were filtered through 1.5 mL spin column and stored on dry ice. All samples were analyzed at one time by Amuza Inc. (San Diego, CA) using an Eicom HTEC-510 HPLC-ECD with a graphite electrode, an Eicom SC-3ODS reverse phase C18 column, and an AS-700 autosampler. HPLC analysis occurred at 25 °C, with a + 750 mV vs. Ag/AgCl detector setting and a 0.4 mL/min flow rate. The mobile phase was 0.1 M citrate-acetate buffer (pH 3.5), 15% methanol, 190 mg/l sodium decane-1-sulfonate, and 5 mg/l EDTA. Concentrations (pg/uL) for DA, DOPAC, HVA, 3MT, NE, Epi, MHPG, 5HT, HIAA, and dopamine and serotonin turnover rates (defined in Table 3) were quantified and analyzed by brain region (LC, VTA) and social condition (control, isolation).

Table 3.

Summary of HPLC neurotransmitter and metabolite data presented with the mean (SEM) for each social condition group in ventral tegmental area (VTA) and locus coeruleus (LC).

HPLC Metabolite Concentration (pg/uL)	Definition	Condition	VTA	LC
Dopamine (DA)	Neurotransmitter	Control	129.56 (22.46)	17.72 (4.76)
		Isolated	296.24 (54.02)	30.83 (12.62)
Norepinephrine (NE)	Neurotransmitter	Control	254.49 (22.50)	361.92 (21.97)
		Isolated	229.90 (29.20)	429.68 (45.634)
Epinephrine (Epi)	Neurotransmitter	Control	5.00 (0.67)	3.15 (0.51)
		Isolated	4.14 (0.31)	3.38 (0.30)
Serotonin (5HT)	Neurotransmitter	Control	68.85 (5.97)	70.29 (6.34)
		Isolated	116.25 (15.51)	83.32 (4.94)
3,4-Dihydroxyphenylacetic Acid (DOPAC)	Metabolite of dopamine	Control	45.13 (8.11)	16.14 (0.66)
		Isolated	77.64 (12.76)	21.38 (4.36)
Homovanillic Acid (HVA)	Metabolite of dopamine catabolism	Control	21.10 (2.02)	15.00 (0.85)
		Isolated	41.18 (8.75)	16.78 (2.38)
3-Methoxy-4-Hydroxyphenethylamine (3MT)	Metabolite of dopamine	Control	2.51 (0.33)	1.40 (0.28)
		Isolated	5.05 (0.87)	1.79 (0.97)
3-Methoxy-4-Hydroxyphenylglycol (MHPG)	Metabolite of norepinephrine	Control	27.03 (3.20)	13.61 (2.07)
		Isolated	34.79 (9.11)	11.81 (2.86)
5-Hydroxyindoleacetic Acid (HIAA)	Metabolite of serotonin	Control	54.31 (4.47)	60.46 (4.66)
		Isolated	76.74 (12.43)	61.10 (6.47)
Dopamine (DA) Turnover	Turnover rate of dopamine	Control	56.50 (5.99)	209.45 (51.46)
		Isolated	42.91 (3.20)	195.33 (34.76)
Serotonin (5HT) Turnover	Turnover rate of serotonin	Control	85.4 (13.00)	92.00 (10.45)
		Isolated	107.36 (18.06)	73.80 (7.11)

2.9. Statistical analyses

All statistical analyses were conducted with SigmaPlot^® 12.5 (Stat Software, Inc., San Jose, CA). To investigate the impact of social condition, two-way repeated measures analysis of variances (ANOVA) with factors of condition (control, isolation) and age (2, 4, 6, 8, and 10-months) were used. Post hoc analyses were performed with Fisher’s Least Significant Difference Method using a Holm-Sidak correction for multiple comparisons. Shapiro-Wilk test of normality and Levene’s test of equal variances confirmed adherence to ANOVA assumptions. When Shapiro-Wilk test failed normality, a Mann-Whitney Rank Sum Test was used. To compare concentrations of brain neurotransmitters and metabolites in the VTA and LC, student’s t-tests were used for each outcome variable. Significance levels were set a priori at 0.05.

