Psychopharmacology

, Volume 218, Issue 1, pp 281–292

Differential sensitivity to amphetamine’s effect on open field behavior of psychosocially stressed male rats

Authors

    • Center of Alcohol StudiesRurgers University
  • April Sweeny
    • Center of Alcohol StudiesRurgers University
  • Patricia Buckendahl
    • Center of Alcohol StudiesRurgers University
Original Investigation

DOI: 10.1007/s00213-011-2339-8

Cite this article as:
Pohorecky, L.A., Sweeny, A. & Buckendahl, P. Psychopharmacology (2011) 218: 281. doi:10.1007/s00213-011-2339-8

Abstract

Rationale

Studies of socially housed rodents have provided significant information regarding the mechanisms of stress and of stress-related disorders.

Objective

Since psychosocial stress is known to alter the functional activity of dopaminergic system, we employed amphetamine (AMP) to evaluate the involvement dopamine in mediating the behavioral consequences of psychosocial stress.

Methods

Male rats housed two per cage were designated as dominant (DOM) or subdominant (Sdom) based on initial evaluations of agonistic behaviors and body weight changes. Diad-housed rats and a group of single-housed (SiH) rats were tested in an open field after injections of saline or amphetamine (0.9 or 2.7 mg/kg IP) prior to and again while diad-housing.

Results

Compared to future DOM rats, saline-injected future Sdom rats entered the open field center less frequently, spent less time in rearing behavior and groomed less. At the pre-diad test AMP treatment elevated locomotor activity of all rats, while stimulation of center entries was more marked in future DOM rats. At the diad test, AMP’s locomotor stimulant effect was evident in all experimental groups with DOM rats showing higher effects compared to Sdom and SiH rats. Amphetamine’s stimulation of center entries in DOM rats was similar to the pre-diad test, but it was diminished in Sdom rats, while stimulation of rearing behavior was most evident in diad-housed rats.

Conclusion

The dopaminergic system modulates the psychosocial stress-induced differences in explorative and emotional behaviors. Furthermore, behavioral traits like frequency of grooming behavior and of center entries were predictive of future hierarchical status.

Keywords

Rank statusOpen field behaviorlocomotor behaviorGrooming behaviorRearing behavior

Introduction

It is well recognized that psychosocial stress strongly influences behavior and physiology of both humans and experimental animals (Björkqvist 2001; Miczek et al. 2008; Pohorecky 1990; 1991; Uhart and Wand 2009). The disruptive role of psychosocial stress is attested by findings of significant alterations in stress-sensitive physiological parameters even in the absence of serious aggression and wounding in group-housed primates (Gust et al. 1991). Research based on different experimental animal models has implicated a variety of neurobiological systems in mediating these effects of psychosocial stressors. One system that is altered by psychosocial stress is the dopamine (DA) neuronal system, though its role in the behavioral consequences of psychosocial stress has not been elucidated.

Considerable evidence has linked the DArgic neuronal system to the response to stressors (for a review, see Marinelli et al. 2006). The mesocorticolimibic DArgic system is believed to mediate various aspects associated with psychosocial stress and has been suggested as the neural substrate for behavioral coping with stressors (Coco and Weiss 2005). Significant evidence has also accrued on the effects of psychosocial stress on the DArgic neuronal system (Lucas et al. 2004). For example, after a resident intruder test, the resident rats have been found to have higher extraneuronal levels DA compared to the intruder rats (Ferrari et al. 2003), and higher levels of DA D2-receptor binding in brain were reported in dominant (DOM) cynomolgus macaques compared to the subordinate animals (Morgan et al. 2002). In rodents, defeat resulted in DArgic hyperactivity and an increase in the phasic DA signaling in the mesolimbic pathway (Anstrom et al. 2009), and repeatedly defeated tree shrews displayed a decline in DA transporter binding in specific brain areas (Isovich et al. 2000). Additionally, psychosocial stress-induced cortisol levels were found to be positively associated with AMP-induced DA release (Wand et al. 2007).

In view of evidence that the functional activity of the DArgic neuronal system is altered by psychosocial stress, we employed amphetamine (AMP) to further evaluate the role of this system in the behavioral consequences of psychosocial stress. Based on extensive literature that implicates DA in the motor-stimulant effects of AMP (e.g., Kalivas and Stewart 1991), the aim of the present study was to evaluate the hypothesis that rats with different levels of psychosocial stress would differ in their behavioral responses to AMP. As a model of psychosocial stress, we employed a diad-housing variant of the triad-housing model developed in our laboratory. This model allows for the rapid development of a stable and robust social hierarchy among co-housed male rats (Blakley and Pohorecky 2006; Pohorecky et al. 1999, 2004a). We evaluated AMP’s behavioral effects on rats using the widely employed open field test. Additionally, we also evaluated whether there are potential behavioral predictors of future rank status.

Methods

Subjects and housing

The subjects were male adult Long–Evans rats weighing approximately 450 g at the start of the experiment (Harlan, Indianapolis, IN, USA). Purina chow and water were available ad libitum throughout the study. The animal room was on a reverse light/dark cycle (12 h each, lights off at 12:30 p.m.), and its ambient humidity and temperature (21°C ± 1°C) were strictly controlled. To adapt to these environmental conditions, rats were initially individually housed in hanging wire-mesh stainless steel cages for 14 days. Rats were assigned to group housing based on body weight; that is, the body weight of the rats to be co-housed differed by <5%. Rats were weighed prior to diad-housing and then weekly for the duration of the study. The housing cages were made of Plexiglas and a wire mesh floor. One of the cage walls had either two (single cages) or four (diad cages) 1-cm opening for drinking spouts. The cages for single-housed (SiH) rats were square (25 × 25 × 30 cm), and those for the diad-housed rats were rectangular (26 cm wide × 82 cm long × 30 cm high). The diad cages had a removable Plexiglas cage divider that partitioned the cage into two equal size compartments. The bottom of the dividers consisted of a 6-cm high wire mesh screen that allowed rats to maintain sensory contact even when separated. These dividers were removed daily for a 1-h period to allow the rats to interact and reinforce their social hierarchy. Our animal facility is certified by AAALAC, and the experimental protocols were approved by the Rutgers University review Committee for the use of Animal Subjects, and all principles of laboratory care were strictly adhered to.

