Abstract
Abstract
The study aimed to test the hypotheses that chronic social instability stress (CSIS) alters behavioral and physiological parameters and expression of selected genes important for stress response and social behaviors. Adult female Sprague-Dawley rats were subjected to the 4-week CSIS procedure, which involves unpredictable rotation between phases of isolation and overcrowding. Behavioral analyses (Experiment 1) were performed on the same rats before and after CSIS (n = 16) and physiological and biochemical measurements (Experiment 2) were made on further control (CON; n = 7) and stressed groups (CSIS; n = 8). Behaviors in the open field test (locomotor and exploratory activities) and elevated-plus maze (anxiety-related behaviors) indicated anxiety after CSIS. CSIS did not alter the physiological parameters measured, i.e. body weight gain, regularity of estrous cycles, and circulating concentrations of stress hormones and sex steroids. QRT-PCR analysis of mRNA expression levels was performed on amygdala, hippocampus, prefrontal cortex (PFC), and hypothalamus. The main finding is that CSIS alters the mRNA levels for the studied genes in a region-specific manner. Hence, expression of POMC (pro-opiomelanocortin), AVPR1a (arginine vasopressin receptor), and OXTR (oxytocin receptor) significantly increased in the amygdala following CSIS, while in PFC and/or hypothalamus, POMC, AVPR1a, AVPR1b, OXTR, and ERβ (estrogen receptor beta) expression decreased. CSIS significantly reduced expression of CRH-R1 (corticotropin-releasing hormone receptor type 1) in the hippocampus. The directions of change in gene expression and the genes and regions affected indicate a molecular basis for the behavior changes. In conclusion, CSIS may be valuable for further analyzing the neurobiology of stress-related disorders in females.
Introduction
Chronic stress is recognized as a major risk factor for several psychiatric disorders, including anxiety and depression (Lee, Jeong, Kwak, & Park, 2010; Lex, Bazner, & Meyer, 2017). The prevalence of these disturbances is highly sex-dependent (Kendler et al., 1995; Piccinelli & Wilkinson, 2000). Women are more likely to develop stress-related disorders, and there is a strong sexual divergence in the susceptibility to adverse life events (Kendler, Gardner, & Prescott, 2002). Recently, individual differences in stress reactivity have been suggested as an important risk factor for the sex-related prevalence rates of mental disorders, in addition to genetic, hormonal, and socio-cultural factors (Kajantie & Phillips, 2006; Young & Altemus, 2004). The dysregulation of the hypothalamic-pituitary-adrenal (HPA) axis and glucocorticoid hypersecretion are linked to the symptomatology and consequences of depression and anxiety (Nestler et al., 2002). Experimental studies indicate that female rodents have elevated baseline corticosterone concentrations (Malisch et al., 2007) and greater and more persistent adrenocorticotropic hormone (ACTH) and corticosterone responses to stress than males (Larkin, Binks, Li, & Selvage, 2010; Young, Altemus, Parkison, & Shastry, 2001). An altered expression of the genes for corticotropin-releasing hormone (CRH) and arginine vasopressin (AVP) in the hypothalamic paraventricular nucleus (PVN) and proopiomelanocortin (POMC) in the anterior pituitary is a frequently observed stress response, reflecting the activation of the HPA axis (Ancelin et al., 2017; Du & Pang, 2015). These processes are partly controlled by the neuronal activity of the hippocampus and prefrontal cortex (PFC), which mediate negative feedback control of HPA axis activation (Nicolaides, Kyratzi, Lamprokostopoulou, Chrousos, & Charmandari, 2015). Stress-coping and anxiety-related behaviors, as well as various types of social behaviors including social recognition and communication, maternal care, and aggression, are modulated by brain vasopressin and oxytocin (Lukas & Neumann, 2013; Neumann, 2008). In the present study changes in AVPR1a, AVPR1b, and OXTR gene expression were selected as endpoints due to their roles in mediating the effects of their respective nonapeptides. Both neuropeptides have generally been described as key regulators of HPA axis activity and stress responses, with both physiological and behavioral aspects (Neumann, 2002). Importantly, vasopressin and oxytocin actions may partly overlap, due to the >85% homology between their receptors (Sala et al., 2011). In animal models, it has been shown that sex steroids are particularly important with regard to central neuropeptide effects (Delville, Mansour, & Ferris, 1996; Uhl-Bronner, Waltisperger, Martinez-Lorenzana, Condes Lara, & Freund-Mercier, 2005). In the study, attention was also focused on estrogen receptor-β (ER-β) because this receptor may mediate a different action from ER-α, such as the non-reproductive actions of estradiol (Handa, Ogawa, Wang, & Herbison, 2012). Furthermore, ER-β was found to co-localize in the rodent hypothalamus with cells expressing OXT and AVP receptors (Hrabovszky et al., 2004).
Considering that susceptibility to mental disease is strongly associated with sex and exposure to environmental factors such as social stress, there is a strong necessity for the development and evaluation of animal models of chronic stress applicable to female rodents. Previously, it was shown that a social instability paradigm is stressful for female rats (Haller, Fichs, Halasz, & Makara, 1999). Herzog et al. (2009) have developed and characterized the 4-week chronic social instability stress (CSIS) protocol, which is based on a rotation of the isolation and overcrowding of female rats. This stress paradigm induced changes in physiological parameters (e.g. adrenal weight and hormone concentrations), but no effects were observed in the forced swim test behavior and hippocampal neurotrophin levels. In order to enhance the unpredictability of the social situation and to avoid a possible adaptation to stress, in the present study, the CSIS procedure was slightly modified by extending the isolation phase and changing the housing parameters during the crowding phase. The effects of CSIS on the central expression of stress-related factors have not previously been tested. Considering that social isolation is an aversive stimulus that increases the activity of the HPA axis (Douglas, Varlinskaya, & Spear, 2004; McCormick, Smith, & Mathews, 2008), it was hypothesized that CSIS alters the expression of the chosen genes important for both social behaviors and stress response, namely POMC, corticotropin-releasing hormone receptor type 1 (CRH-R1), AVPR1a, AVPR1b, OXTR, and ERβ. It was also considered that these changes may be accompanied by alterations in behavioral or physiological parameters. In order to test the study hypotheses, the influence of the modified CSIS procedure was evaluated for (i) anxiety-related behaviors, (ii) peripheral concentrations of HPA axis hormones (ACTH, corticosterone) and sex steroids (estradiol, testosterone), and (iii) the relative gene (mRNA) expression of POMC, CRH-R1, AVPR1a, AVPR1b, OXTR, and ERβ in the brain structures involved in the response to stress.
