Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The -opioid receptor, which binds enkephalins with high affinity (Mansour et al., 1987), is coupled to Gi proteins and inhibits adenylate cyclase (Wood et al., 1981; Mansour et al., 1995). Functionally,  receptors are important mediators of analgesia, locomotor activation, appetitive, and drug reward behaviors (Calenco-Choukroun et al., 1991; Devine and Wise, 1994; Froehlich et al., 1991; Jiang et al., 1990). In contrast to peptide  receptor agonists, the highly selective nonpeptidic -opioid agonist SNC-80 ([(+)-4-[(a-R)-a-((2S,5r)-4-allyl-2,5-dimethyl-1-piperazinyl)-3-methoxybenzyl-N,N-diethylbenzamide]; Calderon et al., 1994), crosses the blood-brain barrier, which facilitates the investigation of the function of central -opioid receptors (Bilsky et al., 1995).

Stress can alter the response to opioid drugs and psychoactive agents. Morphine analgesia was greater in rats stressed by cold water swimming (Vanderah et al., 1993). Morphine produced hyperthermia in nonstressed rats and hypothermia in stressed rats (Ushijima et al., 1985). Stressors also alter endogenous opioids (Vaswani et al., 1987), but can differ in their capacity to alter opioid binding in brain. Ninety hours of water deprivation increased  receptor binding in the striatum, whereas 20 min of foot shock had little effect (Stein et al., 1992). However, little is known about the interaction of social stressors with opioid systems, particularly in the case of chronic social stress. Previous studies have documented that not only do neurochemical and neuroendocrine differences exist between singly housed and group-housed rats (Blanchard et al., 1993; Fulford and Marsden, 1997), but they also exist in group-housed rats of differing rank (Blanchard et al., 1991, 1993). Group-housed male rats readily establish a social hierarchy with the dominant or  rat displaying specific behaviors toward the subdominant cagemates (Blanchard et al., 1991, 1993; Pohorecky et al., 1995).

The present study characterized the functional expression of - opioid central nervous system receptors in long-term differentially housed rats. The impact of psychosocial stress associated with rank status and/or differential housing on  receptor-mediated function is not known. To assess the functional state of the  receptors, rats were injected with the  receptor agonist SNC-80 (Bilsky et al., 1995) and receptor function was evaluated using physiological and behavioral tests. Rectal temperature response, open field behaviors, tail-flick, ultrasonic vocalizations (USVs), and plasma corticosterone levels were determined. To assess the significance of these functional changes, they were correlated with the cyclic[D-penicillamine²-D-penicillamine²]enkephalin (DPDPE)-stimulated guanylyl 5'-[-thio]-triphosphate ([³⁵S]GTPS) binding to the  receptors in brain sections. Agonist-stimulated [³⁵S]GTPS binding is believed to reflect binding to functionally active receptors (Sim et al., 1995). It was expected that psychological stress would modify specific behavioral and physiological responses to SNC-80. That is, compared with the  rat, which displayed some submissive behavior and was the recipient of most of the  rat's aggression, and with the  rats, which displayed only submissive behavior including prominent USVs, the dominant  rat would show less locomotor and exploratory depression after SNC-80 treatment and less analgesia. Our results confirmed this hypothesis by demonstrating that, overall, subordinate rats were more responsive to the sedative effects of SNC-80, whereas dominant rats exhibited greater activation in response to SNC-80 than singly housed controls. Contrary to expectation, nociception was not related to rank status, and was greater in the singly housed rats. These behavioral and physiological differences may be due at least in part to the observed differences in  receptor function as well as to differences in plasma corticosterone levels.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Subjects and Environment. The subjects were 48 male Long Evans rats (Harlan Sprague-Dawley, Indianapolis, IN) weighing approximately 300 g at the beginning of the study. Rats were individually housed in hanging wire mesh, stainless steel cages for 14 days before triad housing, and were transferred to the appropriate housing cages on day 1 of the experiment. Purina chow and water were available ad libitum throughout the study. The vivarium was kept at 21 ± 1°C, with controlled humidity and a reverse light/dark cycle (12 h each, lights off at 12:30 PM). The study was run as two consecutive cohorts of 24 rats each, for a total of 36 triad-housed and 12 individually housed rats. The housing cages were made of Plexiglas and had a wire mesh floor, one of the cage walls had either one (single cages) or two (triad cages) 1-cm openings for a drinking spout. The cages for individually housed rats were square (25 × 25 × 30 cm), and those for the triad housing cages were rectangular (26 cm wide × 82 cm long × 30 cm high). The animal facilities have been certified by American Association of Laboratory Animal Care, and all the experimental protocols were approved by the Rutgers University Animal Care and Use Committee.