Agonistic behavior rating

Agonistic behaviors were first assessed at the time diads were formed. Subjects were placed into a novel diad cage with the dividers in place for a 5-min adaptation period. The dividers were then removed, and social interactions were recorded over the next 10 min. Agonistic behaviors were scored using a modified and expanded version (Pohorecky et al. 1999, 2004a, b) of the method originally described by Peterson and Pohorecky (1989). Twenty-three different behaviors (described in greater detail previously by Pohorecky et al. 2004a) were scored and subsequently grouped for analysis into four major categories: self-centered (rearing, self-grooming, and genital grooming), affiliative (approach, sniff body, sniff genitals, groom other, and mount other), defensive (defensive upright, defensive back kick, immobility, vocalization, and flight/attempt to jump out of the cage), and aggressive (piloerection, aggressive push-under, pounce on, nip other, cage mark, offensive block or pacing, offensive back kick, lateral threat, on top, roll–tumble interaction). Assignment of social status to the members of a diad was based on the combined use of behavioral scores, detection of ultrasonic and audible vocalizations, and body weight changes noted 24 h after diad formation (Pohorecky et al. 2004a, b, 2006). The status of the subdominant (Sdom) rat was further confirmed by identifying the rat that emitted 22 kHz ultrasonic calls using a Mini Bat Detector (QMC Instruments Inc., London, UK). In cases where rank status was difficult to establish because of a lack of significant agonistic interactions, the diad was either discarded, or a given rat was exchanged so that rank differences were noted. Testing was continued until 13 stable diad pairs were established. At intervals during the study, agonistic interactions were verified to assess the stability of rank assignments.

Open field testing

The open field test was carried out using a modified Plexiglas arena (100 × 100 cm) enclosed by 50-cm high walls (Pohorecky et al. 1999). At the pre-diad test, the rats were initially habituated to the apparatus. During the first 2 days of habituation, animals were placed in the open field for 5 min, and the time was then increased to 10 min; all subsequent tests were also 10-minute long. Assessment of open field behaviors was carried out as previously described (Pohorecky et al. 1999). To minimize differences in circadian sensitivity, all tests were conducted during the rat’s active phase (at 1:30 to 5:00 p.m.) in an adjoining testing room illuminated with a red 40-W bulb.

Corticosterone radioimmunoassay

Blood samples were obtained by tail snip on day 139 and from the trunk after decapitation on day 213, at approximately 15 min after the daily social interaction. The samples were centrifuged, and the plasma was frozen at −70°C until analysis. Plasma CORT was quantified using commercially available kits (no. 07-120103, ICN Biomedical, Orangeburg, NY, USA). Aliquots of plasma (10 μl) were assayed in duplicate in a single assay, with an intra-assay coefficient of variability of 6%.

Experimental treatment

Saline and AMP injections (0.9 and 2.7 mg/kg) were given IP 15 min prior to the test. Each subject received every treatment dose (0, 0.9, and 2.7 mg/kg of AMP) at approximately weekly intervals, once at the pre-diad test and then again at the diad test.

Experimental design

All subjects underwent an extensive period of acclimatized to their novel housing environment and to human handling. Table 1 outlines the chronology of the study.
Table 1

Chronology of the study

Test days

Housing

Procedures and tests

1

Single

Start acclimatization to housing environment and human handling

31

 

Start open field testing (habituation, saline, 0.9, and 2.7 mg/kg amphetamine)

89

Diad

Body weight; diad housing and agonistic behavior assessment

90

 

24-h body weight assessment

139

 

Tail blood sampling

140

 

Start open field testing (habituation, saline, 0.9, and 2.7 mg/kg amphetamine)

213

 

Trunk blood sampling

Statistical data analysis

The data were analyzed using StatView version 5. An ANOVA analysis was employed for the analysis of interactive behaviors and body weights data, with housing/rank status (DOM, Sdom, and single) as the main factors. The effect of the saline injection at the baseline test was analyzed using repeated measure ANOVA, with rank/housing status as the dependent variables and pre-diad and diad tests as the repeated factor. To determine the effectiveness of AMP, the behavioral data were corrected for potential differences in baseline sensitivity to stress of the saline injection. For this purpose, the behavioral responses after drug treatment were expressed as a percent change from the corresponding saline group: [(response after an AMP dose/response after saline injection) × 100]. The initial overall two-way repeated measures ANOVA used the pre-diad and diad tests as the repeated factor and rank/housing status (DOM, Sdom, and single) and treatment dose (saline, 0.9 and 2.7 mg/kg of AMP) as the main factors. This test was then followed by separate ANOVA analysis of the pre-diad and the diad tests, as well as separate single ANOVA analysis of each experimental group. When appropriate, between groups statistical significance was assessed using the post-hoc Bonferroni’s test, with significance levels set at P ≤ 0.0167. All the data are presented as the means and standard errors of the means for 13 rats/rank in and 11 SiH rats.