Methods
Animals and housing
Female Sprague-Dawley rats (Centre for Experimental Medicine of Medical University of Silesia, Poland) aged 60–63 days were used in the experiments. Before the study began, the rats were adapted to their new conditions for one week. Rats were housed 3–4 per cage in a climate-controlled room (22 ± 2 °C, humidity: 55 ± 10%) with a 12 h:12 h light/dark cycle starting at 07:00 h, and received food and water ad libitum. All experiments were carried out in accordance with the guidelines of the European Ethical Standards (2010/63/EU) and were approved by the Local Ethics Committee for the Care and Use of Laboratory Animals (Katowice, Poland). The minimum number of rats required to obtain consistent data was used, and every effort was taken to minimize the suffering of the animals.
Study design
The study was comprised of two separate experiments. In Experiment 1, behavioral analyses were performed on the same female rats (n = 16) before and after the CSIS procedure, which lasted for 28 days as described below. These rats were euthanized using CO2 24 h after the last test, on the 34th day of the experiment. In Experiment 2, rats were randomly assigned to the control group (CON; n = 7), or to the stressed group (CSIS; n = 8) in which the rats were subjected to CSIS. These rats were used for body weight measurements, estrous cycle monitoring, and biochemical analysis. The rats were decapitated without anesthesia 24 h after the last overcrowding phase on the 29th day of the experiment (see timeline diagram, Figure 1).
Published online:
15 September 2017Figure 1. Study design. Before the study began, the rats were adapted to the new condition for several days. The study was comprised of two separate experiments. In Experiment 1, the behavioral analyses (open field test – OF, and elevated plus maze – EPM) were performed on the same rats before and after a 28-day chronic social instability stress (CSIS) procedure (before CSIS and after CSIS groups; n = 16). The rats were euthanized using CO2 24 hours following the EPM test. In Experiment 2, animal weight measurements and samplings of vaginal smears, which were initiated concurrently with the CSIS procedure, were conducted on the unstressed, control rats (CON; n = 7) and rats subjected to CSIS (CSIS; n = 8). Rats were euthanized by decapitation without anesthesia 24 hours following the last CSIS procedure day and samples were collected in order to perform biochemical determinations. ACTH: adrenocorticotropic hormone; CORT: corticosterone; E2: 17β; T: testosterone; ELISA: enzyme-linked immunoassay; genes: POMC: pro-opiomelanocortin; CRH-R1: corticotropin releasing hormone receptor1; AVPR1a, b: arginine vasopressin receptor 1a, b; OXTR: oxytocin receptor; ER-β: estrogen receptor β; qRT-PCR: quantitative reverse transcriptase polymerase chain reaction.
![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/ists20/2017/ists20.v020.i06/10253890.2017.1376185/20171127/images/medium/ists_a_1376185_f0001_c.jpg"})
Figure 1. Study design. Before the study began, the rats were adapted to the new condition for several days. The study was comprised of two separate experiments. In Experiment 1, the behavioral analyses (open field test – OF, and elevated plus maze – EPM) were performed on the same rats before and after a 28-day chronic social instability stress (CSIS) procedure (before CSIS and after CSIS groups; n = 16). The rats were euthanized using CO2 24 hours following the EPM test. In Experiment 2, animal weight measurements and samplings of vaginal smears, which were initiated concurrently with the CSIS procedure, were conducted on the unstressed, control rats (CON; n = 7) and rats subjected to CSIS (CSIS; n = 8). Rats were euthanized by decapitation without anesthesia 24 hours following the last CSIS procedure day and samples were collected in order to perform biochemical determinations. ACTH: adrenocorticotropic hormone; CORT: corticosterone; E2: 17β; T: testosterone; ELISA: enzyme-linked immunoassay; genes: POMC: pro-opiomelanocortin; CRH-R1: corticotropin releasing hormone receptor1; AVPR1a, b: arginine vasopressin receptor 1a, b; OXTR: oxytocin receptor; ER-β: estrogen receptor β; qRT-PCR: quantitative reverse transcriptase polymerase chain reaction.
Chronic social instability stress paradigm
According to the experimental design, the rats were exposed over 28 days to the chronic social stress procedure elaborated by Herzog et al. (2009) and modified by Nowacka, Paul-Samojedny, Bielecka, and Obuchowicz (2014). As an unstable social situation and isolation are strong stressors for female rats, uncontrollability is modeled in the CSIS paradigm by alternating the isolation and crowding phases (details in Table 1). For isolation, rats were individually housed in cages (36 × 20 × 15 cm). For crowding, rats were kept in groups of four in small cages (30 × 20 × 15 cm) for 3 h or 6 h (08:00–11:00 h, or 08:00–14:00 h). To enhance unpredictability, rats were individually rotated among the groups during overcrowding. During the entire experimental period, control (non-stressed) rats in groups of seven were kept in standard cages (52 × 31 cm ×19 cm) in separate rooms within the animal facility.
Table 1. The chronic social instability stress (CSIS) procedure.
Experiment 1
All behavioral tests were performed during the illuminated part of the cycle (08:00 h–11:00 h), under conditions of dim light (∼10 lx) and low noise. Female rats (n = 16) were examined by means of the open field (OF) test and the elevated plus maze (EPM) before and after the CSIS procedure. Tests were recorded and analyzed using the video tracking Ethovison XT 10.1 software (Noldus Information Technology, Wageningen, The Netherlands).