Triad Housing. Although subjects were randomly assigned to triad or individual housing, triad assignments were based on body weights to avoid a possible effect of body size on the development of dominance. The initial body weights were 363.10 ± 15.69 g, 363.42 ± 15.14 g, 362.70 ± 11.63 g, and 364.20 ± 13.02 g for the , , , and singly housed rats, respectively. Rats in two triads were euthanized after only 3 days of differential housing, and the remaining behavioral and physiological data was collected from 10 triad- and 10 singly housed rats tested after 3 and 50 days of differential housing. The 12 triad rats differed in the level of aggression, with 4 triad rats showing highly significant agonistic behavior until the end of the study, whereas another 4 triad rats did not. In one of the highly aggressive triads, the  rat died without showing signs of physical injury, and in another two triads, the subdominant rats showed accelerated loss of body weight during the second half of the study, suggesting a high level of stress. The  rat in these triads had to be separated from its cagemates by a wire screen cage divider. This divider was removed for a 10-min period every day until the end of the study, and the body weights of these triad rats were monitored carefully.

Agonistic Behavior Rating. Agonistic behaviors were first assessed at the time the triads were formed. Rats assigned to a triad were placed into their novel cage, and social interactions were noted during the next 30 min. Offensive and defensive aggressive behaviors (including roll-tumble fights, aggressive grooming, on-top, lateral threat, freezing, on-back, defensive upright, and fleeing) were scored using the method described by Peterson and Pohorecky (1989). Additional measures of aggression, such as signs of bodily attack by the  rat on the cohabitants, were also recorded. Lesions were seen most frequently on the tail but were also found on other parts of the body. The behaviors and signs of bodily attack were rated and the the rank status of , , or  was assigned. The  or dominant rat displayed offensive aggressive behavior toward his cagemates. The rat that was least aggressive and had lost the most body weight 24 h after triad formation, the  rat, rapidly developed strategies to diminish further contact with the  rat. Although the full scope of communication signals is uncertain, it included immobility and USVs when approached by the  rat. Conversely, the  rat displayed few defensive behaviors, and consequently was involved in more frequent aggressive interactions with the  rat (overall score for the interaction with s and s was 19.0 ± 5.4 and 9.3 ± 4.5, respectively). The  rat did not appear to change its pattern of behavioral interaction with the  rat, frequently initiating an interaction leading to a roll and tumble fight.

Experimental Protocol. Rats were extensively habituated to handling before testing. The 3-day post-triad formation test day was selected as a compromise to examine the acutely versus chronically induced neurobiological and behavioral consequences of rank-associated stress. Each rat was subjected to a series of behavioral tests over a 40-min period, beginning with the presumably least stressful and ending with the most intrusive of the tests. The experimental protocol and testing schedule are outlined in Table 1. Testing was conducted between 1:00 and 6:00 PM. Each rat was tested after an injection of vehicle and SNC-80 and the test battery was repeated after 50 days of differential housing. After basal rectal body temperature was measured (Digi-Sense Thermistor Thermometer), vehicle (acidified saline) or SNC-80 (35 mg/kg) was administered i.p., and after a 5-min period, the subject was tested in a modified open field (100 cm × 100 cm divided into 20 quadrants, with 8 equidistantly distributed 4-cm diameter holes). The onset, frequency, and total duration of crossover activity, rearing, grooming, headpoke, and center entry were quantified over a 10-min period with the aid of an IBM-XT computer equipped with a manually operated interface, as previously described (Pohorecky et al., 1989). Fifteen minutes after injection, rectal temperature was again recorded and the rat was tested in a tail-flick analgesiometer (Columbus Instruments, Columbus, OH). The time in seconds for the rat to retract its tail away from a focused beam of light was recorded as the latency for tail-flick; if the rat failed to respond within 40 s, the test was terminated and a maximized score of 40 s was assigned. Twenty minutes after injection, the rat was tested for emission of 22 KHz USVs (QMS Mini Bat Detector, QMC Instruments Ltd., London, England) in response to 2.8 psi air puffs delivered by an Airstim Instrument (San Diego Instruments, San Diego, CA) directed over the rat's head. Once vocalizations were initiated, the administration of air puffs ceased, and the number of bouts of vocalizations was recorded; the test was terminated when no vocalizations were heard during the next 3 min (Knapp and Pohorecky, 1995). Thirty minutes after injection, rectal body temperature was measured again, followed by a tail-flick test.

