Psychopharmacology (2001) 154:50–60 DOI 10.1007/s002130000595
O R I G I N A L I N V E S T I G AT I O N
Nathalie Castanon · Rose-Marie Bluthé Robert Dantzer
Chronic treatment with the atypical antidepressant tianeptine attenuates sickness behavior induced by peripheral but not central lipopolysaccharide and interleukin-1β in the rat Received: 9 May 2000 / Accepted: 16 September 2000 / Published online: 20 December 2000 © Springer-Verlag 2000
Abstract Rationale: The hypothesis that proinflammatory cytokines play a causative role in the pathophysiology of depression has been recently tested by studying the effect of antidepressants on production of endogenous cytokines, and on sickness behavior induced by exogenous cytokines. In this last case, however, the effect of antidepressants has been only studied on the effect of peripherally administered cytokines. Objectives: The aim of the present study was to determine whether the antidepressant tianeptine can attenuate both peripheral and central cytokine actions. Methods: Rats were injected IP with acute (10 mg/kg) or chronic (10 mg/kg, 2 times/day, 17 days) tianeptine. The effects of this treatment were assessed on the behavioral (social exploration, locomotion) and metabolic (food intake, body weight) alterations induced by peripheral or central administration of the cytokine inducer lipopolysaccharide (LPS) (250 µg/kg IP; 100 ng/rat ICV) or the prototypical proinflammatory cytokine interleukin-1 (IL-1)β (15 µg/rat IP; 90 ng/rat ICV). Results: Chronic, but not acute, treatment with tianeptine attenuated the behavioral signs of sickness behavior induced by peripheral, but not central, LPS or IL-1β. Conclusions: This work, which is the first in vivo study assessing the effect of an antidepressant on centrally induced immune activation, shows a clear dissociation between peripheral and central cytokine effects, and suggests a peripheral site of action of tianeptine. It also provides the first evidence that the protective effects of classical antidepressants on LPS-induced sickness behavior extend to an atypical antidepressant, and that the protective effect of antidepressants also applies to IL-1β. Keywords Sickness behavior · Tianeptine · LPS · IL-1β · Rat · Central versus periphery N. Castanon (✉) · R.-M. Bluthé · R. Dantzer INRA-INSERM U394, Neurobiologie Intégrative, Institut François Magendie, Rue Camille Saint-Saëns, 33077 Bordeaux Cedex, France e-mail:
[email protected] Fax: +33-5-56 98 90 29
Introduction The host response to infection is mediated by the synthesis and release of proinflammatory cytokines (mainly interleukin-1: IL-1; IL-6, tumor necrosis factor: TNFα) by activated monocytes and macrophages. This response includes non-specific symptoms such as fever, activation of the hypothalamic-pituitary-adrenocortical (HPA) axis, and drastic behavioral modifications typically characterized by depressed activity, reduction of food intake, and little or no interest in the environment (Kent et al. 1992). These changes, collectively referred to as sickness behavior, are mediated by central actions of proinflammatory cytokines, which are transiently expressed in the brain in response to peripherally released cytokines (for review, see Dantzer et al. 1998; Dantzer et al. 2000). In laboratory animals, sickness behavior can be induced either by the administration of a cytokine inducer such as lipopolysaccharide (LPS), the active fragment of endotoxin from Gram-negative bacteria, or by direct injection of proinflammatory cytokines such as IL-1 (Kent et al. 1992). Sickness behavior is accompanied by stress-like alterations of central serotoninergic (5-HT) and noradrenergic (NA) neurotransmission (for review, see Dunn et al. 1999), and increased HPA axis activation (for review, see Besedovsky and Del Rey 1996). Although cytokine-induced sickness behavior deserves to be studied on its own, as the expression of the reorganization of the organism’s priorities in face of infectious microorganisms, it has received additional attention due to its similarity to depression (Dantzer et al. 1999). At the clinical level, depressed patients exhibit mood alterations associated with a large set of behavioral modifications, including appetite and sleep disturbances, fatigue and loss of energy. These changes are often accompanied by important neurochemical alterations, affecting particularly both 5-HT and NA neurotransmission, concomitant with an HPA axis hyperactivity. It is noteworthy that most of these symptoms are also found in humans submitted to cytokine immunotherapy (McDonald et al. 1987; Capuron and Ravaud 1999a,
51