Results

Behavioral ranking and body weight changes

When two naive male rats were confronted in a novel cage, they initially tended to ignore each other while they explored their novel environment. Subsequently, they engaged in mutual investigation and displayed the behaviors summarized in Fig. 1. The data analysis disclosed that a subject’s hierarchical status was associated with a specific set of behaviors. Namely, self-centered behaviors (F1,  24 = 6.981, P = 0.0143), defensive behaviors (F1,  24 = 7.708, P = 0.0105) and offensive behaviors (F1,    24 = 21.881, P < 0.0001) were related to hierarchical status, but affiliative behaviors were not. The DOM rats engaged in more self-centered behaviors compared to the Sdom rats (P = 0.0143) and displayed offensive behaviors (P < 0.0001), while the Sdom rats displayed more defensive behaviors (P = 0.0105). Surprisingly, and in contrast to our previous experience with triad-housed rats, the DOM rats in only four out of the 13 diads displayed some marking behavior (included in the offensive behavioral category).
https://static-content.springer.com/image/art%3A10.1007%2Fs00213-011-2339-8/MediaObjects/213_2011_2339_Fig1_HTML.gif
Fig. 1

Behaviors exhibited by rats during their initial interaction on diad-housing. Twenty-three different behaviors were scored during a 10-min test and were subsequently grouped for analysis into four major categories, as described in “Methods.” The data are presented as the mean ± SEM for 13 rats per diad-housed group. Asterisk designates statistically significant differences from the dominant rat

Compared to the SiH rats, both diad-housed rats lost some body weight 24 h after diad formation (F2,  34 = 4.66, P = 0.0162). The body weight loss of the DOM and Sdom rats was similar −3.13 ± 0.87% and −3.66 ± 0.69%, compared to −0.69 ± 0.40% for SiH rats, statistically significant decline only for the comparison of the Sdom and SiH rats (P = 0.0065).

Open field behavior

Amphetamine’s effect on open field behaviors was tested on days 31–82 prior to diad-housing and on days 140–191 during the diad–housing period. Amphetamine’s behavioral effect was evaluated statistically after correcting for differences in baseline activity produced by the injection of saline. Namely, the data presented in the figures represents the effect of AMP expressed as the percent change from the corresponding baseline saline-injection test. Therefore, values above 100% indicate AMP-induced behavior activation beyond that at the saline baseline test, while values below the 100% level indicate AMP-induced inhibition of behavior.

Peripheral locomotor frequency

The frequency of peripheral locomotor behavior refers to ambulation restricted to the peripheral wall-bounded squares of the open field arena. In saline-treated rats, peripheral locomotor activity at either the pre-diad or diad tests was not related to the subjects’ rank/housing status (Table 2).
Table 2

Open field behaviors of saline-injected diad-housed and single-housed rats

 

Dominant

Subdominant

Single-housed

Locomotor frequency

   

 Pre-diad

135.23 ± 7.01

117.92 ± 7.91

144.00 ± 10.57

 Diad

129.15 ± 7.12

127.38 ± 6.84

142.62 ± 14.20

Center frequency

   

 Pre-diad

11.28 ± 1.14

8.15 ± 1.00D,Si

14.83 ± 0.81

 Diad

12.79 ± 1.62

11.27 ± 1.09*

13.27 ± 2.16

Center duration

   

 Pre-diad

42.17 ± 1.73

41.06 ± 2.04

44.02 ± 2.42

 Diad

27.70 ± 2.09*

70.91 ± 3.32D*

56.68 ± 3.12D,SD*

Rearing frequency

   

 Pre-diad

55.17 ± 5.23

49.17 ± 4.18

56.27 ± 3.25

 Diad

40.57 ± 5.33*

41.83 ± 3.34

42.48 ± 3.33*

Rear duration

   

 Pre-diad

128.09 ± 6.57

107.01 ± 6.05D

113.38 ± 4.32

 Diad

72.05 ± 8.97*

69.84 ± 6.65*

80.27 ± 11.16*

Grooming frequency

   

 Pre-diad

3.85 ± 0.70

1.67 ± 0.47D

1.73 ± 0.50D

 Diad

4.46 ± 0.95

3.62 ± 0.49*

1.73 ± 0.45D

Grooming duration

   

 Pre-diad

21.35 ± 3.56

10.62 ± 2.47D

9.03 ± 1.46D

 Diad

24.92 ± 2.72

29.21 ± 3.25*

18.25 ± 3.23SD*

Open field behaviors of diad and single-housed rats were assessed as described in “Methods”. The data are presented as the mean ± SEM for 13 rats per diad-housed group and n = 11 for the single-housed rats. D, SD, and Si designate statistically significant differences from dominant, subdominant, and single-housed rats, respectively. Asterisks designate statistically significant differences from the pre-diad group

Treatment with AMP clearly enhanced peripheral locomotor activity at both test times (Fig. 2). An overall repeated measures ANOVA of both tests indicated a main effect of test time (F1, 68 = 25.059, P < 0.0001), although the effects of rank/housing status and of drug dose were not statistically significant. While the interaction of treatment dose and test time was significant (F1, 68 = 13.687, P = 0.0004), the interaction of rank/housing status and test time just missed significance (P = 0.0657). Altogether, the analysis indicates that the increase in peripheral locomotor activity was related to the drug dose, was dependent on the test time, and also tended to vary with the subject’s rank/housing status. To further assess the effect of AMP at each of the two test times, we evaluated each test time separately using a repeated measures ANOVA. There was a main effect of AMP treatment at both tests (F1,  34 = 24.684, P < 0.0001 and F1,  34 = 5.765, P = 0.0220, at the pre-diad and the diad tests, respectively). However, at the pre-diad test, neither the main effect of rank/housing status nor the interaction of rank/housing with drug treatment reached statistical significance. In addition, peripheral locomotor activity of the future DOM and Sdom rats was higher after the 2.7 mg/kg dose of AMP compared to the 0.9 mg/kg dose (P = 0.0019 and P = 0.0064, respectively). On the other hand, at the diad test, there was a main effect of drug dose (F1,  34 = 35.087, P < 0.0001) and a significant interaction effect of drug dose and rank/housing status (F2,  34 = 4.390, P = 0.0200). Also in contrast to the pre-diad test, peripheral locomotor activity of the DOM rats after the 2.7 mg/kg dose of AMP was lower compared to the 0.9 mg/kg dose (P = 0.0158). Although the 0.9 mg/kg AMP-treated DOM rats ambulated more frequently compared to the Sdom or SiH rats, these effects just missed statistical significance. In addition, the low dose of AMP stimulated peripheral locomotor activity more effectively at the diad test compared to the pre-diad test (P < 0.0001 and P = 0.0031 for the DOM and Sdom rats, respectively). Taken together, the data indicated that the diad-housed rats displayed enhanced sensitivity to AMP compared to the pre-diad test, while peripheral locomotor activity of the SiH rats was similar at both test times.
https://static-content.springer.com/image/art%3A10.1007%2Fs00213-011-2339-8/MediaObjects/213_2011_2339_Fig2_HTML.gif
Fig. 2