Open field test
The OF test was performed on the 1st and 32nd day of the experiment. Locomotor and exploratory activities were measured using a transparent plastic chamber with a light gray floor (45 × 45 × 40 cm), which was raised 80 cm above the floor. The floor surface of each chamber was thoroughly cleaned with 70% ethanol, dried with paper towels, and left for at least five minutes to completely dry. Each rat was individually placed in the left corner of the OF apparatus and its activity was recorded for 15 min. Movement duration, total distance traveled, and time spent in the center and margins of the chamber were automatically analyzed. The duration of rearing and grooming activities was manually counted.
Elevated plus maze test
The EPM test was performed on the 2nd and 33rd day of the experiment. The EPM apparatus (which was raised 40 cm above the floor) consisted of two opposing open arms (50 cm long and 14 cm wide) and two opposing closed arms of the same size surrounded by side walls (15 cm high) that extended from the central square (14 × 14 cm). The test was conducted on the day after the OF test. At the start of each test, the rats were placed individually on the central platform and their behavior was monitored with a video camera for 5 min. The maze was thoroughly cleaned with 70% ethanol between each rat to eliminate any odor cues. The number of entries and the time spent in each arm (open, closed) were analyzed automatically. The frequency of head dips (protruding the head over the ledge of an open arm and down toward the floor) was counted manually.
Experiment 2
Measurements of body weight
The body weight of each control (n = 7) and stressed (n = 8) rat was measured twice a week for 28 days. The change in body weight during the experiment was calculated as: body weight gain (%) = (final weight ×100)/initial weight).
Estrous cycle monitoring
During the experiment, vaginal smears were collected every morning (07:00–08:00 h) to determine the estrous cycle stage of each rat. A sample of cells of the vaginal canal was obtained with a sterile swab. The sample was placed on a microscope slide, fixed with methanol, stained with Giemsa stain, and dried and examined under a microscope. Cell descriptions were used to classify rats based on the stages of the estrous cycle (proestrus, estrus, metestrus, and diestrus) (Marcondes, Bianchi, & Tanno, 2002).
Sample collection
Brain
Brain regions were isolated according to methods described previously (Sutherland, Burian, Covault, & Conti, 2010). After decapitation, the brain was rapidly removed and put on an ice-chilled metal plate, ventral side up. The entire hypothalamus, excluding the optic nerves, was removed. The brain was turned dorsal side up and hemisected. After the removal of the hypothalamus, septum, and striatum, the left and right hippocampi were removed using curved forceps (approximately 2.0–5.0 mm posterior, 0–4.0 mm lateral, and 2.6–4.0 mm ventral to Bregma) (Paxinos & Watson, 1997). To obtain the amygdala, the brain was placed ventral side up. Using a surgical blade, the piriform cortex was removed from the amygdala and the entire left and right amygdalae were dissected (approximately 3.0–5.4 mm lateral and 7.4–9.4 mm ventral to Bregma). Subsequently, an approximately 2.0 mm-thick slice of the PFC was removed from each hemisphere. The brain structures were weighed and stored in RNAlater® (Sigma-Aldrich, St. Louis, MO) at –20 °C until usage. Trunk blood samples were divided into two parts to obtain the plasma (blood was collected into pre-chilled tubes containing EDTA) or serum (blood was kept at room temperature for 30 min in tubes without anticoagulant). Then, blood samples were centrifuged (10 min, 2500×g, 4 °C) and stored at –80 °C until the hormone concentrations were determined.
Corticosterone, adrenocorticotropic hormone, 17β-estradiol, and testosterone assays
In order to evaluate endocrine changes in the rat model of CSIS used in the present study, the concentrations of stress-related (ACTH, corticosterone) and sex hormones (17β-estradiol, testosterone) were assayed. Plasma corticosterone concentrations were determined using an enzyme-linked immunoassay (ELISA) kit for rats (Cayman Chemical, Ann Arbor, MI), according to the manufacturer’s instructions. The absorbance at 450 nm was measured using a microplate reader (Multiskan RC, Labsystems, Helsinki, Finland) (the intra-assay coefficient of variation (%CV) was 6% and the sensitivity was 30 pg/ml). Plasma ACTH, and serum 17β-estradiol and testosterone concentrations were determined with a chemiluminescent immunometric assay on an IMMULITE 2000 analyzer (Siemens Healthcare Diagnostics, Eschborn, Germany). For the ACTH, 17β-estradiol, and testosterone assays, the % CV were, respectively 4.8%, 4.5%, and 5% and the sensitivities were 3 pg/ml, 2 pg/ml, and 0.05 ng/ml.
RNA isolation and reverse transcription
The brain samples were sonicated in an ice-cold TRI Reagent® (Sigma, St. Louis, MO) (1 ml/100 mg of tissue). The total RNA was extracted according to Chomczynski (1993). RNA samples were qualitatively and quantitatively evaluated by spectrophotometry (BioPhotometer, Eppendorf, Germany). For genomic DNA elimination, 25 ng/μL total RNA, 1 μL double strand-specific(ds) DNase, and 1 μL 10× dsDNase buffer were incubated for 2 min at 37 °C. Then, for reverse transcription, 1 μL (100 pmol) random hexamer primer, 1 μL 10 mM deoxynucleotide mix (0.5 mM final concentration), and nuclease-free water (Thermo Fisher Scientific, Waltham, MA) were mixed with each RNA sample in a total volume of 15 μL. Samples were held for 10 min at 25 °C, followed by 30 min at 50 °C with 4 μL 5× RT buffer (Thermo Fisher Scientific, Waltham, MA), 1 μL Maxima H Minus Enzyme Mix containing reverse transcriptase, and RNase inhibitor. The reactions were terminated by incubation at 85 °C for 5 min. Samples were stored at −70 °C until their use in PCR.
Quantitative real-time PCR
The qPCR analysis of mRNA expression levels was performed on a LightCycler 96 Real-Time PCR System (Roche Diagnostics, Risch-Rotkreuz, Switzerland) and FastStart Essential DNA Green Master (Roche Diagnostics, Risch-Rotkreuz, Switzerland). The primers were designed using ePrimer3 software and are listed in Table 2. PCR cycling conditions were 10 min at 95 °C, 38 cycles of three-step amplification (10 s at 95 °C, 10 s at 55 °C, and 10 s at 72 °C), and melting curve analysis (10 s at 95 °C, 1 min at 65 °C, and 1 s at 97 °C). An analysis of the melting curve verified the specificity of the PCR products. The relative quantification (RQ) of the studied genes was normalized to β-actin using the RQ method 2–ΔΔCt (Livak & Schmittgen, 2001). The mRNA levels were expressed as a fold difference. All experiments were performed in duplicate for each data point.