DPDPE Stimulated [³⁵S]GTPS Binding Autoradiography. The assay was carried out as described by Sim et al. (1995). Briefly, brains were sectioned at 20 µm using a cryostat set at 20°C. Sections were thaw-mounted onto subbed slides, which were stored at 80°C until use. Slides were incubated in assay buffer (50 mM Tris·HCl, 3 mM Mg Cl₂, 0.2 mM EGTA, and 100 mM NaCl, pH 7.4) at 25°C for 15 min. DPDPE-stimulated activity was determined by incubation in 1 mM DPDPE, 0.04 1 nM [³⁵S]GTPS, and 1 mM GDP in assay buffer at 25°C for 2 h. Basal activity was assessed with GDP in the absence of agonist, and nonspecific binding in the presence of 10 mM unlabeled [S]GTP without GDP. Slides were then rinsed twice in ice-cold Tris buffer (50 mM Tris·HCl, pH 7.5 at 25°C), and once briefly in deionized water. After drying overnight, slides were exposed to Hyperfilm-BioMax 20 for 24 h. Films were digitized using a video camera and analyzed using the NIH IMAGE (Scion 1.59) program for Macintosh computers. For quantification of images, densitometric analysis was carried based on standards made with ³⁵S-spiked brain paste.

Plasma Corticosterone Determination. Trunk blood was collected on day 52. After centrifugation at 2000g for 10 min, plasma was frozen at 70°C until analysis. Duplicate 10-µl samples of plasma were used for quantification of corticosterone concentration by radioimmunoassay using ICN kits (ICN Radiochemicals, Irvine, CA).

Adrenal Catecholamine Determination. Adrenals were rapidly frozen on dry ice and stored at 70°C until analysis. After homogenization in 0.4 N perchloric acid and centrifugation for 20 min at 15,000g, fractions of diluted supernatant were analyzed by high-pressure chromatography (Spectra-Physics model SP8770 dual piston pump, and a Biophase ODS C-18 reversed phase column 5 mm, 250 × 4.6 mm i.d. from Bioanalytical Systems, West Lafayette, IN). The detector (LC-4C, BAS) was set at a +0.72 V potential between the glassy carbon electrode and the Ag/AgCl reference electrode. The filtered and degassed mobile phase (0.10 M citric acid, 0.10 M sodium phosphate dibasic and 10% methanol) was pumped at a rate of 1.0 ml/min. Quantification was against external standards.

SNC-80 Dose Determination. An optimal effective dose of 35 mg/kg of SNC-80 (10 mg SNC in acidified saline) was determined by testing the effect of various SNC-80 doses (100, 50, 40, 30, and 25 mg/kg) on rectal temperature and locomotor behavior. At 100 mg/kg, rats showed extreme myoclonus, severe temperature depression, and total loss of locomotor activity for approximately 15 min after injection. Reaction to the drug was not as severe at 50 and 40 mg/kg, however loss of locomotor activity was still prominent for up to 8 min. At the dose of 25 mg/kg, no drug effect was observed.

Statistical Data Analysis. The data are presented as means and S.E.M. The data were analyzed using a SYSTAT-based repeated measures ANOVA. Correlations between measures were determined with the Pearson's product moment correlation using uncorrected probabilities. The between subjects' factor was the rats' housing and rank status (, , , single). The within subjects' factor was drug treatment (saline vehicle or SNC-80). The repeated measures parameter was duration of differential housing (3 and 50 days), and in the case of rectal temperature and tail-flick, the time at which these measures were assessed (T = 1, 15, 30, and 40 min for rectal temperature and T = 15 and 30 min for the analgesia test). Additionally, Bonferroni adjusted linear contrasts were used for the planned comparisons between cell means within the ANOVA. Significance level was set at p  .05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Body Weight Changes

Differential triad housing resulted in significant rank-related long-term changes in the body weights of all rats (Fig. 1). After the initial 24 h of housing, the overall body weights of triad-housed rats were lower than their pretriad weights (F2,32 = 12.942, p < .001). However, the body weight loss of only the  and  rats reached statistical significance (p < .001 for both). The rank-related differences in body weight were maintained throughout the study (F3,36 = 7.942, p < .001). The  rats gained significantly more body weight (39%) over the course of the study than did the subdominant rats (about 19%, p < .001 for both) or the singly housed rats (22%, p < .001).