1999b). Similarly, viral infections and inflammatory diseases such as rheumatoid arthritis and atherosclerosis are also accompanied by a depressed mood and other symptoms found in major depressive episodes (for review, see Neveu and Castanon 1999). Additionally, experiments performed in laboratory rodents have shown that pretreatment with the endogenous IL-1 receptor antagonist (IL-1ra) prevents development of the behavioral deficits in a learned helplessness model of depression (Maier and Watkins 1995). In view of these clinical and experimental observations, particular attention has been paid to the immune alterations associated with depression. For decades, the neuro-hormonal changes observed in depressed patients were believed to be exclusively associated with an immunosuppression (Irwin et al. 1990). However, recent studies have shown that major depression can be accompanied by immune activation in the form of an acute phase reaction mediated by the peripheral synthesis and release of cytokines (Connor and Leonard 1998). Some authors have even gone so far as to use these findings as evidence for implication of cytokines in the causation of depressive episodes (Maes et al. 1995; Smith 1991). One way to test this hypothesis is to assess the effect of antidepressant medication on the immune alterations associated with depression, in order to determine whether antidepressants normalize immunity and/or alter the production, release, or actions of cytokines. If cytokines play a causative role in the pathophysiology of depression, then antidepressants should interfere with cytokine production and/or actions. Until now, only a few investigations have been carried out to answer this question. Most available clinical and experimental data concern the common and more widely used antidepressants, e.g. monoamine oxidase inhibitors (MAOI), tricyclic antidepressants (TCA) blocking both 5-HT and NA neurotransmission, and selective 5-HT reuptake inhibitors (SSRI) (for review Neveu and Castanon 1999). For example, in vitro incubation of human monocytes with different antidepressants markedly inhibited LPS-induced IL-1β, TNFα and IL-6 production (Xia et al. 1996). In addition, the enhanced lymphocyte proliferation and stimulated splenic cell production of IL-1 and IL-2, that was observed in rats subjected to a chronic mild stress model of depression, were reversed by chronic treatment with imipramine (Kubera et al. 1996). Some of the LPSinduced behavioral alterations observed in rodents were attenuated by chronic treatment with TCAs (Shen et al. 1999; Yirmiya 1996). Conversely, olfactory bulbectomy in rats, a classic animal model of depression (Jancsar and Leonard 1983), elicited an acute phase reaction that was attenuated by chronic treatment with several antidepressants (Song and Leonard 1995). Chronic administration of desipramine to olfactory bulbectomized rats also attenuated the IL-1β and TNFα production induced by an in vivo challenge with LPS (Connor et al. 2000). In addition to the classic antidepressants, there are some atypical molecules with demonstrated experimental and clinical therapeutic effects, but devoid of the classic
actions of antidepressants on central monoamine activities. For example, tianeptine is a unique antidepressant that increases the reuptake of 5-HT into rat brain (Fattaccini et al. 1990) and into human and rat platelets (Kato and Weitsch 1988; Ortiz et al. 1993), while paradoxically having well documented therapeutic activity compared with other antidepressants (Lôo et al. 1999). Its antidepressant efficiency has also been demonstrated in different animal models of depression (Whitton et al. 1991; Thiebot et al. 1992; Kelly and Leonard 1994). Despite a growing number of studies focusing on tianeptine’s properties (for review, see Wilde and Benfield 1995), the mechanisms underlying its therapeutic effects are still unknown, including a possible effect on neuroimmune interactions. The present study was therefore set up to examine the possibility that tianeptine alters cytokine-induced sickness behavior in a way that is related to its antidepressant properties. This hypothesis was tested through the analysis of the effect of tianeptine on LPSand IL-1β-induced behavioral changes in the social exploration test. This test, which is classically used in rodents to study central actions of cytokines (Kent et al. 1992), allows the precise quantification of changes in the duration of exploration of a juvenile rat that is temporarily introduced into the home cage of the adult rat being tested. This social stimulus normally induces a full sequence of close following and olfactory investigation that is profoundly depressed in rodents injected with LPS or recombinant proinflammatory cytokines. Behavioral effects of peripheral immune activation are mediated centrally, via the brain production of cytokines (for review, see Dantzer et al. 1998; Dantzer et al. 2000). Therefore, it was important to determine first whether tianeptine acts on the peripheral and/or central activation steps elicited by the initial immune stimulation. To answer this question, the effects of tianeptine were studied on an immune activation originating either peripherally or centrally. Since IL-1β released in response to LPS is the predominant molecular signal for the reduction in social behavior (Bluthé et al. 1992), the second major question was whether tianeptine acts preferentially on LPS-induced synthesis of cytokines including IL-1β, or whether it alters the interaction between IL-1β and its cell targets.