3. Results

3.1. Ultrasonic vocalizations

All variables (means +/− SEM) for each acoustic parameter of complex vocalization are in Table 1. Additionally, while not discussed below, all variables (means +/− SEM) for each acoustic parameter of simple vocalizations are located in Table A.1 (appendix).

Table 1.

Mean (SEM) for each acoustic parameter of complex ultrasonic vocalizations at each testing timepoint for each social condition.

Acoustic Parameters	Condition	2 months	4 months	6 months	8 months	10 months
Tonality (#)	Control	0.40 (0.01)	0.44 (0.09)	0.43 (0.01)	0.43 (0.01)	0.41 (0.01)
	Isolated	0.37 (0.01)	0.40 (0.10)	0.41 (0.01)	0.40 (0.01)	0.39 (0.01)
Peak Frequency (kHz)	Control	59.37 (1.31)	55.44 (1.10)	57.34 (1.20)	56.19 (1.20)	56.72 (0.99)
	Isolated	59.15 (1.24)	58.43 (1.05)	59.74 (1.05)	58.85 (1.05)	61.04 (1.05)
Max Frequency (kHz)	Control	74.40 (1.37)	73.15 (1.15)	73.34 (1.26)	71.54 (1.26)	73.56 (1.03)
	Isolated	72.32 (1.30)	74.54 (1.09)	77.64 (1.09)	75.98 (1.09)	78.49 (1.09)
Duration (s)	Control	0.041 (0.004)	0.054 (0.004)	0.046 (0.004)	0.046 (0.004)	0.042 (0.003)
	Isolated	0.037 (0.003)	0.039 (0.003)	0.040 (0.003)	0.047 (0.003)	0.042 (0.003)

3.1.1. Percent complex vocalizations

There was a significant interaction between condition and timepoint (F(4,50) = 2.59, p = 0.048) for percent of complex vocalizations with a decrease from the 2- to 10-month timepoint within the control group (p = 0.046) (Fig. 2A). There were no other significant differences in number of USVs or percent of complex USVs.

Fig. 2..

Summary of ultrasonic vocalization data. A.) Percentage of complex vocalization out of total vocalizations. B.) Average tonality of complex vocalization collapsed by timepoint to demonstrate main effect of condition. C.) Average tonality of complex vocalizations collapsed by condition to illustrate main effect of timepoint. D.) Max frequency of complex vocalizations. Box and whisker plots illustrate min and max values, upper and lower quartile range and the median for each group. Bars indicate significance with asterisk indications showing level of significance: *p < 0.05, **p < 0.01.

3.1.2. USV acoustic parameters

There were no significant interactions for condition and timepoint for tonality of complex vocalizations. There were significant main effects of condition (F(1,50) = 5.67, p = 0.030) and timepoint (F(4,50) = 5.12, p = 0.002) (Fig. 2B and C). Post hoc analyses showed that the control group produced calls with increased tonality compared to the isolated group (p = 0.044) and that there were increases in tonality from 2 months to 4 months (p = 0.005), 6 months (p = 0.008), and 8 months (p = 0.010) of age regardless of condition.

There were no significant interactions or main effects for peak frequency of complex vocalizations. There was a significant interaction between condition and timepoint (F(4,50) = 2.94, p = 0.029) for the max frequency of complex vocalizations (Fig. 2D). Within the isolated group, there were significant increases between the 2- and 6-month (p = 0.026) and the 2- and 10-month (p = 0.007) timepoints, for maximum frequency. There were no significant interactions or main effects for duration of complex vocalizations.

3.2. Cognitive function

There were no significant interactions or main effects for time exploring the novel object or for the discrimination ratio (D) (Table 2).

Table 2.

Mean (SEM) for each behavioral parameter at each testing timepoint for each social condition.

