Interleukin 1 beta enhances conditioned fear memory in rats: possible involvement of glucocorticoids Cai Song, 1 Anthony G. Phillips 1 and Brian Leonard 2 1 Department of Psychiatry, University of British Columbia, 2255 Wesbrook Mall, Vancouver, Canada, V6T 2A1 2 Brain and Behaviour Research Institute, Academic Hospital Maastricht, University of Maastricht, The Netherlands Keywords: corticosterone, IL-10, mifepristone (RU486), passive avoidance, PGE 2 Abstract Central administration of 15ng interleukin (IL)-1b in the rat signi®cantly enhanced conditioned fear memory assessed by a passive avoidance task, when retested at 24 and 48h post-training. Pain threshold was unaffected by 15ng IL-1b administration. IL-1b treatment also increased serum corticosterone. This increase in serum corticosterone was further enhanced in rats given both IL-1b andfootshock.Furthermore,theglucocorticoidreceptorantagonistmifepristoneblockedIL-1b-inducedelevationincorticosteroneand also attenuated the enhanced conditioned fear memory. Central administration of IL-1b signi®cantly increased prostaglandin E 2 and decreased the anti-in¯ammatory cytokine IL-10 release from whole blood cultures; therefore this treatment appears to be effective in inducinganin¯ammatoryresponseinboththeperipheryandthebrain.Thepresentstudycon®rmsthatIL-1b canenhanceconditioned fearmemory,aneffectwhichiscorrelatedwithchangesinglucocorticoidfunction.Thisfacilitationofdefensivebehaviourcouldre¯ect adaptive responses which may enhance survival during sickness. Introduction A growing body of evidence indicates that the immune, endocrine and neurotransmitter systems form a complex interacting network which can have signi®cant effects on behaviour and cognition (Maier & Watkins, 1998). The proin¯ammatory cytokine interleukin (IL)-1b, produced by activated macrophages in the periphery and by microglia and astrocytes in the brain (Johnstone et al., 1999), can induce an in¯ammatory response in the brain, activate the hypothalamic±pitui- tary±adrenal (HPA) axis and alter brain monoamine metabolism (Song, 2000; Hoozemans et al., 2001). Central effects of IL-1b are mediated in part by IL-1b receptors in the hippocampus and amygdala (Loddick etal., 1998). These effects include changes in sleep pattern, anhedonia, loss of libido and mental disturbance (Hickie & Lloyd, 1995; Krueger etal., 1998), which also may contribute to the cognitive changes in man and `sickness behaviour' in rodents (Dantzer et al., 1998). Intracerebroventricular (i.c.v.) injections of murine IL-1b to rats, or peripheral injection to mice, before training impairs spatial memory (Gibertini et al., 1995; Maier & Watkins, 1998). These behavioural ®ndings are consistent with the presence of IL-1b binding sites in the hippocampusandamygdala(Ericsson etal.,1995;Loddick etal.,1998). IL-1 may also be involved in different types of memory; for example, i.c.v. administration of IL-1 receptor antagonist (RA) blocks the enhancement of conditioned fear produced by inescapable shock (Maier & Watkins, 1995). Selective effects on contextual vs. cued fear con- ditioning occur following central infusion of IL-1b (10 or 20ng), with the former dependent on the dorsal hippocampus and the latter on the amygdala (Pugh et al., 1999). IL-1b also plays an important role in neuroendocrine responses to stress, in particular glucocorticoid secretion, which is through the activation of prostaglandin (PG) E 2 receptor gene expression and increase in PGE 2 synthesis (Zhang & Rivest, 1999; Ek et al., 2000). Both IL-1RA and a cyclooxygenase inhibitor can block IL-1b-induced increase in the secretion of CRF and glucocorticoid (Shintani et al., 1995; Ericsson et al., 1997). In part, disruptive effects of IL-1b on cognition may be mediated by glucocor- ticoid receptors in the hippocampus (Oitzl etal., 1998; Brabham etal., 2000). Contextual fear conditioning also depends on normal function of thehippocampusandcanbefacilitatedbyglucocorticoids(Cordero etal., 1998). This in turn raises the possibility that IL-1b may facilitate fear- related memory function while exerting disruptive effects on spatial learning and memory mediated by the hippocampal formation. Yirmiya et al. (2002) partially con®rm this hypothesis in their recent report that i.c.v. administration of human IL-1b enhanced passive avoidance in rats. The present study extended these ®ndings in several new directions. First, IL-1b was administered i.c.v. subchronically (3±4days) to better mimic the pathological condition over a course of in¯ammation or a short-term of illness. Second, pain threshold was measured to rule out the possibility that the effect of IL-1b on passive avoidance condition- ing may re¯ect increased sensitivity to noxious stimuli. Third, the role of glucocorticoids in modulating fear-related memory by IL-1b was assessed by coadministration of the glucocorticoid antagonist mife- pristone (RU486). Fourth, given the important role of PGE 2 in IL-1b- induced increase in glucocorticoid secretion, following behavioural tests, serum concentrations of corticosterone and the release of PGE 2 were measured in rats given central injections of IL-1b. Fifth, IL-10, produced by T-helper 2 cells (Th2), was also assayed because this anti- in¯ammatory cytokine antagonizes the effect of proin¯ammatory cytokines (such as IL-1b) produced by Th1 and macrophages (Lucey et al., 1996). Further justi®cation for measuring the release of IL-10 from blood culture comes from reports that an imbalance between pro- and anti-in¯ammatory cytokines is associated with several psychiatric disorders (Schwarz et al., 2001). European Journal of Neuroscience, Vol. 18, pp. 1739±1743, 2003 ß Federation of European Neuroscience Societies doi:10.1046/j.1460-9568.2003.02886.x Correspondence: Dr Cai Song, 1 Neuroscience Division, as above. E-mail: caisong@interchange.ubc.ca Received 18 March 2003, revised 11 June 2003, accepted 11 July 2003