From hippocampus to dorsal horn: The pervasive impact of IL-1 on learning and memory spans the length of the neuroaxis

Abstract

In 1988, Benjamin Hart published a highly influential review arguing that behavioral manifestations of acute illness were recuperative in nature and did not represent a sign of behavioral debilitation (Hart, 1988). Since that time, the expression of sickness behaviors has come to be viewed as a goal-directed (i.e., motivational) process that can be suspended when reproductive fitness is at stake (Dantzer, 2004). Moreover, advances in biochemical and molecular techniques have led to exponential growth in the detection of soluble factors (cytokines, chemokines, prostanoids, etc) that play an identifiable role in cross-talk among immune cells, endocrine systems and the CNS. Today, it is largely accepted that soluble factors more traditionally associated with peripheral immune function can serve as signaling factors within the CNS and that inflammatory signaling pathways modulate a multitude of CNS-mediated events. Of the many cytokines and immune-related factors that have been isolated, Interleukin-1 (IL-1) repeatedly emerges as a fundamental factor seemingly involved in all facets of neuroimmune and neuroinflammatory processes, ranging from fever (Kluger, 1991) to sickness behavior (Kent et al., 1992) to neuropathic pain (Watkins et al., 2001) and even consequences of stressor exposure (Deak et al., 2005). The impact of cytokines on CNS functioning is no longer restricted to affective and motivational processes, but has naturally extended into the realm of cognition, with particular focus on learning and memory processes. Indeed, immune activation by LPS has been shown to interfere with spatial learning in the Morris water maze (Oitzl et al., 1993) as well as contextual (but not auditory-cued) fear conditioning (Pugh et al., 2001). These data guided mechanistic studies towards the hypothesis that increased IL-1 in the hippocampus produced during immune challenge can impair learning that relies largely on this structure for its expression. This hypothesis has been well-supported by empirical data at multiple levels of analysis, including findings that microinjection of IL-1 into the hippocampus produces impairments similar to immune activation (Barrientos et al., 2002) and retards the development of long-term potentiation (Lynch, 1998). The manuscript by Young et al. (2007) extends the influence of immune activation on learning and memory to the spinal cord, and in doing so, binds together several research areas within the broader fields of psychoneuroimmunology, traumatic CNS injury and neuropathic pain. The authors employed a well-established spinal learning paradigmwhere spinalized rats were given the opportunity to terminate shocks by performing an instrumental (leg flexion) response. Spinalized rats readily learn this contingency and there is a growing body of literature examining the mechanisms of spinal plasticity that underly this spinal learning phenomenon (Grau et al., 2006). Interestingly, prior exposure to noncontingent shock produces a learning deficit when rats are later tested in a contingent shock paradigm, and rats can be protected against the ill-effects of noncontingent shock by prior contingent shock exposure. Double-speak aside, what this means is that prior experience with controllable shock produces a veritable resilience against the maladaptive consequences of uncontrollable shock. Indeed, these effects were borne out of the classic literature with controllable and uncontrollable stress and are reminiscent of ‘learned helplessness’ and ‘behavioral immunization’, respectively (Maier and Watkins, 2005). Recall, however, that these effects are observed in rats with full spinal transections, suggesting that spinal learning may rival the complexity that is typi-