Behavioral Parameters	Condition	2 months	4 months	6 months	8 months	10 months
USV % Complex (%)	Control	59.48 (4.10)	48.73 (3.46)	48.47 (3.78)	47.06 (3.78)	44.28 (3.09)
	Isolated	45.75 (3.91)	48.88 (3.28)	54.90 (3.28)	49.90 (3.28)	51.44 (3.28)
EPM Time in Closed Arms (s)	Control	184.63 (16.30)	227.65 (13.87)	165.32 (22.30)	241.46 (13.87)	214.84 (13.87)
	Isolated	206.67 (14.71)	218.63 (14.71)	162.65 (22.06)	216.69 (14.71)	155.66 (14.71)
Sucrose Consumed (mL)	Control	23.17 (2.62)	24.30 (2.11)	21.40 (2.11)	22.70 (2.11)	24.58 (2.26)
	Isolated	17.61 (2.57)	20.89 (2.57)	23.00 (2.36)	28.13 (2.36)	31.13 (2.36)
Novel object exploration time (s)	Control	6.63 (4.98)	5.75 (2.90)	9.50 (17.01)	2.67 (2.52)	1.5 (0.71)
	Isolated	7.56 (6.18)	4.13 (2.06)	3.13 (1.75)	3.25 (3.30)	4.75 (4.35)

3.3. Anxiety

There were no significant interactions for time in the closed arms of the elevated plus maze. There was a significant main effect of timepoint (F(4,49) = 5.04, p = 0.002), with post hoc analysis showing a significant decrease from 4- to 6-months (p = 0.023), an increase from 6- and 8-months (p = 0.010), and a decrease from 8- to 10-months (p = 0.028) (Table 2; Fig. 3).

Fig. 3.

Elevated plus maze time in closed arms with data collapsed by condition to illustrate main effect of timepoint. There were significant differences between 4- and 6-month, 6- and 8-month, and 8- and 10-month timepoints. Box and whisker plot illustrate min and max values, upper and lower quartile range and the median for each timepoint. Bars indicate significance with asterisk indications showing level of significance: * p < 0.05.

3.4. Anhedonia

There were no significant interactions for the volume of sucrose consumed. There was a main effect of timepoint (F(4,58) = 3.07, p = 0.023; specifically, that there was a significant increase from the 2- to 10-month timepoints (p = 0.035) (Table 2; Fig. 4) for volume of sucrose consumed.

Fig. 4.

Sucrose volume consumed (mL) during sucrose preference test. Data for sucrose consumption is collapsed to illustrated significance between 2- and 10-month timepoint. Box and whisker plot illustrate min and max values, upper and lower quartile range and the median for each timepoint. Bars indicate significance with asterisk indications showing level of significance: * p < 0.05.

3.5. High-performance liquid chromatography

Summary of HPLC neurotransmitter and metabolite data (means +/− SEM) are presented in Table 3. There were significant differences between the control and isolated groups in the VTA for concentrations of dopamine (DA) (T = 86.00, p = 0.006), 3,4-dihydroxyphenylacetic acid (DOPAC) (t(14) = −2.24, p = 0.02), homovanillic acid (HVA) (T = 83.00, p = 0.015), 3-methoxy-4-hydroxyphenethylamine (3MT) (t (14) = −3.016, p = 0.005), serotonin (5HT) (t(14) = −3.13, p = 0.004), and 5-hydroxyindoleacetic acid (5HIAA) (t(14) = −1.80, p = 0.048) (Fig. 5), such that isolated group had significantly increased concentrations compared to the control group. There were no significant differences between the control and isolation groups in the VTA in norepinephrine (NE), epinephrine (Epi), or 3-methoxy-4-hydroxyphenylglycol (MHPG) concentrations, in DA or 5HT turnover rates (Fig. 6), or for any HPLC variables in the LC (Table 3).

Fig. 5.