Table 2. Description of the primers used in this study for qRT-PCR.
Statistical analysis
Statistical analyses were performed using the GraphPad Prism 5.01 (GraphPad Software, La Jolla, CA). Differences were considered statistically significant for p < .05. Data are presented as mean ± standard error of the mean (SEM). The distribution of each data set was checked for normality using the Shapiro–Wilk test (alpha = .05). The data from behavioral tests were analyzed with a paired t-test or Wilcoxon’s signed-rank test. Rat body weights were analyzed with two-way ANOVA and estrous cycles were analyzed with the Chi-squared-test. Hormone concentrations and gene expression analyses were undertaken using an unpaired t-test with Welch’s correction or the Mann–Whitney test. The correlation analyses were performed using Pearson’s or Spearman’s correlation coefficient.
Results
Behavior analysis
In the OF test, after CSIS, a decrease in the total movement duration was observed (df = 13, t = 2.25, p = .04; Figure 2(a)). The difference in the total distance traveled before and after CSIS did not reach statistical significance (p = .06; Figure 2(b)). CSIS exposure reduced the duration of rearing (df = 7, t = 3.70, p = .007; Figure 2(c)) and grooming (df = 13, t = 3.75, p = .002; Figure 2(d)) activity, as compared to the results in rats before CSIS. The stressed rats did not spend significantly less time in the center of the OF chamber (df = 14, t = 1.77, p = .09; Figure 2(e)). In the EPM test, rats subjected to stress entered the open arms less frequently than they did before CSIS (df = 15, t = 2.20, p = .04; Figure 3(a)), spent less time in the open arms (p = .007; Figure 3(b)), and spent more time in the closed arms (p = .018; Figure 3(c)). Moreover, a decrease in head dip frequency was observed in the rats after CSIS exposure, as compared to the results before CSIS (p = .046; Figure 3(d)).
Published online:
15 September 2017Figure 2. The spontaneous locomotion and exploratory behaviors of female rats in the open field before and after exposure to chronic social instability stress (CSIS): (a, b) movement duration and total distance traveled, (c, d) grooming and rearing duration, (e) percentage of time spent in the center of open field (OF) chamber. Data were collected from Experiment 1 (n = 16). Values are means ± SEM. **p < .01, *p < .05 vs. before CSIS (a–e: Paired t-test).
![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/ists20/2017/ists20.v020.i06/10253890.2017.1376185/20171127/images/medium/ists_a_1376185_f0002_c.jpg"})
Figure 2. The spontaneous locomotion and exploratory behaviors of female rats in the open field before and after exposure to chronic social instability stress (CSIS): (a, b) movement duration and total distance traveled, (c, d) grooming and rearing duration, (e) percentage of time spent in the center of open field (OF) chamber. Data were collected from Experiment 1 (n = 16). Values are means ± SEM. **p < .01, *p < .05 vs. before CSIS (a–e: Paired t-test).
Published online:
15 September 2017Figure 3. Anxiety-related behavior of female rats on the elevated plus maze (EPM) before and after exposure to chronic social instability stress (CSIS): (a) percentage of time spent in the open arms, (b) percentage of entries into the open arms, and (c) time spent in the closed arms, and (d) head dip frequency. Data were collected from Experiment 1 (n = 16). Values are means ± SEM. **p < .01, *p < .05 vs. before CSIS (a: Wilcoxon’s signed-rank test; b–d: Paired t-test).
![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/ists20/2017/ists20.v020.i06/10253890.2017.1376185/20171127/images/medium/ists_a_1376185_f0003_c.jpg"})
Figure 3. Anxiety-related behavior of female rats on the elevated plus maze (EPM) before and after exposure to chronic social instability stress (CSIS): (a) percentage of time spent in the open arms, (b) percentage of entries into the open arms, and (c) time spent in the closed arms, and (d) head dip frequency. Data were collected from Experiment 1 (n = 16). Values are means ± SEM. **p < .01, *p < .05 vs. before CSIS (a: Wilcoxon’s signed-rank test; b–d: Paired t-test).
Body weight gain
Within the 4-week experimental period, the body weight of female rats increased by 23.5%. The two-way ANOVA test revealed significant effects of the time factor (F[7,107] = 86.7, p < .0001), but no significant effects for the stress factor (F[1,107] = 3.92, p = .05) and time × stress interaction (F[7,107] = 0.32, p = .94) (Figure 4(a)).
Published online:
15 September 2017Figure 4. The effect of chronic social instability stress (CSIS) on physiological parameters: (a) body weight gain expressed as a percentage of the initial body weight measured on the first experimental day, (b) percentage of rats having a regular estrous cycle each week (note axis does not begin at zero). Data were collected from Experiment 2 (the control (CON) group n = 7, the CSIS group n = 8). Values are means ± SEM (no significant group differences; a: two-way ANOVA, b: chi-squared test).
![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/ists20/2017/ists20.v020.i06/10253890.2017.1376185/20171127/images/medium/ists_a_1376185_f0004_c.jpg"})
Figure 4. The effect of chronic social instability stress (CSIS) on physiological parameters: (a) body weight gain expressed as a percentage of the initial body weight measured on the first experimental day, (b) percentage of rats having a regular estrous cycle each week (note axis does not begin at zero). Data were collected from Experiment 2 (the control (CON) group n = 7, the CSIS group n = 8). Values are means ± SEM (no significant group differences; a: two-way ANOVA, b: chi-squared test).
Estrous cyclicity
Vaginal smears were collected for 28 consecutive days from the control and CSIS rats. The Chi-squared-test was performed to compare the percentage of rats having a regular estrous cycle each week. During the 4 weeks, there was no significant difference between the control and the stress groups (df = 3, χ2=1.31, p = .72; Figure 4(b)). The estrous cycles of rats were regular and followed a 4 or 5-day pattern. On the last day of Experiment 2, 45.4% of all 15 rats were in the diestrus phase, 27.3% were in proestrus, and 27.3% were in estrus.
