**1. Introduction**

Arginine vasopressin (AVP) plays an important role in saltwater homeostasis and the regulation of blood pressure as a peripheral hormone [1]. Besides these functions, it also works as a neuropeptide and contributes to the regulation of learning and memory, social behavior and emotionality. Most of the AVP is synthesized in the magnocellular neurons of the hypothalamic paraventricular (PVN) and supraoptic nuclei. It may have an effect on the stress axis by strengthening the function of corticotropin-releasing hormone (CRH) in the PVN [1]. Thus, AVP can increase the CRH-induced adrenocorticotropin (ACTH) levels, then finally lead to glucocorticoid (corticosterone in rodents) secretion from the adrenal gland [1,2]. This forms the hypothalamic–pituitary–adrenocortical (HPA) axis, the fundamental component of adaptation to stress. Disturbances of these mechanisms

may contribute to the development of many diseases among them being anxiety. Many research studies supported the observation that the AVP level is positively correlated with the manifestation of anxiety symptoms in adulthood [3].

The perinatal period is less studied in relation to anxiety-like behavior, most probably because of the low number of available tests. One of the best tools seems to be maternal separation-induced ultrasonic vocalization (MS-USV) [4,5]. Our previous study indicated the contribution of AVP to anxiety and the emission of MS-USV using a genetic model: the natural AVP-deficient Brattleboro strain [6]. We found significantly less anxiety-like behavior and low ACTH levels compared with the wild type, both in adults [7] as well as in pups [6]. However, the contributing receptor subtype remained to be elucidated.

AVP has three vasopressin receptors (VR) in mammals: vasopressin (V) 1aR, V1bR and V2R, with different localizations in the body [8]. V1aRs can be found in the vessels' endothelium, and their most important function is to regulate vasoconstriction. We can also find this VR subtype in the limbic system (lateral septum, amygdala and hippocampal areas), the most important regions in the regulation of anxiety-related behaviors. V1bR signaling seems to also be important in anxiety [8]. Most of the V1bRs are present in the adenohypophysis, regulating the HPA axis. V1bR, as the main VR in the pituitary, was the main target of anxiety research in the 2000s [9]. However, among others, Bayerl et al. investigated the anxiogenic role of the VRs in the PVN of adult rats with the elevated plus maze test and found that V1aR, but not the V1bR antagonist, decreased anxiety-like behavior [10]. Moreover, in 2012, clinical studies by Griebel et al. showed that the nonpeptide V1bR antagonist SSR149415 may not be useful for the treatment of generalized anxiety disorder in adult patients [11]. V2Rs are localized in the kidney, regulating saltwater homeostasis. In the brain, its appearance is restricted to the cerebellum. Thus, its contribution to anxiety is questionable but cannot be excluded [12].

Here, we aimed to establish which VR subtype could be responsible for the observed behavioral (MS-USV) and HPA axis effects of the Brattleboro rat pups. As any stress, even a single saline injection, could influence the anxiety measured by MS-USV, we first confirmed that the effect of genetic AVP mutation was still detectable after the mild stress of a saline injection. Then, the effect of pharmacological antagonism by different VRs (V1aR-antagonist SR49059, V1bR-antagonist SSR149415 and V2R-antagonist SR121463B) was studied in increasing doses (3, 10 or 30 mg/kg) in comparison with vehicle injection in association with stress hormone changes (ACTH and corticosterone) [13]. Finally, a combination of V1aR and V1bR antagonists (10 + 10 mg/kg) was used.

#### **2. Materials and Methods**

#### *2.1. Subjects*

We used 7–8-day-old Brattleboro (*n* = 19) and Wistar (*n* = 149) male and female rat pups. There was no sex effect in either case, therefore we pulled results from two sexes together. The litter sizes of 7–10 pups, and 4–8 litters were included in each experiment, leading to 1–3 pups/treatment group/litter [14]. We chose this age group based on previous data [6]. Brattleboro rats were maintained at the Institute of Experimental Medicine in a colony started from breeder rats from Harlan, Indianapolis, IN, USA. Parental Wistar rats were purchased from Charles River (Budapest, Hungary) and were kept in the local animal facility. After birth, the litters were left undisturbed with their mothers until experimentation on postnatal day 7 or 8. The families were kept under standard laboratory conditions: a 12 h light–dark cycle with the lights on at 7:00 a.m., room temperature (20 ± 2 ◦C), 50–70% humidity and food and tap water ad libitum.The experiments were carried out between 10:00 a.m. and 2:00 p.m.