Average concentration (pg/uL) (+/− SEM) of DA (A), DOPAC (B), HVA (C), and 3MT (D), 5HT (E), and HIAA (F) in micro punches of the VTA (Bregma −4.80). Rats in the isolation group had significant increases in levels of DA, DOPAC, HVA, 3MT, 5HT, and HIAA compared to the control group (housed in pairs). Sample sizes are as follows: N = 16, n = 8 per group. Box and whisker plots illustrate min and max values, upper and lower quartile range and the median for each group. Bars indicate significance with asterisk indications showing level of significance: *p < 0.05, **p < 0.01.

Fig. 6.

Average turnover rates of DA and 5HT in the ventral tegmental area (VTA) (A.) and the locus coeruleus (LC) (B). There were no significant differences between groups. Sample sizes are as follows: N = 16, n = 8 per group. Box and whisker plots illustrate min and max values, upper and lower quartile range and the median for each group.

4. Discussion

The purpose of this study was to determine the impact of chronic social isolation on vocal communication, cognitive function, affect, and neurotransmitters in the VTA and LC in adult male rats. We hypothesized that USVs would be disrupted in response to social isolation in vocal parameters. Results show that socially isolated rats had reduced vocal tonality (signal-to-noise ratio) in USVs compared to the control rats that were housed in pairs, regardless of timepoint. These findings are complementary to results from an acute social isolation paradigm of aged rats which demonstrated a reduction of vocal loudness in the social isolation group compared to the control rats [64]. Additionally, all rats showed an increase in vocal loudness at 4, 6, and 8-months compared to the 2-month timepoint, regardless of social condition. This has been replicated in previous work from our lab and is an expected consequence of growth as laryngeal structures increase in size during adulthood [73]. Of note, advanced aging in the rat (past 20 months of age) leads to a gradual decline in vocal loudness [95–97]. This should be considered for future work, as the advanced aging population is also high risk for isolation and effects could be exacerbated by advanced age.

Here, we also show a significant increase in maximum frequency at 6 and 10-months compared to the 2-month timepoint in the isolated rats. This pattern was not observed in the control (pair-housed) rat group. Interestingly, previous work from our lab has identified a decrease in peak frequency over time in control-housed WT rats at these timepoints thought to be due to normal aging [73]. Additionally, Peterson et al. (2013), has found age-related decreases in peak frequency as well as diminished elastin, hyaluronic acid, and collagen in the lamina propria of vocal folds of advanced aged rats (18 months). Because the isolated rat group in our study demonstrated increased frequency, an effect in the opposite direction of what is observed in healthy aging, we suspect this alteration is due to social isolation and not due to changes we could expect from the normal aging process. A similar increase in frequency is observed when rat pups are isolated from their mother and littermates, thought to be in response to an adverse event and need for social contact [98,99]. While our study investigated only adult rats, this may contribute to the understanding that USV frequency is an important indicator for affective state, however, further investigation is warranted to determine the ecological and social relevance of this change.

We also hypothesized that social isolation would disrupt cognition, anxiety, and anhedonia compared to control conditions. While our results did not find a decline in cognitive abilities or increase anxiety or anhedonia, which is not consistent with other findings [100–102], there are several factors that could account for this and make the results of our work difficult to compare to the current literature. One possible explanation is that our frequent testing and handling procedures may have attenuated the effects of being socially isolated by being housed singly. As part of our animal care and use protocols, all rats were handled daily to inspect for overall health and well-being. Rats in the control condition were also housed in pairs (with a cagemate) and housing them in groups should be explored in the future. Additionally, because our aim was to measure function over time, rats underwent regular testing in multiple assays (at 2, 4, 6, 8, and 10-months), which subsequently led to frequent acclimation to testing procedures (on average 20 days per month), handling by laboratory staff, and exposure to multiple stimulus female rats to elicit vocalizations in the social-mating paradigm for several days prior to testing. This could have inadvertently served as social stimulation to overcome affective differences caused by isolation and thus diminished the behavioral impacts we anticipated. Finally, it should be noted that age was likely a factor as many studies implemented isolation protocols in various stages of adolescence, in contrast to our study, which began isolation procedures when rats were young adult [103–105]. This period of adolescence in both rat and humans is a critical developmental period for stress and emotional regulation and behavior [106–109]. Thus, previous work that has demonstrated increased anxiety behavior during the post-weaning period may be capturing changes during adolescent development, resulting in more robust alterations to cognition, anxiety, and anhedonia behaviors.