Hormone assays
The rats subjected to CSIS showed no significant changes in plasma ACTH (df = 11, t = 1.22, p = .24; Figure 5(a)) or in plasma corticosterone concentrations (df = 11, t = 1.08, p = .30; Figure 5(b)), as compared to the unstressed, control group. However, in the stressed rats, an increase in the plasma ACTH/corticosterone ratio was observed (df = 9, t = 2.30, p = .04, Figure 5(c)), as compared to the control group. After stress exposure, serum estradiol and testosterone concentrations remained unchanged (estradiol: p = .53, testosterone: p = .58; Figure 5(d,e)).
Published online:
15 September 2017Figure 5. The effect of chronic social instability stress (CSIS) on circulating concentrations of hormones: (a) plasma concentrations of adrenocorticotropic hormone (ACTH) and (b) corticosterone, (c) ACTH/corticosterone ratio, (d) serum concentrations of estradiol, and (e) testosterone on the last experimental day. Data were collected from Experiment 2 (the control (CON) group n = 7, the CSIS group n = 8). Values are means ± SEM (a–c: unpaired t-test, d–e: Mann–Whitney’s test).
![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/ists20/2017/ists20.v020.i06/10253890.2017.1376185/20171127/images/medium/ists_a_1376185_f0005_c.jpg"})
Figure 5. The effect of chronic social instability stress (CSIS) on circulating concentrations of hormones: (a) plasma concentrations of adrenocorticotropic hormone (ACTH) and (b) corticosterone, (c) ACTH/corticosterone ratio, (d) serum concentrations of estradiol, and (e) testosterone on the last experimental day. Data were collected from Experiment 2 (the control (CON) group n = 7, the CSIS group n = 8). Values are means ± SEM (a–c: unpaired t-test, d–e: Mann–Whitney’s test).
Correlation of the peripheral concentrations of hormones
In order to evaluate if stress hormones are influenced by sex steroids, or vice versa, appropriate correlation analyses were performed. There was no correlation between plasma ACTH and 17β-estradiol concentrations (p = .38, Pearson r = –0.29), but ACTH concentrations showed a negative correlation with serum testosterone concentrations (p = .018, Spearman r = –0.70). There was also no correlation between the plasma concentrations of corticosterone and serum estradiol or testosterone (corticosterone/estradiol: p = .63, Pearson r = 0.16, corticosterone/testosterone: p = .13, Spearman r = 0.47) (results not detailed).
qRT-PCR analysis
First, the relative expression of POMC was significantly increased in the amygdala of stressed rats (df = 2, t = 7.40, p = .01), as compared to the control group, while in the PFC and hypothalamus POMC expression decreased (PFC: df = 2, t = 4.49, p = .04; hypothalamus: df = 2, t = 10.2, p = .009; Figure 6(a)). Exposure to CSIS decreased the relative expression of CRH-R1 in the hippocampus (df = 2, t = 7.38, p = .01) and increased the expression of this receptor in the PFC (df = 2, t = 5.07, p = .03; Figure 6(b)), as compared to the control females. Then, the influence of stress on the relative expression of receptors important for social behavior was studied. Levels of AVPR1a expression were reduced in the hypothalamus (df = 3, t = 3.24, p = .04), but elevated in the amygdala of stressed rats (df = 3, t = 4.15, p = .02; Figure 6(c)), as compared to the controls, while no significant differences were revealed in the hippocampus and PFC. Similarly, the second type of vasopressin receptor (AVPR1b) was significantly reduced in the hypothalamus of rats subjected to CSIS (df = 3, t = 4.73, p = .01; Figure 6(d)). OXTR expression was increased after chronic stress exposure in the amygdala (df = 3, t = 5.24, p = .01; Figure 6(e)), as compared to the non-stressed rats, while there were no significant differences in the other structures. Finally, ER-β expression was reduced only in the PFC of CSIS females (df = 2, t = 5.04, p = .03; Figure 6(f)), as compared to the control rats.
Published online:
15 September 2017Figure 6. The effect of chronic social instability stress (CSIS) on relative gene expression (RQ) in the amygdala (A), hippocampus (H), prefrontal cortex (PFC), and hypothalamus (HT). (a) RQ for proopiomelanocortin (POMC), (b) RQ for corticotropin-releasing hormone receptor type 1 (CRH-R1), (c, d) RQ for arginine vasopressin receptors (AVPR1a, AVPR1b), (e) RQ for oxytocin receptor (OXTR), and (f) RQ for estrogen receptor beta (ERβ). Note that the scale of the y-axis varies depending on the gene. Data were collected from Experiment 2 (the control (CON) group n = 7; the CSIS group n = 8). Values are means with range. *p<.05 vs. the control group (a–f: unpaired t-test with Welch’s correction).
![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/ists20/2017/ists20.v020.i06/10253890.2017.1376185/20171127/images/medium/ists_a_1376185_f0006_c.jpg"})
Figure 6. The effect of chronic social instability stress (CSIS) on relative gene expression (RQ) in the amygdala (A), hippocampus (H), prefrontal cortex (PFC), and hypothalamus (HT). (a) RQ for proopiomelanocortin (POMC), (b) RQ for corticotropin-releasing hormone receptor type 1 (CRH-R1), (c, d) RQ for arginine vasopressin receptors (AVPR1a, AVPR1b), (e) RQ for oxytocin receptor (OXTR), and (f) RQ for estrogen receptor beta (ERβ). Note that the scale of the y-axis varies depending on the gene. Data were collected from Experiment 2 (the control (CON) group n = 7; the CSIS group n = 8). Values are means with range. *p<.05 vs. the control group (a–f: unpaired t-test with Welch’s correction).
