Lipopolysaccharide induces delayed FosB/DeltaFosB immunostaining within the mouse extended amygdala, hippocampus and hypothalamus, that parallel the expression of depressive-like behavior

Summary

Proinflammatory cytokines induce both sickness behavior and depression, but their respective neurobiological correlates are still poorly understood. The aim of the present study was therefore to identify in mice the neural substrates of sickness and depressive-like behavior induced by lipopolysaccharide (LPS, 830 μg/kg, intraperitoneal). LPS-induced depressive-like behavior was dissociated from LPS-induced sickness by testing mice either at 6 h (at which time sickness was expected to be maximal) or at 24 h post-LPS (at which time sickness was expected to be minimal and not to bias the measurement of depressive-like behavior). Concurrently, the expression of acute and chronic cellular reactivity markers (c-Fos and FosB/ΔFosB, respectively) was mapped by immunohistochemistry at these two time points. In comparison to saline, LPS decreased motor activity in a new cage at 6 h but not at 24 h. In contrast, the duration of immobility in the tail suspension test was increased at both 6 and 24 h. This dissociation between decreased motor activity and depressive-like behavior was confirmed at 24 h post-LPS in the forced swim test. LPS also decreased sucrose consumption at 24 and 48 h, despite normal food and water consumption by that time. At 24 h post-LPS, LPS-induced depressive-like behavior was associated with a delayed cellular activity (as assessed by FosB/ΔFosB immunostaining) in specific brain structures, particularly within the extended amygdala, hippocampus and hypothalamus, whereas c-Fos labeling was markedly decreased by that time in all the brain areas at 6 h post-LPS. These results provide the first evidence in favor of a functional dissociation between the brain structures that underlie cytokine-induced sickness behavior and cytokine-induced depressive-like behavior, and provide important cues about the neuroanatomical brain circuits through which cytokines could have an impact on affect.

Introduction

Several features of cytokines-induced sickness behavior overlap with the clinical symptoms of depression (for reviews: Dantzer et al., 1999; Yirmiya, 2000; Leonard, 2001; Castanon et al., 2002; Matthews et al., 2004). These points of convergence include mood and cognitive alterations, behavioral changes such as decreased motor activity, attenuated sensitivity to reward and reduced food intake, alterations in brain monoaminergic neurotransmission (Hirschfeld, 2000) and activation of the hypothalamic–pituitary–adrenal axis (Checkley, 1996). However, there is only partial overlap between cytokine-induced sickness behavior and depression as revealed by longitudinal clinical studies on patients treated by cytokine immunotherapy and showing a chronic activation of the immune system. Whereas all patients responded to first injections of cytokine immunotherapy by sickness symptoms, only one-third of them developed major depressive episodes after a few days or several weeks of treatment depending on the cytokines (Capuron et al., 2000, Capuron et al., 2001). Furthermore, pretreatment with the selective serotonin reuptake inhibitor paroxetine had no effect on neurovegetative symptoms such as fatigue and anorexia experienced by patients receiving cytokine immunotherapy (Capuron et al., 2002a) whereas it prevented the occurrence of anhedonia and depressed mood. Taken together, these data indicate that there is a functional dissociation between cytokine-induced sickness behavior and depression.

The mechanisms of the behavioral action of cytokines are usually studied in animals injected with the cytokine inducer lipopolysaccharide (LPS) or with recombinant proinflammatory cytokines (Dantzer, 2001; Konsman et al., 2002). Attempts to assess the neurobiological bases of cytokine-induced depression in animal models have concentrated on LPS or cytokine-induced anhedonia, as assessed by decreased preference for sucrose (De La Garza, 2005), reduced response to rewarding brain stimulation (Anisman et al., 2002) and attenuation of cocaine-induced place preference (Suzuki et al., 1994). Proinflammatory cytokines have also been shown to be able to increase the duration of immobility in the forced swim test (FST, Makino et al., 2000; Yamano et al., 2000; Dunn and Swiergiel, 2005). However, the depressive-like behavior induced by administration of LPS or specific cytokines is confounded in most of these experiments by the profound reduction in motor activity that occurs during the hours following treatment. Based on the clinical evidence showing that depression develops later than sickness in cytokine-treated patients, we decided in the present experiment to assess depressive-like behavior induced by cytokines at a time at which animals had recovered from LPS-induced depression of motor activity. Based on our previous extensive studies on LPS-induced sickness behavior (Dantzer, 2001), we hypothesized that sickness should be maximal at 6 h post-LPS and minimal at 24 h post-LPS whereas LPS-induced depressive-like behavior should still be apparent at 24 h post-LPS. We tested this hypothesis by submitting mice to well-accepted tests of depressive-like behavior including the tail suspension test (TST), the FST and the sucrose consumption test (Cryan et al., 2002, Cryan et al., 2005). At the neuroanatomical level, activation of the peripheral innate immune system has already been shown to induce the expression of c-Fos in a rapid and transient manner in a number of brain areas that have been proposed to underlie the different behavioral and metabolic components of cytokine-induced sickness behavior (e.g., Hansen et al., 2000a; Konsman et al., 2000). To determine the neuroanatomical substrates of LPS-induced depressive-like behavior, we decided to map the brain distribution of Delta FosB in addition to c-Fos since this truncated splice variant of FosB is characterized by a long half-life leading to its accumulation during repeated or long-lasting stimulation (Nestler et al., 2001). Our results show a partial dissociation in the neural correlates of LPS-induced sickness behavior and depressive-like behavior.