Specifically with regard to cognitive function, our study did not demonstrate differences in cognition as a function of social isolation, in contrast to other studies [61,101,110,111]. Again, the results of other studies are difficult to extract and compare to our work, however, due to the different periods of isolation, age of isolation, as well as cognitive testing protocols. For example, Schiavone et al. (2012), began social isolation of male rats at postnatal day 21 and found differences in novel object discrimination abilities after four weeks of social isolation [101]. However, the rats in this study were raised in litters and socially housed until the commencement of experimental procedures at 2 months of age. We also suspect that our different findings could be due to the cognitive task we selected, as there are a variety of cognitive assays that measure different aspects of cognitive performance. For example, Krupina et al. (2020) and Zorzo et al. (2019) found that socially-isolated adult rats performed worse during the Moris Water Maze, which assays spatial memory as opposed to recognition memory, as is in the case in the novel object task [61,111].

We expected to see an increase in anxiety behaviors as a result of social isolation over time. In contrast, we found reduced anxiety behaviors at 6 months of age, but for both isolated and control rats, which differs from what other studies have shown [109,112–114]. Importantly, the length and implementation timeline of isolation varies across studies, and thus, comparisons are difficult to discern. For example, Chappell et al. (2013), began isolation at postnatal day 28 and after six weeks, found increased anxiety behavior in the elevated plus maze [100]. Conversely, Viana Borges et al. (2019) began isolation protocols at three months of age (young adulthood) and after 50 days of isolation found no difference in anxiety behavior using the elevate plus maze or the open field test, another assay to measure anxiety in rodents [115]. Together, these findings reiterate the importance of age of isolation implementation and the long-term impact on anxiety behavior. Additionally, as described above, our frequent handling and behavioral testing may have attenuated anxiety behaviors in both groups over time, effectively washing-out any influence that social isolation may have had on anxiety behaviors associated with this assay.

Likewise, we did not see an effect of social isolation on anhedonia, a clinical indicator for depression. Zlatkovic et al. (2014) and Manouze et al. (2019) instead found that isolated adult male rat groups demonstrated a decreased preference for sucrose compared to rats housed in pairs [102,116]. Interestingly, Manouze et al. (2019) also found that isolated rats that were handled frequently by laboratory staff showed increased sucrose water consumption as compared to the pair-housed group [116]. This is an important consideration when determining the impact on social isolation in regard to anhedonia behaviors in adult rats, as housing condition may not be the only type of social interactions that could influence behaviors and thus, potentially improve affective state.

Based on the results of our HPLC analysis in the VTA, social isolation increased DA, DOPAC, HVA, 3MT, 5HT and HIAA in adult rats. This agrees with Hamed et al. (2015), that found adult rats isolated for a brief, three-week period, starting at 3 months of age had increased DA, DOPAC, HVA, and 3MT in the VTA compared to groups housed in pairs measured with HPLC [26]. In contrast to our study, however, they did not find differences in 5HT or HIAA in the VTA and also found that this period of social isolation increased the number of 50-kHz vocalizations during pair-encounters when compared to pair-housed rats. Together, this suggests that ages and/or length of time of social isolation may influence changes in the dopaminergic and serotonergic systems in the VTA, and thus, impact vocal communication in different ways.