#### *2.2. Measurements*

#### 2.2.1. Maternal Separation-Induced Ultrasonic Vocalization

We measured the weights of the pups, marked them with waterproof ink and gave them an intraperitoneal (ip) injection. Afterward, they were placed back with the litter and were left undisturbed for 30 min. Then, the rat pups were separated from their mother and littermates and placed in a 2 L empty glass beaker without bedding or heating. The experiment was carried out in an empty, closed, soundproof room. MS-USV was recorded for 10 min by an ultrasonic sensitive frequency division detector (CIEL Electronique, CDB205 R2) connected to a personal computer and used Audacity 2.0.5 free software. The detector was put on a platform 10 cm to the side from the top opening of a glass beaker. The recordings were later analyzed by a rat call counter developed by S. Zseb˝ok [6]. The signals were filtered, and the power spectrum was analyzed, ranging from 30 kHz to 50 kHz. In previous studies, the large portions of MS-USV emitted by the 8-day-old rats were found from 30 kHz to 50 kHz [15,16]. The threshold value was set at a signal amplitude of 0.4 V to exclude background noise. The MS-USV duration and number of calls were studied in each group. The MS-USV duration was the sum of the emitted USV durations in seconds during the 10 min observation period, and the number of calls was the average number of emitted USVs per minute.

#### 2.2.2. Testing Sedative Side Effects

At the end of 10 min of MS-USV detection, possible sedative side effects were evaluated.

For the righting reflex, the rat pups were placed on their back on a smooth, flat surface, and the latency necessary to reach the normal upright position with all four feet on the table was measured. The cut-off time was 15 s.

For the negative geotaxis, the offspring were placed on a 45◦ inclined foam rubber board with their nose pointing down. The animals had 30 s to rotate their body through 180◦ .

Both tests are widely used to assess neurobehavioral development and evaluated as being positive (the pup can conduct it within the given timeframe) or negative [17]. At this age, a control animal should be able to perform well, and a negative outcome could be judged as a sedative effect of the treatment.

#### 2.2.3. Hormone Measurements

Right after the sedative tests, the rats were decapitated, and trunk blood was collected into ice-cold Eppendorf tubes and centrifuged at 3000 rpm for 30 min at −4 ◦C. The serum was stored at −20 ◦C until hormone measurements were conducted. From the serum samples, the ACTH and corticosterone concentrations were measured by a specific radioimmunoassay (RIA) without previous extraction. Both antibodies were developed in our institute as described elsewhere [18–20]. The detection limits were 4 fmol/mL for ACTH and 2.7 pmol/mL for corticosterone. The intraassay coefficients of variation were 4.7% for ACTH and 12.3% for corticosterone. From the Brattleboro rats (Experiment 1), we also collected the hypophysis of the pups to determine the AVP content. Pituitary samples were stored in 100 µL 0.1 N HCl at −20 ◦C and then homogenized by ultrasound and centrifuged. Then, the AVP content was measured from the hundredfold diluted supernatant using a specific RIA. The rabbit antibodies were donated by Dr. M. Vecsernyés (University of Debrecen, Debrecen, Hungary). The limit of detection was 1 pg AVP/assay tube. The intraassay coefficient of variation was 10.7%. All the samples from a particular experiment were assayed in the same RIA.

#### *2.3. Experiments*

#### 2.3.1. Experiment 1: Genetic AVP Deficiency

Brattleboro rats born from heterozygous (di/+) mothers and homozygous diabetes insipidus (di/di) fathers were given ip injections with physiological saline 30 min before the test and placed back with their mothers. MS-USV was measured for 10 min, and at termination, trunk blood was collected for ACTH and corticosterone measurement, and hypophysis was collected for determination of the genotype (di/+ or di/di disposition based upon the AVP content) [17].

### 2.3.2. Experiment 2: Pharmacological AVP-Effect Deficiency

A (2A) V1aR antagonist (SR49059), (2B) V1bR antagonist (SSR149415) or (2C) V2R antagonist (SR121463B) was suspended in 0.4% Tween 80 (1µL/g volume for every animal), then delivered intraperitoneally 30 min before MS-USV recording in three different doses: 3, 10 or 30 mg/kg (a generous gift from the Sanofi-Synthélabo company). In a further experimental series, a (2D) 10 mg/kg V1aR antagonist was mixed with a 10 mg/kg V1bR antagonist. The control treatment was the solvent in a 1 µL/g volume washed with 15 µL saline. MS-USV was measured for 10 min, and at the end, the righting reflex and negative geotaxis were evaluated within 2 min. At the termination of the experiments, trunk blood was collected for ACTH and corticosterone determination.

The specificity of these drugs was strongly confirmed by previous studies, and each of them effectively influenced their main target symptom. More precisely, SR49059 showed high affinity to V1aRs in the rat liver (Ki/inhibition constant = 1.6 ± 0.2 nmol/L) and human platelets, adrenals and myometrium (Ki ranging from 1.1 to 6.3 nmol/L). In vivo, SR49059 inhibited the pressor response to exogenous AVP in rats with a long duration of action [21].