Discussion
Recently, significant progress has been accomplished with regard to understanding the neurobiological link between stress-coping and anxiety as well as social behaviors. Despite this progress, the available treatment options are far from being mechanism-based, including the role of sex differences. This problem is becoming increasingly recognized, as several preclinical studies investigating the impact of social stress on female rodents have recently been conducted (Pittet, Babb, Carini, & Nephew, 2017; Zanier-Gomes et al., 2015). Thus, the necessity to validate animal models for studying the pathophysiology of stress-related disorders in women is quite evident. Here, the CSIS model is characterized based on an unstable social situation. In the present study, hypotheses were tested for the influence of the modified CSIS procedure on behavioral and physiological parameters and the expression of chosen genes important for stress response and social behaviors.
The main findings of the study are that CSIS resulted in a constellation of behavioral changes in adult females without an alteration in physiological parameters (body weight gain, estrous cycle regularity, hormone concentrations). Second, CSIS induced alterations in the central expression of molecules mediating various aspects of the stress response (POMC, CRH-R1, AVPR1a, AVPR1b, OXTR, and ERβ). After CSIS exposure, decreased spontaneous locomotion and a reduced exploration in the OF chamber were observed, indicated by a significant decrease in the total duration of rearing. Grooming behavior was reduced after CSIS. Previously, it has been shown that a reduction in grooming activity might reflect anhedonia (Kalueff, Aldridge, LaPorte, Murphy, & Tuohimaa, 2007; Willner, 2005). Although this observation should be addressed with great caution considering that self-grooming is a complex innate behavior and that there are many symptoms that may be modeled in rodents by an assessment of self-grooming behavior (Kalueff et al., 2016). In the EPM, increased anxiety-related behavior was revealed by the reduced frequency of entries into open arms, time spent in open arms, and reduced head dip frequency after CSIS exposure. These data are in line with the results obtained by others using chronic stress paradigms in males (Ieraci, Mallei, & Popoli, 2016; Sterlemann et al., 2008), but stand in contrast to several studies conducted on females in which chronic social stress did not induce any changes in anxiety and/or depressive-related behaviors (Herzog et al., 2009; Schmidt, Wang, & Meijer, 2011). However, in Experiment 1, there was no separate control group and behavioral tests were conducted on the same rats twice (before and after CSIS), as in previous investigations (Boehmerle, Huehnchen, Peruzzaro, Balkaya, & Endres, 2014; Murphy & Burnham, 2006). This protocol might be acknowledged as a limitation of the study. The impact of the estrous cycle and sex hormones on behavioral responses to stress could be an important factor affecting the obtained results (Marcondes, Miguel, Melo, & Spadari-Bratfisch, 2001). Considering this, the decision was made not to include the measurements of the cycle phase in Experiment 1, as frequent vaginal sampling could disrupt the behavioral tests, which may be considered a limitation of the study. On the last day of Experiment 2, no alteration in body weight gain was noted in the stressed rats as compared to the control group. It is well-established that in males, chronic stress results in a decrease in body weight gain or a loss of body weight (Tamashiro et al., 2007), but female rats are less affected by chronic stress with respect to body weight, based on previously observed effects (McCormick, Robarts, Kopeikina, & Kelsey, 2005; Ter Horst, Wichmann, Gerrits, Westenbroek, & Lin, 2009). The effects of chronic social stress on physiological and behavioral parameters may differ due to rapidly changing ovarian hormone fluctuations (Frye, Petralia, & Rhodes, 2000; Ter Horst et al., 2009). Taking this into consideration, the effect of CSIS exposure on the estrous cycle and on the circulating concentrations of sex steroids in the adult females was evaluated. Chronic stress did not disrupt estrous cycle regularity. The estrous cycles in both studied groups were regular and followed a 4- or 5-day pattern. On the last day of the experiment, serum estradiol and testosterone concentrations were not altered in the stressed females, as compared to the control rats. In females, HPA axis activity is influenced by estradiol (Burgess & Handa, 1992), which may partially explain the fact that mood disorders are diagnosed in women twice as often as in men. However, no significant correlations between serum estradiol concentrations and the studied stress hormones were found. Interestingly, a negative correlation between ACTH and testosterone concentrations was observed. Previous experimental studies showed that estradiol enhances HPA axis activity in female rodents (Burgess & Handa, 1992) whereas testosterone inhibits this activity in male rats (Viau & Meaney, 2004). In the present study, the ACTH and corticosterone concentrations also remained unchanged after CSIS exposure. Based on the finding that the hormone concentrations vary during the estrous cycle, the lack of the effect of stress might be related to the different phases of the estrous cycle among the experimental groups on the last day of the procedure (blood sampling). In female rats, the resting plasma corticosterone concentrations depend on the stage of the estrous cycle, being low at estrus and maximal at proestrus. Plasma corticosterone concentration is positively associated with plasma ACTH concentrations; however, the nature of this association may change across the estrous cycle (Atkinson & Waddell, 1997). A significantly increased ACTH/corticosterone ratio was observed in the stressed rats, which may reflect lower sensitivity of the adrenal cortex to ACTH (Daskalakis, Enthoven, Schoonheere, Ronald de Kloet, & Oitzl, 2014). Lower adrenal sensitivity to ACTH is altered by chronic stress exposure, long-term voluntary exercise, or aging (Bornstein, England, Ehrhart-Bornstein, & Herman, 2008).