Effect of amphetamine on peripheral locomotor frequency of diad- and single-housed rats. A saline or an amphetamine injection (0.9 or 2.7 mg/kg IP) was administered 30-min before the 10-min open field test. The data (mean ± SEM) presents the percent change in locomotor frequency after each drug treatment compared to the saline treatment at both test times (n = 13 for the two diad-housed rats and n = 11 for the single-housed rats). Asterisk designates statistically significant differences of the 2.7 mg/kg dose compared to the 0.9 mg/kg dose at each of the two tests. Caret designates statistically significant differences after the 0.9 mg/kg treatment dose for the pre-diad test compared to the diad test

Center frequency

A repeated ANOVA analysis disclosed a statistically significant main effect of rank/housing on the frequency of center entries of saline-injected rats at the pre-diad test (F2,  34 = 4.179, P = 0.0238) (Table 2). Specifically, the saline-injected future Sdom rats entered the open field center less frequently than did the future DOM or the SiH rats (P = 0.0243 and P < 0.0001, respectively), but they did not differ at the diad test. Furthermore, the frequency of entries to the central arena was similar at the two test times for both the DOM and the SiH rats, while the Sdom rats made more frequent entries at the diad test compared to pre-diad test (P = 0.0217).

Similarly to its effect on peripheral locomotor activity, treatment with AMP also enhanced entries to the central arena of the open field, but the drug’s effect appeared to be overall greater (i.e., the percent change was higher). An overall repeated measure ANOVA indicated a main effect of rank/housing status on center entries (F2,  68 = 16.166, P < 0.0001) and a significant interaction of rank/housing and test time (F2,  68 = 7.095, P = 0.0016); however, the main effect of drug dose missed statistical significance (P = 0.0871), while the main effect of test time was not significant (Fig. 3a). These analyses indicate that the overall drug effect was similar at both tests, and it was dependent on the subject’s rank/housing status. Overall, the SiH rats made fewer center entries compared to the DOM or Sdom rats irrespective of test time (P < 0.0001 and P = 0.0007, respectively). Separate repeated ANOVA analyses indicated that rank/housing status had a significant main effect on center frequency at both the test times (F2,  34 = 6.149, P = 0.0053 and F2,  34 = 10.843, P = 0.0002, pre-diad and diad tests, respectively). At the pre-diad test, the SiH rats made fewer entries to the central arena compared to both diad-housed rats (P = 0.0135 and P = 0.0101, for the DOM and Sdom rats, respectively). The main effect of drug dose and interaction of rank/housing status with AMP treatment were not statistically significant. By contrast, at the diad test, the AMP dose had a significant effect on center frequency (F1,  34 = 5.905, P = 0.0205). The DOM rats made more frequent center entries than either the Sdom or the SiH rats after both the 0.9 mg/kg (P = 0.0005 and P = 0.0001, respectively) and after the 2.7 mg/kg AMP doses (P = 0.0026 and P = 0.0030, respectively). Lastly, the Sdom rats entered the arena’s center more frequently after the 0.9 mg/kg dose compared to the 2.7 mg/kg dose (P = 0.0126). Hence, at the pre-diad test, differences in center entries were primarily dependent on the future rank/housing status of the subjects, while at the diad test, for at least the Sdom rats, differences in center entries were also dependent on the treatment dose.
https://static-content.springer.com/image/art%3A10.1007%2Fs00213-011-2339-8/MediaObjects/213_2011_2339_Fig3_HTML.gif
Fig. 3

Effect of amphetamine on the frequency and duration of entries to the open field center by diad- and single-housed rats. A saline or an amphetamine injection (0.9 or 2.7 mg/kg IP) was administered 30-min before the 10-min test. The data (mean ± SEM) represents the percent change in the frequency (a) and on duration (b) of center entries after each drug treatment compared to saline treatment at both test times (n = 13 for the two diad-housed rats and n = 11 for the single-housed rats). D, SD, and Si designate statistically significant differences from the dominant, subdominant and singly housed groups, respectively. Asterisk designates statistically significant differences for the 0.9 mg/kg group compared to the 2.7 mg/kg treatment group at both tests. Caret designates statistically significant differences for the pre-diad test compared to the diad test at each dose level

Center duration

While center duration of all the saline-injected subjects was similar at the pre-diad test, at the diad test, a subject’s rank/housing status had a significant main effect on the time spent in the open field center (F2,  34 = 64.475, P < 0.0001). The DOM rats spent less time compared to the Sdom and SiH rats (P < 0.0001, for both comparisons), while the Sdom rats spent less time in the center than the SiH rats (P = 0.0016). Additionally, compared to the pre-diad test, at the diad test, the DOM rats spent less time, while the Sdom and SiH rats spent more time, in the open field center (P = 0.0005; P < 0.0001, and P = 0.0114, respectively).