SRR149415 had high affinity to V1bRs (hypophysis, human: Ki = 4.2 ± 1.1 nmol/L; rat: Ki = 3.7 ± 1.3 nmol/L). Its in vivo activity could be characterized as anxiolytic-like and stress relieving [13].

The human V2R bonds SR121463 with high affinity (Ki = 0.54 ± 0.09 nmol/L) [22]. SR121463 normalized serum Na<sup>+</sup> levels, abolished hyponatremia and restored normal urine excretion, urine osmolality and renal function in a rat model of cirrhosis [23].

#### *2.4. Statistical Analysis*

Data were expressed as mean ± SEM and analyzed using the STATISTICA 13.0 software package (StatSoft, Inc., Tulsa, OK, USA) by analysis of variance (ANOVA) using one-way ANOVA (factor = treatment). In case of the Levene's assumption being significant, the data were transformed logarithmically, and the analysis was conducted on the transformed data. Post hoc comparison was made by the Fisher's Least Significant Difference method, and the results were presented on the figures. Correlations were calculated by the Pearson method. The level of statistical significance was taken as *p* < 0.05.

#### **3. Results**

#### *3.1. Genetic AVP Deficiency*

The AVP-deficient Brattleboro rats (di/di) were less anxious, based upon their emitted MS-USV number of calls (F(1,17) = 7.600; *p* = 0.014; Figure 1a) and duration (F(1,17) = 7.600; mboxemphp = 0.014; Figure 1b) compared with their heterozygous littermates. In connection, their ACTH levels were significantly lower (F(1,17) = 15.380; *p*= 0.001; Figure 1c) without significant alterations in corticosterone levels (F(1,17) = 1.915; *p* = 0.184; Figure 1d).

The righting reflex (F(1,17) = 1.850; *p* = 0.193) and negative geotaxis (F(1,17) = 0.042; *p* = 0.840) values were comparable in the two genotypes (data not shown).

#### *3.2. Pharmacological AVP Deficiency*

#### 3.2.1. V1aR Antagonist

The V1aR antagonist treatment decreased MS-USV only with the 30 mg/kg dose (number of calls: F(3,42) = 1.788; *p* = 0.164; Fisher post hoc control vs. 30 mg/kg: *p* = 0.026; Figure 2a; duration: F(3,42) = 1.551; *p* = 0.215; Fisher post hoc control vs. 30 mg/kg: *p* = 0.040; Figure 2b). The ACTH levels showed no alterations (F(3,42) = 1.219; *p* = 0.315; Figure 2c), while the corticosterone levels were significantly higher in the group with the 30 mg/kg antagonist treatment (F(3,42) = 1.857; *p* = 0.152; Fisher post hoc control vs. 30 mg/kg: *p* = 0.030; Figure 2d).

There was no difference between the groups in the latency of the righting reflex as well as the negative geotaxis (Table 1).

*Brain Sci.* **2021**, *11*, x FOR PEER REVIEW 5 of 14

**Figure 1.** Brattleboro rats. The 7–8-day-old pups emitted reduced ultrasonic vocalization to maternal separation, measured for 10 min both in terms of (**a**) the number of calls (count/min) and (**b**) the duration (s) 30 min after a single intraperitoneal saline injection. Moreover, they showed reduced (**c**) adrenocorticotropin (ACTH, fmol/mL) elevation at the end of separation without significant changes in (**d**) corticosterone (pmol/mL) levels. di/+ = heterozygous; di/di = homozygous diabetes insipidus pups, without functional vasopressin; *n* = 8–11. \* *p* < 0.05. \*\* *p* < 0.01 vs. di/+ control. The righting reflex (F(1,17) = 1.850; *p* = 0.193) and negative geotaxis (F(1,17) = 0.042; *p* = **Figure 1.** Brattleboro rats. The 7–8-day-old pups emitted reduced ultrasonic vocalization to maternal separation, measured for 10 min both in terms of (**a**) the number of calls (count/min) and (**b**) the duration (s) 30 min after a single intraperitoneal saline injection. Moreover, they showed reduced (**c**) adrenocorticotropin (ACTH, fmol/mL) elevation at the end of separation without significant changes in (**d**) corticosterone (pmol/mL) levels. di/+ = heterozygous; di/di = homozygous diabetes insipidus pups, without functional vasopressin; *n* = 8–11. \* *p* < 0.05. \*\* *p* < 0.01 vs. di/+ control. *Brain Sci.* **2021**, *11*, x FOR PEER REVIEW 6 of 14