Materials and methods Animals and housing conditions Animals used in the present study were male Wistar rats (approximately 300 g) obtained from Charles River (Saint-Aubin-lesElbeuf, France). They were housed in groups of five until the beginning of the experiments, under a reverse 12:12 h light:dark cycle (lights on at 9:00 p.m.) to allow behavioral observations during the rat active phase. Food and water were available ad libitum, and room temperature was controlled (22±1°C). Rats were isolated 24 h before the first behavioral test in transparent cages (30×45×19 cm). All rats were handled daily for five days before the onset of the experiment to minimize stress reactions to manipulation. Juvenile male Wistar rats (21–28 days), serving as stimulus animals for behavioral observations, were housed collectively in the same condi-
52 tions as experimental subjects. This study was conducted in conformity with French legislation on animal experiments. Surgery After a week of acclimatization, rats used for central injection experiments were anesthetized with a mixture of ketamine and xylazine (61 mg and 9 mg/kg, respectively), and were equipped unilaterally with a 23-gauge stainless steel guide cannula over the lateral ventricle for intracerebroventricular (ICV) injections. Stereotaxic coordinates for the guide cannula were 1.5 mm lateral to midline, 0.6 mm posterior to bregma and 3.2 mm under the surface of the skull, with tooth bars 5 mm above the interaural line. Rats were allowed 10 days of recovery before behavioral testing. Behavioral observations All rats were given experience with test conditions and the injection procedure before the experiment to reduce reaction to novelty and to ensure stability of the behavioral baseline. Behavior was monitored via a video camera by a trained observer blind to drug treatment. Each session started with the introduction of a juvenile into the home cage of the rat being tested, for a 4-min period. Different juveniles were presented on each session. The amount of time (in seconds) spent by the rat following, grooming, or sniffing the juvenile was recorded. Social exploration was defined as the total time spent performing these behaviors. The locomotor effect of drug treatment was also evaluated by monitoring the duration of total immobility (a posture exclusively observed in LPS- or IL-1β-treated rats). At the end of each session, rats were weighed, and food and water consumption was measured by weighing the wire-tops and the bottles (except for experiment 1). As both consumption profiles were similar, only results concerning food intake are presented. Drugs Drugs used were phenol-extracted LPS from Escherichia coli (0127:B8, Sigma L 3129, St Louis, Mo., USA), rat recombinant IL-1β (biological activity: 317 IU/mg, NIBSC, Potters Bar, UK), and tianeptine sodium salt (Servier, France). LPS and tianeptine were dissolved in sterile physiological saline. Recombinant rat IL1β was prepared in sterile saline containing 0.1% endotoxin-free bovine serum albumin (BSA) (A-8806, Sigma). All compounds were administered in a volume of 2 ml/kg or 2 µl/rat for intraperitoneal (IP) and ICV injections, respectively. Fresh solutions were made on every test day. The doses of tianeptine, LPS, and IL-1β used were selected on the basis of the available literature (Bluthé et al. 1992; Delbende et al. 1994; Anforth et al. 1998) and unpublished data from our laboratory. Treatments and experimental conditions For all experiments, the duration of social exploration recorded during the habituation sessions was used to allocate rats to different experimental groups matched for mean social exploration. Tianeptine was always injected IP. Experiment 1 This experiment was set up to assess the effect of tianeptine on behavioral and metabolic changes induced by a peripheral or a central LPS injection. Peripheral LPS injection For the acute experiment, the first behavioral observation (serving as baseline value) was carried out, on the test session, just before
treatment. Each rat then received an IP injection of either saline or tianeptine (10 mg/kg), followed immediately by an IP injection of saline or LPS (250 µg/kg). All rats were then submitted to four additional test sessions conducted 2, 4, 6, and 24 h after treatment. For the chronic experiment, this procedure was applied to naive rats pretreated with tianeptine (10 mg/kg, IP, twice a day) for 17 days, days 16 and 17 corresponding to the habituation sessions. Central LPS injection Because of the duration of the ICV injection procedure, baseline social exploration was assessed on the day before the test. For the acute experiment, rats were first injected on the day of test, with saline or tianeptine (10 mg/kg), and immediately after they received an ICV injection of BSA or LPS (100 ng/rat). Their behavior was then measured 1.5, 3, 6, and 24 h after treatment. For the chronic experiment, this procedure was applied to rats pretreated with tianeptine (10 mg/kg, IP, twice a day) for 17 days. Experiment 2 This experiment was designed to assess tianeptine’s effect on the behavioral and metabolic changes induced by IL-1β. Therefore, all the procedures described for experiment 1 were repeated on naive rats, the only difference being that LPS was replaced by IL-1β at doses of 15 µg/rat and 90 ng/rat for the IP and ICV injections, respectively. Data analyses Data were submitted to an ANOVA with pretreatment (saline versus tianeptine) and treatment (saline versus LPS or IL-1β) as between-subject factors and time as a within-subject factor. Since immobility was only observed after LPS or IL-1β, it was analyzed for these two groups by a two-way ANOVA with pretreatment and time as between- and within-subject factor, respectively.
Results Experiment 1: effect of tianeptine on LPS-induced sickness behavior Acute treatment with tianeptine does not alter the behavioral effects of IP LPS but tends to attenuate the body weight loss induced by LPS As previously described, LPS significantly reduced the duration of social exploration in a time-dependent manner [treatment: [F(1,30)=190.0; P