View larger version (38K): [in this window] [in a new window]   Fig. 1.   Effect of rank status and housing condition on body weight loss at 1 day, and body weight gain at 50 days post-triad formation. Columns and error bars represent mean ± S.E.M. (n = 12/group). * and , indicate significant difference from the corresponding  rats (p  .001).

Rectal Temperature

Basal rectal temperature was dependent on housing condition and rank status (F3,76 = 6.259, p = .001; Fig. 2A), and was lower at the 50-day compared with the 3-day test (F1,76 = 3.988, p = .049). The basal temperature of singly housed rats was lower than that of triad-housed rats (p .026 and p.007 for the 3- and 50-day tests, respectively), but the basal temperatures of  rats tended to be higher at the 50-day compared with the 3-day test. SNC-80 treatment decreased the rectal temperature of all rats (F1,33 = 5.599, p = .024), with greater hypothermia 30 min after injection (F3,66 = 53.755, p < .001) (Fig. 2B). At the 3-day test all rats were similarly affected by SNC-80 treatment. However, after 50 day, rank and housing status modified the hypothermic effect of SNC-80 (F3,32 = 3.653, p = .021). SNC-80-treated  rats were less hypothermic compared with  rats and singly housed rats (p = .034 and p = .004, respectively).

View larger version (68K): [in this window] [in a new window]   Fig. 2.   Effect of rank status and housing condition and of treatment with SNC-80 (35 mg/kg, i.p.) on body temperature measured at 3 and 50 days after triad formation. Temperature was measured before and at 15, 30, and 40 min following SNC-80 injection. A, basal rectal temperatures before injection. B, decline in rectal temperature 30 min after SNC-80 treatment. Control rats were injected with an equivalent volume of saline. Columns and error bars represent mean ± S.E.M. (n = 8-12/group). + and *, indicate significant differences from singly housed rats at the 3- and 50-day tests (p = .05). , indicates significant differences from the corresponding 3-day groups (p  .05).

Open Field Behavior Locomotor Activity

Locomotor activity of triad-housed rats was lower compared with singly housed rats (F3,29 = 4.044, p = .016; Table 2). Although rats were less active overall at the 50-day compared with the 3-day test (F1,27 = 4.368, p = .046), this effect was less prominent in  and  rats. SNC-80-treated rats were more hypoactive compared with saline-injected rats (F1,27 = 25.090, p < .001; Fig. 3A). This depressant effect of SNC-80 was dependent on the duration of differential housing (F1,27 = 5.465, p = .027). Although at the 3-day test SNC-80 depressed locomotor activity of all rats, at the 50-day test the drug did not produce hypoactivity in the  rats. As a consequence, the  rats and singly housed SNC-80- treated rats differed in locomotor activity at the 50-day but not the 3-day test (F1,27 = 6.487, p = .017).

View larger version (31K): [in this window] [in a new window]   Fig. 3.   Effect of SNC-80 (35 mg/kg, i.p.) on open field behavior of rats tested at 3 and 50 days after triad formation. Controls were injected with an equivalent volume of saline. Columns and error bars represent percentage of change in the activity of rats injected with drug versus those of rats injected with saline ± S.E.M. (n = 7-10/group). A, frequency of locomotor activity in the peripheral area of the open field. B, frequency of locomotor activity in the central area of the open field. C, frequency of headpoke behavior. *, indicates significant differences from  rats at 3 days (p = .02). , indicates significant difference from corresponding  rats.