Consistent with its action on center frequency, treatment with AMP also enhanced the time rats spent in the open field center. Although its effect at the pre-diad test was relatively small, particularly in the future diad-housed rats, at the diad test center behavior of the DOM rats was particularly sensitive to AMP (Fig. 3b). The overall repeated measures ANOVA indicated that the time subjects spent in the center of the arena was significantly influenced by the main effect of rank/housing status (F2,  68 = 24.697, P < 0.0001) and of test time (F1,  68 = 90.160, P < 0.0001). Furthermore, there were several significant interaction effects, namely, of rank/housing status with treatment dose (F2,  68 = 11.078, P < 0.0001), of rank/housing status with test time (F2,  68 = 62.008, P < 0.0001), of treatment dose with test time (F1,  68 = 69.785, P < 0.0001), and lastly, of rank/housing status with drug dose and with test time (F2,  68 = 14.827, P < 0.0001). These analyses indicate that this behavioral measure was particularly sensitive to the experimental variables, since they all contributed to the subject’s propensity to remain in the center of the open field. Separate analysis of AMP’s effect at each test time indicate that at both tests rank/housing status had a significant main effect on the time subjects spent in the center of the open field (F2,  34 = 18.712, P < 0.0001 and F2,  34 = 33.419, P < 0.0001 for the pre-diad and diad tests, respectively). The treatment dose also had a significant effect on center duration at both test periods (F1,  68 = 136.637, P < 0.0001 and F1,  34 = 30.946, P < 0.0001 for pre-diad and diad, respectively). Although the interaction of rank/housing status with drug treatment at the diad test was significant (F2,  34 = 28.813, P < 0.0001), it missed statistical significance at the pre-diad test (P = 0.0775). Focusing first at the pre-diad test, all the rats spent more time in the central arena after the higher dose of AMP compared to the lower dose (P < 0.0001 for both diad rats and P = 0.0007 for the SiH rats). Furthermore, the DOM rats treated with the lower AMP dose spent less time in the center compared to similarly treated Sdom and SiH rats (P < 0.0001 for both groups). After the higher drug dose, only the SiH rats spent significantly more time in the center compared to the DOM rats (P = 0.0226). By contrast, at the diad test, both diad-housed rats spent less time in the center after the larger AMP dose compared to the smaller dose (P = 0.0107 and P < 0.0001 for the dominant and Sdom rats), while the SiH rats spent more time in the center after the higher dose (P = 0.0011). Focusing on the rank/housing differences at each dose level, we find that after the 0.9 mg/kg AMP dose, the DOM rats spent significantly more time in the center compared to similarly treated Sdom and SiH rats (P < 0.0001 for both comparisons), while after the higher drug dose, the Sdom rats spent significantly less time in the center than did either the DOM or the SiH rats (P < 0.0001 for both comparisons). Lastly, comparing the effect AMP at the two test times, we note that the lower AMP dose significantly prolonged the time the three groups spent in the central arena (P < 0.0001, P = 0.0070, and P = 0.0001, for the DOM, Sdom and SiH rats, respectively). On the other hand, after the higher AMP dose, the DOM rats spent more time (P = 0.0107), but the Sdom rats spent less time (P < 0.0001) in the central arena at the diad compared to the pre-diad test time. Altogether, at the pre-diad test, all rats displayed a dose-dependent increase in the time spent in the open field center, and unlike the SiH rats, the diad-housed rats displayed a dose-related decline in the time spent in the arena’s center.

Rearing frequency

Neither rank/housing status nor treatment with AMP had a significant effect on the frequency of rearing behavior at both test times (Table 2). While all three experimental groups tended to rear less frequently at the diad compared to the pre-diad test, these differences attained statistical significance for only the DOM and SiH rats (P = 0.0199 and P = 0.0028, respectively)

In accordance with AMP’s effect on behaviors discussed so far, the frequency of rearing behavior was also elevated in a dose-related fashion, at least at the pre-diad test (Fig. 4a). The rank/housing status did not have an overall significant effect on the frequency of rearing. Although AMP’s main effect on rearing frequency was significant (F1,  68 = 6.569, P = 0.0126), the effect of test time and the interaction of rank/housing and test time just missed significance (P = 0.0689 and P = 0.0510, respectively). Nevertheless, the interaction of treatment dose and test time was statistically significant (F1,  68 = 5.816, P = 0.0186), reflecting the somewhat higher frequency of rearing at the diad test. These analyses indicate that drug dose was the major influence on rearing frequency, with a smaller contribution of the time of testing. Separate repeated ANOVA analysis of the two test times indicated that at the pre-diad test the AMP dose had a significant effect on the frequency of rearing (F1,  34 = 17.624, P = 0.0002); however, the effect of rank/housing status just missed significance (P = 0.0662), and its interaction with rank/housing status was also not significant. The Sdom rats treated with the 2.7 mg/kg of AMP reared with greater frequency compared to their response to the lower drug dose (P = 0.0432). By contrast, at the diad test, rearing behavior was not significantly affected by either rank/housing status or drug dose. Lastly, the pre-diad/diad repeated measure ANOVA disclosed that the 0.9 mg/kg AMP-treated DOM rats reared more frequently at the diad compared to the pre-diad test (P < 0.0001). Summarizing, all AMP-treated rats displayed a dose-related increase in rearing frequency at the pre-diad test and an enhanced responsiveness to the lower dose only at the diad test.
https://static-content.springer.com/image/art%3A10.1007%2Fs00213-011-2339-8/MediaObjects/213_2011_2339_Fig4_HTML.gif
Fig. 4