Essential to the function of the HPA axis are feedback loops, in which the POMC processing pathways play an important role (Nicolaides et al., 2015). It is well-established that CRH regulates anterior pituitary POMC gene activity and the production/release of ACTH, β-endorphin, and melanocyte stimulating hormone (MSH) peptides. ACTH reaches the adrenal gland and activates the melanocortin receptor 2 (MC2-R), thereby inducing the production and secretion of corticosterone. Moreover, the same steroid acts to terminate the stress response by interacting directly with the central nervous system (CNS) or anterior pituitary receptors to attenuate CRH and POMC peptide production (Slominski, Wortsman, Luger, Paus, & Solomon, 2000). Although POMC peptides are mainly associated with the pituitary gland and neurons of the hypothalamus (Cawley, Li, & Loh, 2016), low levels of POMC mRNA have been detected in other brain regions, including the cerebral cortex, amygdala, and hippocampus (Civelli, Birnberg, & Herbert, 1982). So far, the expression and/or function of POMC in the above-mentioned regions have only been explored in relation to stress-related responses during heroin withdrawal and abstinence (Niikura, Zhou, Ho, & Kreek, 2013; Zhou, Leri, Cummins, & Kreek, 2015). After a 12-h withdrawal, heroin-treated mice showed higher levels of POMC mRNA in the amygdala but lower levels in the hippocampus than the control. Conversely, after 7-d withdrawal, the mice showed fewer POMC-EGFP (enhanced green fluorescent protein)-positive cells and lower POMC mRNA levels in the amygdala than did the control (Niikura et al., 2013). It was also reported that chronic immobilization stress may decrease the level of POMC mRNA in the rat hippocampus (Chen, Tang, & Yang, 2008). Another interesting study showed that chronic ethanol intake reduced a-MSH expression both in the amygdala and hippocampus and modified the mRNA POMC level in adult rats (Lerma-Cabrera et al., 2013). In the present study, female rats exposed to CSIS displayed an elevation in POMC mRNA in the amygdala, and a reduction in prefrontal cortical and hypothalamic POMC expression levels. There were no changes in the POMC mRNA in the hippocampus. These results show that direction of changes in central POMC expression is region-specific, suggesting the interesting but still inadequately explained role of this peptide in the regulation of stress responses.
In addition to the well-known function of CRH in the HPA axis, CRH is expressed in numerous sites within the CNS where it acts as a neurotransmitter mediating the behavioral, autonomic, and metabolic responses to stress (Owens & Nemeroff, 1991). CRH-binding receptor type 1 (CRH-R1) has been shown to mediate the classic neuroendocrine response to stress (Reul & Holsboer, 2002). In rodents, CRH-R1 expression has been found in numerous location within the CNS and the anterior and intermediate lobes of the pituitary (Potter et al., 1994; Van Pett et al., 2000). The current study shows that after CSIS exposure, mRNA expression for CRH-R1 differed depending on the brain region. In the hippocampus, levels of CRH-R1 mRNA were reduced, while in the PFC they were elevated. No changes in amygdalar and hypothalamic CRH-R1 mRNA were observed. Previous studies showed inconsistent results. Pournajafi-Nazarloo et al. (2013) showed no effect of chronic isolation on hypothalamic and hippocampal CRH or CRH-R1 mRNAs in prairie voles. However, it has been revealed that exposure to stress during the adolescent period caused an increase of CRH-R1 expression in the hypothalamus during adulthood, with amygdala and PFC concomitantly causing decreased CRH-R1 expression in the hippocampus (Li et al., 2015; O’Malley, Dinan, & Cryan, 2011). CRH also acts as a neurotransmitter in the amygdala, medial PFC, and hippocampus, mediating the development of anxiety and depression as well as learned helplessness via the activation of CRH-R1 (Spencer, Buller, & Day, 2005; Todorovic, Jahn, Tezval, Hippel, & Spiess, 2005). The results suggest that exposure to CSIS has region-specific consequences for the modulation of gene expression for CRH-R1 in the female brain. The PFC and hippocampus are the stress-sensitive brain structures that are involved in the inhibition of the HPA stress response (Herman, Ostrander, Mueller, & Figueiredo, 2005), and observed alterations in CRH-R1 expression following stress exposure might be a causative factor for the persistent dysregulation of the HPA axis and increased vulnerability to stress in these regions.
Central vasopressin acts on maternal behavior, cognitive functions, and emotion via two receptors subtypes in the brain: AVPR1a and AVPR1b (Bosch & Neumann, 2008). AVPR1a receptors are involved in the development of psychiatric disorders such as anxiety, depression, and post-traumatic stress disorder (Rotondo et al., 2016). AVPR1b receptors are prominently expressed in the corticotroph cells of the anterior pituitary and modulate the release of ACTH. Vasopressin, especially via AVPR1a, and oxytocin play a crucial role in sustaining the activation of the HPA axis during exposure to chronic stress (Aguilera, Subburaju, Young, & Chen, 2008; Neumann, 2008). Oxytocin neurons are activated in response to various stressors (e.g. restraint stress, conditioned fear), which increase the release of central oxytocin (Onaka, 2004; Onaka, Takayanagi, & Yoshida, 2012). The central administration of oxytocin has been shown to attenuate the release of stress-induced corticosterone and exert an anxiolytic effect both in female and male rats (Blume et al., 2008; Windle, Shanks, Lightman, & Ingram, 1997). To date, this anxiolytic effect of oxytocin has been localized within the central amygdala (Neumann, 2008), whose efferents to the hypothalamus and brainstem trigger the autonomic expression of fear (Huber, Veinante, & Stoop, 2005). By contrast, the activation of vasopressin receptors exert anxiogenic effects in the central amygdala (Huber et al., 2005). In the present investigation, it was found that AVPR1a and AVPR1b mRNA levels were significantly decreased in the hypothalamus of the stressed females, while in the amygdala increased AVPR1a and OXTR expression were observed. Previously, the sole related study showed that chronic isolation stress did not affect AVPR1a expression in the hypothalamus of female and male prairie voles (Pournajafi-Nazarloo et al., 2013). Duque-Wilckens et al. (2016) have shown that social stress increased AVPR1a in the nucleus accumbens (NAc), which is known to interact with the amygdala. The authors also have found that in females, the anxiolytic and prosocial effects of AVPR1a were mediated independently by receptors in the NAc shell and the bed nucleus of the stria terminalis (BNST). Moreover, it might be suggested that only one of the two types vasopressin receptor plays a role in certain behaviors within a specific brain area, whereas in other regions, both AVPR1a and AVPR1b may act in a similar manner or even counteract each other. These findings reveal some major points. First, in the adult females, alterations in the amygdalar mRNA expression of the AVP and OXT receptors as well as the mRNA of hypothalamic AVPRs might be important for the stress-related response, including anxiety-related behaviors. Second, the increase in AVPR1a and OXTR expression in the amygdala may be the result of maintaining an endogenous balance between the anxiogenic and anxiolytic actions of vasopressin and oxytocin following CSIS. Finally, the observed increased OXTR expression in response to chronic social stress supports the hypothesis that the central release of oxytocin after stress exposure may attenuate the stress-induced increase in HPA axis reactivity, thereby enabling beneficial social behaviors such as reproduction and feeding (Kudwa, Mcgivern, & Handa, 2014).