Locomotor Activity in Central Area. Singly housed rats entered the central area more frequently than did triad-housed rats (F3,27 = 37.004, p = .046; Table 3). At the 50-day test, triad-housed rats, particularly the  rats, made fewer entries to the central area compared with the singly housed rats (p = .05). Compared with saline injection, SNC-80 treatment depressed center entries at the 3-day test (F1,27 = 3.944, p = .057; Fig. 3B). After 50 days of differential housing, SNC-80-treated singly housed rats still made few entries to the center of the field, but the  rats made comparatively more entries than saline-injected rats. Because of differences in the SNC's effect on locomotor activity in the center versus periphery of the field, the ratio of center entries to total activity in the open field was calculated (Table 4). Interestingly, although SNC-80-treated triad-housed rats increased center entries by 96% by the 50-day test, there was no change in singly housed rats. Thus, these long-term triad-housed rats differentially increased entries to the central area in contrast to singly housed rats.

Headpoke Behavior. Chronicity of differential housing also depressed the frequency of headpoke behavior (F1,24 = 6.396, p = .018), particularly in the  rats (Table 5). SNC-80 treatment decreased the frequency of headpoke behavior (F1,24 = 47.253, p < .001; Fig. 3C). SNC-80's effect on headpoke frequency depended on the rank and duration of housing (F3,24 = 3.099, p = .046). Again the  rats showed no drug-induced depression of headpoking at the 50-day test, and headpoked more frequently than  rats or singly housed rats (p = .09 and p = .015, respectively).

Rearing Behavior. Saline-injected triad-housed rats reared less frequently than did saline-injected singly housed rats (F3,26 = 2.991, p = .049), particularly at the 50-day test (Table 5). SNC-80 treatment abolished the rearing behavior of all rats (F1,26 = 15.591, p = .001 and F1,26 = 114.028, p < .001 for the 3- and 50-day tests, respectively).

Grooming Behavior. About 84% of all the saline-injected rats engaged in grooming behavior during the open field test. Grooming was almost entirely abolished in rats treated with SNC-80 (F1,28 = 47.637, p < .001). None of the  and  rats and only 10% of the  rats and singly housed SNC-80-treated rats engaged in grooming behavior.

Fecal Boli. Overall, saline-injected rats defecated little during the open field test. After SNC-80 treatment defecation by triad-housed rats declined further. Interestingly, drug-treated singly housed rats showed a remarkable increase in defecation (3.5-fold at the day-3 test and 5-fold at the day-50 test).

Analgesia

Compared with saline-injected rats, SNC-80-treated rats had longer tail-flick latencies (F1,28 = 255.65, p < .001; Table 6). Irrespective of the duration of differential housing, this analgesic drug effect was similar in triad-housed and singly housed rats; moreover, there were no rank-related differences.

USV

Housing condition, but not rank status of rats, had a significant effect on USVs induced by air-puff stimuli (Table 7). Compared with triad-housed rats, saline-injected singly housed rats had shorter latencies, and emitted fewer USVs, irrespective of the duration of differential housing (F1,37 = 11.667 and F1,38 = 8.275, p = .007, respectively). SNC-80 treatment completely suppressed USVs by all rats.

Plasma Corticosterone

Rank status and housing condition significantly altered plasma corticosterone levels (F3,28 = 3.262, p = .036; Fig. 4).  and  rats had lower corticosterone levels compared with the  rats (p = .018 and p = .033, respectively). Corticosterone levels of  rats were also lower compared with singly housed rats (p = .042).

View larger version (69K): [in this window] [in a new window]   Fig. 4.   Plasma corticosterone concentration of rats housed for 50 days in triads or singly. Columns and error bars represent the mean concentration of corticosterone in plasma of rats injected with drug versus saline ± S.E.M. (n = 8-9/group). *, indicates significant difference from  rats (p  .033). +, indicates significant difference from singly housed rats (p = .042).

Adrenal Catecholamines

After 50 days of differential housing, the concentration of adrenal catecholamines was similar in triad-housed and singly housed rats, and was not correlated with rank status (Table 8).

Correlations between Rank Status/Housing and Behavioral and Physiological Measures

The drug-induced hypothermia correlated with several of the measures assessed in the open field test. In long-term triad-housed rats, SNC-80-induced hypothermia correlated positively with drug-induced hypoactivity, with the frequency of entries to the center of the field, and with the frequency of headpoking (r = .673, p < .001, r = .719, p = .029 and r = .792, p = .014, respectively). Drug-induced hypothermia also correlated positively with the frequency of grooming behavior of saline-injected rats (r = 1.00, p < .001). Additionally, frequency of locomotor activity correlated negatively with tail-flick latencies of SNC-80-treated rats (r = .776, p < .001).