Effect of amphetamine on frequency and duration of rearing behavior of diad- and single-housed rats. A saline or an amphetamine injection (0.9 or 2.7 mg/kg IP) was administered 30-min before the open field test. Rearing behavior was assessed during a 10-min long test. The data (mean ± SEM) represents the percent change in the frequency of rearing behavior after each drug treatment compared to saline treatment at both test times (n = 13 for the two diad-housed rats and n = 11 for the single-housed rats). D designates significant differences from the dominant group. Asterisk designates statistically significant differences for the 0.9 mg/kg group compared to the 2.7 mg/kg treated group at both tests. Caret designates statistically significant differences for the pre-diad test compared to the diad test at each dose level

Rearing duration

At the pre-diad test, the future rank/housing status had a significant effect on the time subjects engaged in rearing behavior (F2,  34 = 3.541, P = 0.0401) (Table 2). Single ANOVA analysis indicated that at the pre-diad test the saline-treated DOM rats reared longer than their Sdom counterparts (P = 0.0136). In addition, the duration of rearing of all the rats was shorter at the diad compared to the pre-diad test (P < 0.0001, P = 0.0015, and P = 0.0358 for the DOM, Sdom and SiH rats, respectively). However, at the diad test, the duration of rearing behavior was not influenced by their rank/housing status.

By contrast to the rearing frequency measure, and the two previously described behaviors, at the pre-diad test, the effect of AMP on rearing duration was relatively small, though it was clearly greater at the diad test. The overall repeated ANOVA analysis indicated a significant main effect of rank/housing, drug dose as well as test time on the duration of rearing (F2,  68 = 11.567, P < 0.0001, F1,  68 = 6.769, P = 0.0114, and F1,  68 = 40.094, P < 0.0001, respectively) (Fig. 4b). However, none of the interaction effects attained statistical significance. Overall, both the DOM and the SiH rats spent less time rearing than did the Sdom rats (P < 0.0001 and P = 0.0003, respectively). Analyzing the two test times independently, we note that rank/housing status and AMP treatment had significant main effects on rearing duration at the pre-diad test (F2,  34 = 4.448, P = 0.0192 and F1,  34 = 12.259, P = 0.0013, respectively) and also at the diad test (F2,  34 = 4.357, P = 0.0207 and F1,  34 = 5.478, P = 0.0253, respectively). At the pre-diad test, the 2.7 mg/kg AMP-treated future Sdom reared longer than the correspondingly treated future DOM rats (P = 0.0040), and these Sdom rats also reared longer after the 2.7 mg/kg dose compared to the lower drug dose (P = 0.0017). Lastly, comparing the two test times, we find that rearing duration of the 0.9 mg/kg AMP treated DOM and Sdom rats was longer at the diad compared to the pre-diad test (P = 0.0182 and P = 0.0067, respectively), with a similar effect also evident after the higher AMP dose (P = 0.0063 and P = 0.0170, respectively). However, the SiH rats reared significantly longer at the diad compared to the pre-diad test only after the larger AMP dose (P = 0.0391). In summary, primarily at the diad test, treatment with AMP enhanced rearing duration of all rats, with Sdom rats displaying the greatest sensitivity to the drug.

Grooming frequency and duration

Rank/housing status had a significant effect on the frequency of grooming behavior of the saline-injected rats at both the pre-diad and the diad tests (F2,  34 = 4.981, P = 0.0127 and F2,  34 = 3.782, P = 0.0329, respectively) (Table 2). Single ANOVA analysis disclosed that at the pre-diad test the future DOM rats engaged in more frequent grooming behavior than either the future Sdom or the SiH rats (P = 0.0097 and P = 0.0206, respectively), while at the diad test, the DOM rats groomed more frequently than the SiH rats (P = 0.0130). Moreover, both diad rats tended to groom more frequently at the diad compared to the pre-diad test, though this effect reached statistical significance only for the Sdom rats (P = 0.0023), while grooming frequency of the SiH rats was the same at both tests.

Rank/housing status had a significant effect on the duration of grooming behavior of rats at the pre-diad test (F2, 34 = 17.711, P < 0.0001), but the effect missed statistical significance at the diad test (Table 2). The single ANOVA analysis indicated that at the pre-diad test the future DOM rats groomed significantly longer than did either the future Sdom or the SiH rats (P = 0.0082 and P = 0.0005, respectively). On the other hand, at the diad test, the DOM and Sdom rats groomed significantly longer than the SiH rats (P = 0.0002 and P < 0.0001, respectively). Furthermore, the diad-housed rats tended to groom longer at the diad test compared to the pre-diad test, but this effect was statistically significant only for the Sdom rats (P = 0.0067).

Interestingly, grooming behavior was exceptionally sensitive to AMP treatment. Even the lowest AMP dose nearly completely obliterated grooming behavior in all three experimental groups. After the 0.9 mg/kg AMP dose, only five DOM, three Sdom, and two SiH rats displayed some minimal grooming, with the number decreasing, respectively, to 3, 0, and 2, after the 2.7 mg/kg drug dose (n = 13 for the diad rats and n = 11 for the SiH group). Considering that grooming behavior was practically absent, AMP’s effect on grooming behavior could not be analyzed statistically.