As the VTA projects to several brain regions that co-innervate and modulate behavior, this observed increase in dopamine, serotonin, and metabolites, may have downstream effects on emotional response and regulation of ultrasonic vocalization production. For example, dopamine projections from the VTA innervate the nucleus accumbens and increased dopamine release here regulates pro-social 50-khz vocalizations [26,117]. However, this relationship is complex, and an increase in dopaminergic content in the VTA alone does not confirm a positive-state reaction, as increased VTA dopamine is also associated with aversive stimuli such as pain [20,118,119]. While HPLC provides us with excellent quantification methods that allows for the detection of more sensitive changes to neurotransmitter concentrations, the particular dissection technique used here, while focused on the specific brain targets of interest, also includes adjacent regions. Other immunohistochemistry assays or neuroimaging techniques such as electroencephalography (EEG), functional magnetic resonance imaging (fMRI), or positron emission tomography (PET) techniques would afford substantial insight into disease-specific functional and structure outcomes and should be considered in future work.

This study had a few notable limitations. First, we did not include female rats, which have been shown to demonstrate differential responses to stressful stimuli [66,67]. For example, chronic environmental stress in female rats leads to dysregulation of estrus cycles and causes more anxiety, depressive behaviors, and stress-related hormonal increases compared to male rats that underwent the same chronic stress protocols [120]. As such, female rats may demonstrate more robust adverse reactions to social stresses in terms of behavior and in neurological findings and the results of this study is not generalizable to the female sex. Thus, the inclusion of female rats in the study of social isolation on vocal communication, cognition, affect, and neurochemical changes is warranted in additional investigations by our lab and others. Second, as mentioned above, our frequent health checks and behavioral assays may have attenuated some social isolation effects on vocalizations, condition, anxiety, and anhedonia. Third, the behavioral measures only reflect specific aspects of complex cognitive behavior, anxiety, and anhedonia. Using a broader range of or alternative behavioral assays to more comprehensively assess these behaviors of interest may result in a more nuanced understanding of these complex behavioral outcomes with regard to isolation.

In summary, this study demonstrated that vocal dysfunction and changes to neurotransmitters in the VTA of adult male rats were present following chronic social isolation, in the absence of cognitive, anxiety, or anhedonia disruptions. The results of this work support the need for continued research on the impact of social isolation (stress) on the progression of communication and affective function in the normal aging process as well as disease states that confer more vulnerability to these systems.

Abbreviations:

USV

ultrasonic vocalizations

WT

wild-type rat

VTA

ventral tegmental area

LC

locus coeruleus

DA

dopamine

5HT

serotonin

DOPAC

3,4-dihydroxyphenylacetic acid

HVA

homovanillic acid

3MT

3-methoxy-4-hydroxyphenethylamine

5HIAA

5-hydroxyindoleacetic acid

NE

norepinephrine

Epi

epinephrine

MHPG

3-methoxy-4-hydroxyphenylglycol

EPM

elevated plus maze

Appendix A

See Table A1.

Table A1.

Mean (SEM) for each acoustic parameter of simple ultrasonic vocalizations at each testing timepoint for each social condition (control, isolation).

Parameters	Condition	2 months	4 months	6 months	8 months	10 months
Tonality (#)	Control	0.41 (0.02)	0.45 (0.02)	0.45 (0.02)	0.47 (0.02)	0.46 (0.01)
	Isolated	0.36 (0.02)	0.45 (0.02)	0.43 (0.02)	0.43 (0.02)	0.43 (0.02)
Peak Frequency (kHz)	Control	57.66 (1.12)	54.99 (1.04)	56.20 (1.04)	55.36 (1.02)	57.85 (0.95)
	Isolated	56.17 (1.07)	56.68 (1.10)	58.48 (1.07)	57.33 (1.10)	58.05 (1.10)
Max Frequency (kHz)	Control	60.18 (1.09)	58.21 (1.02)	59.29 (1.02)	58.44 (1.00)	60.91 (0.93)
	Isolated	58.46 (1.05)	59.27 (1.08)	61.49 (1.05)	60.02 (1.08)	60.42 (1.08)
Duration (s)	Control	0.03 (0.003)	0.04 (0.002)	0.03 (0.002)	0.03 (0.002)	0.03 (0.002)
	Isolated	0.03 (0.002)	0.03 (0.003)	0.03 (0.002)	0.03 (0.003)	0.03 (0.003)