Locomotor activity following saline injection correlated positively with adrenal norepinephrine (r = .968, p = .032). Plasma corticosterone correlated negatively with tail-flick latency following saline injection (r = .429, p = .018), and positively with locomotor activity of the  rats (r = .835, p = .039). Additionally, plasma corticosterone levels correlated positively with adrenal epinephrine and norepinephrine (r = .834, p = .039 and r = .661, p = .05, respectively).

DPDPE-Stimulated [³⁵S]GTPS Binding

DPDPE-stimulated [³⁵S]GTPS binding in the left prefrontal cortex (PFCx) depended on the subject's rank and housing condition (F3,20 = 4.928, p = .010; Fig. 5A). Singly housed and  rats had more binding in the left PFCx compared with  rats (p = .018) and  rats (p = .002). Moreover, DPDPE-stimulated [³⁵S]GTPS binding in the right PFCx of singly housed rats was higher than that of triad-housed rats, which showed no rank-related differences in binding.

View larger version (46K): [in this window] [in a new window]   Fig. 5.   DPDPE-stimulated [35S]GTPS binding in the PFCx (A) and the central nucleus of the amygdala (B) of rats housed in triads or singly for 50 days. The columns and error bars represent the stimulation in fmol of [35S]GTPS bound per mg tissue ± S.E.M. * Indicates significant difference from corresponding singly housed rats (p = .01). +, indicates significant difference from corresponding  rats (p = .05).

In the central nucleus of the amygdala, rank status and differential housing also altered DPDPE-stimulated [³⁵S]GTPS binding (F3,20 = 3.82, p = .03 for the right, and F3,20 = 4.68, p = .01 for the left side; Fig. 5B). Binding in the left amygdala of singly housed rats was also higher than that of triad-housed rats (p = .003 and p < .001 for  and  rats, respectively). Similarly, in the right amygdala, DPDPE-stimulated [³⁵S]GTPS binding in singly housed rats was higher than that of , , and  rats (p = .033, p < .001, and p < .001, respectively); moreover, the binding in  rats was higher compared with the  rats (p = .017). It should be stressed that, as in the PFCx, DPDPE-stimulated [³⁵S]GTPS binding in the amygdala was lateralized (F3,11 = 4.527, p = .037). Lateralization of binding in the amygdala was significant for the  rats (p = .039),  rats (p = .047), and singly housed (p = .012) rats.

In the posterior aspects of the nucleus accumbens DPDPE-stimulated [³⁵S]GTPS binding was dependent on the subject's rank status and housing condition (F3,25 = 6.35, p < .001). However, in contrast to the PFCx and the amygdala, there was no lateralization of binding in this brain area (Fig. 6A). Singly housed rats had higher DPDPE-stimulated [³⁵S]GTPS binding in the accumbens compared with , , and  rats (p < .0001, p = .05 and p = .0002, respectively). Among the triad-housed rats, the  rats had the highest binding (p = .033 and p = .04 compared with  and  rats, respectively).

View larger version (66K): [in this window] [in a new window]   Fig. 6.   DPDPE-stimulated [35S]GTPS binding in the nucleus accumbens (A), the median eminence (B), the arcuate nucleus (C), and the thalamus (D) of rats housed in triads or singly for 50 days. The columns and error bars represent the stimulation in fmol of [35S]GTPS bound per mg tissue ± S.E.M. *, indicates significant difference from singly housed rats for A, B, and D (p  .01). +, indicates significant difference from  rats from A and B and from  rats for D (p  .01).

In the median eminence (Fig. 6B) and the thalamus (Fig. 6D), DPDPE-stimulated [³⁵S]GTPS binding was also related to rank status and housing condition (F3,25 = 3.965, p = .012, and F3,25 = 4.007, p = .009, respectively). In the median eminence, the binding in  and  rats was lower compared with  rats (p = .006 and p = .007, respectively) and singly housed rats (p = .001 and p = .002, respectively). In the thalamus, binding in  rats was lower compared with , , and singly housed rats (p = .004, p = .019 and p = .003, respectively). Finally, in the arcuate nucleus, differential housing condition had no effect on  receptor-mediated binding, possibly because of the larger variability of the data (Fig. 6C). The trend in binding, however, indicated a pattern similar to that in the nucleus accumbens and median eminence.