Plasma corticosterone

Plasma CORT levels were determined twice, the first tail blood sample was obtained 51 days from the start of diad-housing (day 139 of the study), and the second trunk blood sample was obtained on the final day of the study (day 213). Housing/rank status had a significant effect on plasma CORT levels (F2,  34 = 4.305, P = 0.0216), as did the test time (F1,  34 = 8.797, P = 0.0055) (Fig. 5). Overall, plasma CORT levels of the DOM rats were higher than those of the Sdom rats (P = 0.0060). Specifically, at the diad test, plasma CORT levels of the DOM rats were higher compared to those of the Sdom and SiH rats (P = 0.0005 and P = 0.0011, respectively). On the last day of the study, there was a similar trend in plasma CORT levels, but the group differences missed statistical significance.
https://static-content.springer.com/image/art%3A10.1007%2Fs00213-011-2339-8/MediaObjects/213_2011_2339_Fig5_HTML.gif
Fig. 5

Plasma corticosterone levels of diad and single-housed rats. Blood samples were obtained on days 139 and 213 of the study. Data are presented as the mean ± SEM for 13 rats per diad-housed group and n = 11 for the single-housed rats. D designates significant difference from the dominant rats

Discussion

Basal open field behaviors

Behavior of rodents in an open field arena is generally taken to reflect exploratory behavior (Dai et al. 1996). All saline-injected rats displayed similar peripheral locomotor activity at both tests. The lack of hierarchical or housing differences in locomotor activity at the diad test confirms previous reports (del Pozo et al. 1978). Other investigators noted that Sdom rodents and rodents defeated in a resident-intruder test were either more (Blanchard et al. 2001; Nikulina et al. 2004) or less active than DOM rodents (Blakley and Pohorecky 2006; Meerlo et al. 1996). These apparent inconsistencies may be attributed to experimental variables, such as handling and pre-test injection, which may have had significant biobehavioral consequences (Gariépy et al. 2002; Hall et al. 1997; Sciolino et al. 2010). The subjects in our study were extensively handled, injected, and tested in a large non-novel open field arena at both tests, minimizing the anxiety generated by the fear of open spaces, nevertheless their contribution to group differences in these variables remains to be determined.

Avoidance of a novel open field center has been interpreted to reflect an “anxiety-like” state (Ramos et al. 2003). The behavioral differences observed at the pre-diad test may indicate differences in innate emotionality and explorative drive, encompassing such distinct aspects as arousal, impulsivity, and risk taking behavior. Based on previous literature, the future Sdom rats exhibited enhanced emotionality. By contrast, at the diad test, saline-injected rats did not differ in frequency of center entries. At the diad test, only the future Sdom rats displayed increased center behavior indicating a less emotional state, while a 32% decline in center duration of the saline-injected future DOM rats suggested a more emotional state. The enhanced emotionality of the saline-injected DOM rats could indicate a more successful coping behavior of the Sdom rats (Haller et al. 1996; Peterson and Pohorecky 1989). At the diad test, the DOM rats were more emotional than the Sdom rats, in line with the greater “anxiety” of DOM mice compared to non-dominant mice (Ferrari et al. 1998) and the lack of “anxiety-like” behavior of defeat rats (Duncan et al. 2006; Zelena et al. 1999). Chronic psychosocial stress may also obliterated trait-like behavioral difference of future DOM and Sdom rats and indicated the former were “natural risk takers” (Davis et al. 2009; Duncan et al. 2006). Enhanced “anxiety” in defeated rodents has also been reported (Heinrichs et al. 1992; Ruis et al. 1999). These diverse findings point to a critical dependence on experimental context when assessing the emotional state of rodents and indicate that interpretation of “anxiety-like” state based on explorative behaviors is complex.

Rearing behavior in an open field has been interpreted as exploratory activity, emotionality, and vigilance (Ducottet et al. 2004; Thiel et al. 1998). Since future DOM rats reared longer, though not more frequently, they were possibly more vigilant than future Sdom rats. Rearing frequency declined at the diad compared to the pre-diad test, but there were no rank/housing differences as noted previously (Blanchard et al. 2001). Lastly, grooming may serve a variety of functions in rodents including as displacement and de-stressing behaviors (Spruijt et al. 1992). Since stressors do influence grooming behavior (van Erp et al. 1994), one would expect rats to groom less at the pre-diad test compared to the diad test. In fact, except for Sdom rats, grooming was similar at both tests. Thus, enhanced grooming of the Sdom rats at the diad test may reflect an active coping strategy and/or an enhanced de-stressing response. At both test times, the DOM rats displayed the highest level of grooming behavior, while the SiH rats spent the least time grooming at the diad test. Whether these differences in grooming behavior were due to altered responsiveness to stress is unclear.

Effect of amphetamine

Amphetamine increased locomotor activity, confirming its well-known stimulant effect (Smith et al. 1997), with enhanced sensitivity in chronically psychosocially stressed animals (Araujo et al. 2006; Covington and Miczek 2001; 2005). Sensitivity to AMP’s locomotor activation depended on drug dose, the subject’s rank/housing status, and the test period. While SiH rodents were more responsive to an AMP challenge compared to group-housed rodents (Ahmed et al. 1995; Gaytan et al. 1996), these studies did not focus on hierarchical distinctions in their subjects. All rats, particularly the DOM rats, were more sensitive to the lower AMP dose at the diad test, while sensitivity to the higher dose was similar at both tests, confirming the suggestion of qualitative differences in the behavioral effects AMP in group-housed compared to SiH animals (Nikulina et al. 2004; Smith and Byrd 1984). A recent report described a blunted locomotor response to AMP in rats defeated as adolescents (Burke et al. 2010). Hence, the long-term effects of psychosocial stress remain to be further explored.

Amphetamine also enhanced the frequency of entries to the central arena, particularly at the lowest dose. Enhanced center activity of DOM and SiH rats was similar at both test times, but was smaller at the diad test in Sdom rats. At the diad test, center time was increased by the low dose of AMP particularly in the DOM rats, but the diad-housed rats spent less time in the center after the highest dose. The difference in effectiveness of the two AMP doses might be related to its distinct dose-related electrophysiological effects (Homayoun and Moghaddam 2006).