Finally, chronic social stress significantly decreased ERβ expression in the PFC, and no changes were observed within the other brain structures. Some reports have shown the predominance of ERβ and not ERα in the adult PFC, which may indicate the importance of these receptors (Westberry & Wilson, 2012). It was reported that ERβ mediates in many non-reproductive actions, for example, the cardio- or neuroprotective effects of estradiol (Kim, Torcaso, Asimes, Chung, & Pak, 2017). The ERβ regulates different aspects of social behavior and anxiety (Cushing et al., 2008). Finally, the PFC is an important region involved in stress-related behaviors, working memory, and moderating correct social behaviors, and displays sex differences in response to stressors (McEwen, Nasca, & Gray, 2016). It may be concluded that the region-specific decline in the prefrontal cortical ERβ expression may be responsible for some of the behavioral disturbances observed in the stressed rats. Moreover, it is interesting that after CSIS exposure, the prefrontal cortical levels of AVPR1b and OXTR mRNA decreased or tended to decrease (AVPR1b: p = .05, OXTR: p = .08). Our results suggest that in the PFC, chronic social stress may modulate mRNAs for the ERβ as well as other studied stress-related molecules (POMC, AVPR1a, AVPR1b, and OXTR) in a similar manner.
Conclusions
In conclusion, the CSIS procedure decreased spontaneous locomotion and exploration and induced anxiety-related behaviors in adult females. However, exposure to CSIS did not affect the physiological parameters (i.e. body weight gain, regularity of estrous cycles and circulating concentrations of stress hormones). One of the limitations of the study is that the protein concentrations of POMC and the studied receptors were not examined. Moreover, the results of mRNA expression should be interpreted with caution. The main finding of this study is that CSIS affected the central levels of mRNAs for molecules involved in the stress response in a region-specific manner. CSIS altered POMC mRNA expression in the amygdala, PFC, and hypothalamus, which suggests an intriguing role of POMC peptide(s) in the stress response. The up-regulation of CRH-R1 mRNA in the PFC, and down-regulation in hippocampal CRH-R1 expression following CSIS exposure, might be a causative factor for the dysregulation of the neuroendocrine response to stress. Moreover, the increase in AVPR1a and OXTR expression in the amygdala may be a result of maintaining an endogenous balance between the anxiogenic and anxiolytic actions of vasopressin and oxytocin following CSIS. Finally, the lower expression of ER-β in the PFC suggests that this receptor is involved in anxiety-related behaviors following CSIS. Knowledge of CSIS-induced disturbances in the central expression of stress-related and socially important factors and their receptors may contribute to an understanding of the mechanisms underlying the detrimental effects of social stress. Further studies, especially immunohistochemical studies and additional behavioral tests, are required in order to explain the nature of the effects of CSIS on the signaling pathways involving the factors studied here.
| Day of experiment | Experimental procedures |
|---|---|
| 1 | Isolation (single rat/cage) |
| 2 | Crowding (four rats/small cage) for 3 hours |
| 3 | Isolation (single rat/cage) |
| 4 | |
| 5 | Crowding (four rats/small cage) for 6 hours |
| 6 | Isolation (single rat/cage) |
| 7 | Crowding (four rats/small cage) for 3 hours |
| 8 | Isolation (single rat/cage) |
| 9 | Crowding (four rats/small cage) for 6 hours |
| 10 | Isolation (single rat/cage) |
| 11 | Crowding (four rats/small cage) for 3 hours |
| 12 | Isolation (single rat/cage) |
| 13 | |
| 14 | Crowding (four rats/small cage) for 6 hours |
| 15 | Isolation (single rat/cage) |
| 16 | Crowding (four rats/small cage) for 3 hours |
| 17 | Isolation (single rat/cage) |
| 18 | Crowding (four rats/small cage) for 6 hours |
| 19 | Isolation (single rat/cage) |
| 20 | |
| 21 | Crowding (four rats/small cage) for 3 hours |
| 22 | Isolation (single rat/cage) |
| 23 | Crowding (four rats/small cage) for 6 hours |
| 24 | Isolation (single rat/cage) |
| 25 | |
| 26 | Crowding (four rats/small cage) for 3 hours |
| 27 | Isolation (single rat/cage) |
| 28 | Crowding (four rats/small cage) for 6 hours |
| Gene | Primers (5′–3′) | Accession number |
|---|---|---|
| POMC | F-CGACGGAGGAGAAAAGAGGTT R-CTGAGGCTCTGTCGCGGAA | NM_139326.2 |
| CRH-R1 | F-CTACGGTGTCCGCTACAACA R-TGTACTTTGCTCTTCTCTTCGTTG | NM_001301812.1 |
| AVPR1a | F-AAGCGCCTACATCCTTTGCT R-TGGAAGGGTTTTCTGAATCGGT | NM_053019.2 |
| AVPR1b | F-ATGGCAGCATAGGAGCCAAC R-AGCTCGGAGGAGCGTTTTAT | NM_017205.3 |
| OXTR | F-TCGTACTGGCCTTCATCGTG R-TGAAGGCAGAAGCTTCCTTGG | NM_012871.2 |
| ERβ | F-AGGTGCTAATGGTGGGACTG R-CCCCTCATCCCTGTCCAGAA | NM_012754.1 |
| βactin | F-CGCGAGTACAACCTTCTTGC R-CGTCATCCATGGCGAACTGG | NM_031144.3 |
Genes, primers sequences and accession numbers are shown: POMC: proopiomelanocortin; CRH-R1: corticotropin-releasing hormone receptor type 1; AVPR1a: arginine vasopressin receptor a; AVPR1b: arginine vasopressin receptor b; OXTR: oxytocin receptor; ERβ: estrogen receptor beta.
Disclosure statement
All authors declare that there are no financial, personal, or other potential conflicts of interest to report.