To summarize, DPDPE-stimulated [³⁵S]GTPS binding was lateralized in the PFCx and in the amygdala but not in the nucleus accumbens. The data indicate that after 50 days of differential housing,  receptor activation in the right PFCx, the amygdala, and the nucleus accumbens was higher in singly compared with triad-housed rats. In triad-housed rats,  receptor activation tended to be highest in the  rats.

Correlations between Rank and Housing and Functional  Receptor Binding

To further assess the significance in the obtained measures, correlations between behavioral and physiological parameters and the functional binding of the  receptor were determined. Laterality of DPDPE-stimulated [³⁵S]GTPS binding in the PFCx correlated positively with the frequency headpokes of SNC-80-treated rats (r = .695, p = .015). Lateralization in DPDPE-stimulated [³⁵S]GTPS binding in the amygdala also correlated positively with locomotor activity of saline-injected rats (r = .611, p = .012) and negatively with SNC-80- induced hypothermia (r = .61, p = .013).

In saline-injected  rats, locomotor activity correlated negatively with DPDPE-stimulated [³⁵S]GTPS binding in the left amygdala (r = .993, p = .007). Also in saline- but not in SNC-80-treated rats, locomotor activity correlated positively with binding in median eminence (r = .606, p = .010) and the arcuate (r = .509, p = .040). Frequency of rearing of SNC-80-treated rats correlated positively with the binding in amygdala and in arcuate nucleus (r = .645, p = .04 and r = .709, p = .02, for the left and right amygdala, and r = .60, p = .050 for the arcuate). Both the frequency and duration of vocalizing of SNC-80-treated rats, and latencies to tail-flick, correlated positively with DPDPE-stimulated [³⁵S]GTPS binding in the left amygdala (r = .453, p = .50, r = .484, p = .042, and r = .612, p = .035, respectively).

Lastly, plasma corticosterone concentrations correlated positively with DPDPE-stimulated [³⁵S] GTPS binding in the right amygdala (r = .858, p = .003). Adrenal norepinephrine correlated negatively with DPDPE-stimulated binding in the left amygdala (r = .735, p = .03) and in the median eminence (r = .759, p = .02), and positively with binding in the left PFCx of  and singly housed rats (r = 1.00, p = .010, and r = .968, p = .032, respectively). Therefore, rats with greater binding in the left amygdala had lower norepinephrine levels in their adrenals, whereas those with higher binding in the right amygdala had higher plasma corticosterone levels.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The present data demonstrate that in rats the stress of housing and of rank status altered the functional activity of -opioid receptors. Moreover, housing and rank status influenced the physiological and behavioral effects produced by SNC-80 stimulation of the -opioid receptors.  and singly housed rats showed the largest analgesia after SNC-80 treatment, whereas triad-housed rats were more resistant to the drug's analgesic effect. The behavior of SNC-80-treated  and  rats in the open field differed from that of  rats. In general, rats with higher stress levels were more sensitive to the depressant effects of SNC-80.

Housing Stress and Body Weight Changes. Although all rats regained the initially lost body weight within 5 days after triad formation,  rats continued to weigh more than their cagemates. Rank- and housing-associated differences in body weight have been reported for rats (Blanchard et al., 1993; Pohorecky et al., 1995) and marmosets (Johnson et al., 1996). Surprisingly, the weight gain of singly housed rats was similar to that of subdominant rats. Thus body weight responds readily to housing conditions and social stressors.

Housing Stress and Body Temperature. SNC-80- induced hypothermia was similar in all short-term differentially housed rats, despite the higher basal temperatures of triad-housed rats. As rats adjusted to their new housing conditions, their basal temperatures declined. Probably because  rats were the more frequent recipients of the offensive aggression of  rats, their basal temperature did not habituate.  Receptors can modulate body temperature, and DPDPE decreased temperature of both nonstressed and stressed subjects (Spencer et al., 1988). However, because the  receptor-mediated [³⁵S]GTPS binding did not correlate with rectal temperature, it is doubtful that the hypothermia was directly mediated by these receptors, although an effect in drug-induced hypoactivity is suggested by its positive correlation with the binding.