Amphetamine also stimulated rearing behavior, mainly at the diad test. The sensitivity of the diad-housed rats, particularly the Sdom rats, to AMP was greater than of SiH rats. Again, the significance of these social stress-induced differences in sensitivity to AMP requires further evaluation. In sharp contrast to the behaviors discussed so far, grooming behavior was exceptional sensitive to AMP. Even the lowest dose of AMP practically abolished grooming behavior. Hence, under our experimental conditions, grooming was the most sensitive behavior to AMP treatment, showing drastic depression rather than activation at the pre-diad test.

Behavioral sensitization induced by AMP depends on a number of factors, some of which have already been discussed. Handling of subjects increased AMP-induced hyperactivity (West and Michael 1987) and sensitization was more prevalent in SiH rather than in group-housed rats (Gaytan et al. 1996). A potential caveat of our study is that AMP-induced behavioral stereotypy (Antoniou et al. 1998) may have interfered with the expression of open field behaviors. Both test dose and test time are important parameters for expression of AMP-induced stereotypy (Segal 1976). Under our experimental conditions, we did not note any AMP stereotypy, confirming the negative findings of others (Nordquist et al. 2008). One must also consider whether tolerance to AMP contributed to the observed behaviors. Tolerance generally involved continuous, longer term AMP administration (Biała and Kruk 2007; Nielsen 1981). By generously spacing treatment doses and interval between the two tests, we aimed to minimize development of AMP tolerance; nevertheless, this possibility cannot be entirely discounted.

Role of the hypothalamo–pituitary–adrenal axis and stress coping

Social factors are believed to play a significant role in the ability of organisms to cope with chronic stressors. For example, the effects of defeat were mitigated when rats were subsequently pair- or group-housed rather than SiH (de Jong et al. 2005). In contrast with previous work (Pohorecky 2008; Ruis et al. 1999), DOM rats were more emotional and had higher CORT levels compared to the Sdom rats, reflecting better coping behavior (Ebner et al. 2005; Peterson and Pohorecky 1989). This conclusion is supported by the lower display of urine-marking behavior during the social interaction test, and a lack of diad-formation related differential body weight loss (Pohorecky et al. 1999; 2004a, b). Hence, the present diad variant of our earlier triad model was more stressful to the DOM rat, most likely because the removal of the cage barrier exposed its territory to the daily challenge from its cage-sharing partner. The higher plasma CORT levels of the DOM rats may have enhanced their sensitization to AMP (Deroche et al. 1992).

Role of dopamine

Dopamine has been implicated in an organism’s response to both social stress and to psychostimulant drugs such as AMP. Amphetamine-induced locomotor activation has also been associated with enhanced DArgic activity (Ventura et al. 2004; Nordquist et al. 2008). Greater AMP-induced DArgic activation was noted in group-housed rats compared to socially isolated rats (Leng et al. 2004). Prior evidence documents defeat-enhanced DArgic hyperactivity (Anstrom et al. 2009; Ferrari et al. 2003; Tidey and Miczek 1996), and it has been suggested that “ increased phasic DArgic neuron firing in the ventral tegmental area could be considered amongst the features defining the lasting imprint of social defeat stress” (Razzoli et al. 2011). Hence, evidence from defeated subjects highlights the neuroadaptive changes induced by repeated episodes of defeat. One may speculate whether social stress-associated differences in corticotropin-releasing factor (Schmidt et al. 2010; Wood et al. 2010) contributed to the differences in DArgic function associated with social stress, which may underlie the greater sensitivity to AMP of the DOM rats; extra-hypothalamic effects of corticotropin releasing factor include activation of DArgic system and of locomotor and rearing behaviors (Kalivas et al. 1987; Wanat et al. 2008). Lastly, since a substantial literature documents the involvement of noradrenergic and serotonergic systems in mediating AMP’s actions (Lanteri et al. 2008), a significant caveat to our interpretation is the extent of such non-DArgic actions of AMP to the behavioral findings reported here.

Predictive traits of hierarchical status

Previous evidence suggested that hierarchical status may be predicted by certain traits. For example, indices of a less emotional state or a greater “risk taking” trait predicted future social rank (Blakley and Pohorecky 2006; Blanchard et al. 1998, 2001; Davis et al. 2009). Furthermore, DOM rats tended to be stress hyper-responsive, while subordinate rats had an impaired, hyposensitive hypothalamo–pituitary–adrenal axis (Blanchard et al. 1995; Pohorecky et al. 2004a). Our study confirmed that future Sdoms rats made fewer center entries than future DOM or SiH rats, suggesting that they were innately more emotional. Furthermore, the future DOM rats groomed significantly more. The reported individual differences in responsiveness and sensitization to DArgic drugs, including AMP (Ranje and Ungerstedt 1977), may have contributed to these innate rank/housing related behaviors.

In conclusion, administration of the monoamine-releasing drug AMP demonstrated differential involvement of the DArgic neuronal system in the behavioral consequences of chronic psychosocial stress in rats. Compared to the SiH rats, both the DOM and Sdom rats were differentially susceptible to AMP and displayed higher levels of sensitization. One may suggest that differential activation of the hypothalamo–pituitary axis engendered by diad-housing may have contributed to the altered DArgic function, and resulted in altered sensitization to AMP. Since future DOM rats tended to be less emotional in the open field compared to the future Sdom rats, our data suggest that open field center entries following saline injection might be a predictor of future social rank. Furthermore, changes in the DArgic neuronal system may underlie the behavioral and physiological consequences of psychosocial stress.

Acknowledgment

This research was supported in part by funds from the National Institute of Alcoholism and Alcohol Abuse, Grant 1RO1AA10124 and the Center of Alcohol Studies.

Copyright information

© Springer-Verlag 2011