Effects of social defeat and of diazepam on behavior in a resident–intruder test in male DBA/2 mice

Abstract

Social stress induces robust behavioral and physiological changes, some of which may alter the responsiveness to pharmacological agents, including diazepam (DZP). We used a resident–intruder paradigm to (1) develop a comprehensive ethogram of behavioral changes following social defeat (SD) in the socially reactive strain, DBA/2 male mice, (2) determine whether acute exposure of DBA/2 mice to low-dose DZP would induce flight or aggressive behavior, both of which have been observed in other rodent models and (3) to test whether prior social stress affects responses to DZP. Behavioral responses to a nonaggressive intruder (NAI) mouse 24 h post-SD were measured in resident subject mice exposed to DZP (0, 0.5, 2.0 mg/kg, ip) either prior to the resident–intruder test (Experiment 1) or immediately post-SD (Experiment 2); control mice were not defeated (NOSD). In general, SD mice displayed increased passive and active avoidance, defense, immobility, and risk assessment relative to NOSD mice. In Experiment 1, mice treated acutely with 0.5 mg/kg DZP had more approach and flight behavior, while those treated with 2.0 mg/kg DZP had more avoidance than vehicle-treated mice, independent of SD. In Experiment 2, acute DZP (2 mg/kg) induced effects 24 h later, possibly secondary to withdrawal. In a nonsocial context (Experiment 3), DZP increased exploratory activity.

Introduction

Social stress induces marked physiological changes including increased activity of the sympathetic nervous system and of the hypothalamic-pituitary axis, and has pervasive behavioral effects (reviewed in [7], [33], [43]). Social defeat (SD) has been used as an animal model of depression [33], which is often a symptom of, or comorbid with, anxiety disorders [49]. Some social stress-induced behaviors may be analogous to symptoms of anxiety disorders, such as exaggerated avoidance often seen in post-traumatic stress disorder (PTSD; [47]). In the diverse class of clinical anxiety disorders, several symptoms are not responsive to the benzodiazepines (BZs). Symptoms of panic disorder, PTSD, obsessive–compulsive disorder and specific phobias, all of which are classified as anxiety disorders in the DSM-IV [14] and are associated with fear and autonomic arousal, typically are not responsive to standard BZ treatment (reviewed in [26], [46]). The potential utility of ethological models in the study of preclinical pharmacology of anxiety disorders has recently received increased consideration [5].

We used a modified resident–intruder test to develop a comprehensive ethogram of behavioral changes following SD in male DBA/2 mice and tested their responses to acute diazepam (DZP) exposure. In the resident–intruder test, a resident mouse usually approaches and attacks an intruder placed within its cage (reviewed in Ref. [38]). After SD, resident mice display fewer approaches and greater avoidance, defense, and active flights than do nondefeated (NOSD) mice, even in response to nonaggressive intruders (NAIs) [30], [43], [44]. In addition to these measures, we presently report on the effects of SD on risk assessment. Although measures of risk assessment including stretched-approach and stretched-attend posture (SAP) have typically been measured in nonsocial contexts (reviewed in [15], [42]), these behaviors have also been measured in response to a predator's odor [2], [24] and a “stretched posture” in mice confronting a conspecific has been described [25]. We currently use the term stretched approach to refer to slow approach of the subject towards the intruder, with its stretched body lowered to the ground. We use the term SAP to refer to when the subject's neck is extended, moving its head forward while its hind legs are positioned together on the ground. These behaviors appear to be investigative, yet cautious in nature. We used three subtests, each with increasing threat exposure, to broaden our ethological analysis of agonistic behavior.

DBA/2 mice are considered to have an intermediate level of anxiety relative to other mouse strains, based on performance in the elevated plus maze, open field test and light–dark test [11], [51]. In most animal models of anxiety, including tests of conflict, social interaction, exploratory behavior, fear-potentiated startle and stress-induced vocalizations, BZs are anxiolytic (reviewed in [21], [32], [42]). However, depending on the context, dose, treatment regimen, particular BZ, route of administration, and prior experience, BZs may induce enhanced flight behavior (reviewed in Ref. [15]). In resident hamsters, acute DZP exacerbated flight from an NAI in a generalization test 24 h after SD, both in hamsters that received DZP immediately after SD, and in those that received DZP prior to the test [28]. Dixon and Kaesermann [15] reviewed animal studies in which DZP induced approach and flight and suggested that this behavioral disinhibition might be analogous to “paradoxical aggression” sometimes elicited by anxiolytics in humans. Low doses of DZP can increase aggressive behavior, while high doses inhibit it (reviewed in Ref. [39]). Acute DZP exposure increased aggressive behavior in humans with low-anxious individuals responding more aggressively than high-anxious individuals [54]. During SD, DBA/2 mice are highly reactive [44]. These mice continue to display escape attempts following repeated defeats, while C57BL/6 mice inhibit escape attempts [36]. In addition, DBA/2 mice did not display anxiolytic responses to BZ when tested in the light–dark test [11]. We hypothesized that during an agonistic encounter, this mouse strain would display enhanced flight or reactivity to low doses of DZP, rather than aggressive behavior.

Pharmacological agents, including BZs, often affect socially stressed animals differently than they do nonstressed animals [52]. A variety of stressors, including swim, handling, noise and tail pinch affect the GABA/BZ complex; the nature and direction of effects are dependent on the type of the stressor (reviewed in Ref. [19]). Stress may also modify the effect of DZP through the induction of DZP binding inhibitor [48]. Stress effects on the pharmacokinetics of DZP may affect further stress responses. Social stress in mice increases BZ receptors, an effect dependent on adrenal integrity [40]. A stressful event remote in time might alter the effects of DZP. Antelman et al. [1] reported that a single episode of restraint stress in rats was shown to block the effects of DZP one month later. We hypothesized that mice that received SD prior to DZP exposure would respond differently than nonstressed mice.

Goals of the current experiments included (1) the expansion of our ethological analysis of social-stress-induced behavioral changes in DBA/2 mice, (2) the determination of how this “intermediate anxious” mouse strain would respond under the influence of, and following withdrawal from, DZP, and (3) the determination of whether previous social stress would affect the responsiveness of DBA/2 mice to DZP during an agonistic encounter. In Experiment 1, resident subject mice received DZP prior to the resident–intruder test, 24 h after SD or NOSD, in order to test whether DZP would affect behavior in this generalization test. Since DZP has amnesic effects [41], in Experiment 2 we tested whether DZP administered immediately after SD would block acquisition of defeat-induced behavioral changes tested 24 h later. In Experiment 3, we tested whether DZP would affect exploratory activity in a nonsocial situation, in order to determine whether DZP effects on activity levels are context-dependent. These experiments provide additional measures of social stress induced behavioral changes and provide further support that acute DZP exposure may have paradoxical effects depending on the social context.