Housing Stress and Open Field Behaviors. Time- related habituation to housing conditions and the testing apparatus is evident from the decline in locomotor activity from the 3-day to the 50-day test. In contrast to singly housed rats, triad-housed rats became desensitized to novel environments, in support of the "hyperarousal" theory for isolated rats (Einon and Morgan, 1978). Group-housed rats are known to be less active in the open field than singly housed rats (Dalrymple-Alford and Benton, 1981).

Housing Stress and Sensory Reactivity. The longer tail-flick latencies of SNC-80-treated rats attested to its analgesic effect, originally described in mice (Bilsky et al., 1995). There were no rank- and housing-related differences in nociception, as previously reported (Woodworth and Johnson, 1988), although longer latencies in isolated rats have also been reported (Naranjo and Fuentes, 1985). Latencies for tail-flick correlated positively with DPDPE-stimulated [³⁵S]GTPS binding in the amygdala, corroborating the role of these receptors in nociception (Spencer et al., 1988; Calcagnetti et al., 1989). The analgesia of rats with high corticosterone levels supports the finding that glucocorticoids decrease opioid nociception (Ratka et al., 1988). Because stress produces tolerance to DPDPE-induced nociception (Calcagnetti et al., 1989), tolerance of  receptors may explain the lack of rank-related differences in nociception in saline-injected triad-housed rats. Thus both -opioid receptors and glucocorticoids may have modulated nociception in the present study.

Housing Stress and Adrenocortical and Adrenomedullary Hormones. and  rats had lower plasma corticosterone levels compared with  and singly housed rats. The level of control subjects have in stressful conditions may determine the severity of the stress response and the balance between the hypothalamo-pituitary-adrenal axis and sympathoadrenal activity (Henry, 1992). Because plasma corticosterone levels are generally taken to reflect the level of stress,  and  rats appeared to have implemented successful coping strategies. The high plasma corticosterone levels of  rats, on the other hand, suggests that their coping strategies were not successful, because they continued to be challenged by the  rats. Corticosterone levels of singly housed rats were similar to those of  rats, and of rats sacrificed 3 days after triad formation (227.80 ± 9.53 ng/ml, n = 4), when colonies were still highly unstable. By this measure, individual housing of rats appeared to have been stressful.

Housing Stress and  Receptor-Activated Binding. Although the presence of -opioid receptors and its mRNA in brain is well known (Mansour et al., 1995), to our knowledge, lateralization of receptor binding has not been reported. It is interesting that lateralization of  receptor binding in both the PFCx and amygdala was totally absent in  rats, was prominent in  rats, which in contrast to the other rats, showed sensitization rather than depression to SNC-80 in open field behaviors. Similarly, rats with left-biased amphetamine-induced rotation were reported to be more sensitized by stress (LaHosta et al., 1988). Rats with lateralized  receptor binding in the amygdala were more active in the open field. Those with greater binding in the left amygdala were more sensitive to air-puff-induced vocalizations, but the SNC-80- treated rats were less motorically active, and showed less sensory sensitivity (air puffs and pain). Thus, lateralized binding in the amygdala, high locomotor activity, and sensory sensitivity correlated positively with greater sensitivity to SNC-80-induced depression in these measures. Higher right amygdala binding in  rats, on the other hand, correlated positively with more headpoke behavior and higher plasma corticosterone levels. The higher corticosterone of  rats most likely reflects the stress from frequent interactions with  rats. These findings are in agreement with the notion that activation of the amygdala is crucial for the retrieval and analysis of information relevant to a stressor, for the stimulation of the HPA axis and the sympatho-adrenomedullary system by emotional stressors (Huether,1996), and for amygdala's role in vocalizations and reactions to pain (Garcia et al., 1998). Saline-injected  and singly housed rats with higher left PFCx binding had higher adrenal norepinephrine levels.  and singly housed rats with higher PFCx binding made more center entries, moreover  rats with lateralized binding in the PFCx engaged in more exploratory headpoke behavior, suggesting that the  rats may have had a greater impulsive drive, which is supported by their engagements with the  rats. These findings are consistent with the currently held view that psychological stressors activate the PFCx. By interpreting sensory and emotional stimuli, the PFCx is crucial for the assessment of potential danger, the generation of fear and anxiety, and the activation of the HPA axis (Huether et al., 1996).