Sustained hippocampal IL-1β overexpression impairs contextual and spatial memory in transgenic mice

Abstract

Neuroinflammatory conditions such as traumatic brain injury, aging, Alzheimer’s disease, and Down syndrome are often associated with cognitive dysfunction. Much research has targeted inflammation as a causative mediator of these deficits, although the diverse cellular and molecular changes that accompany these disorders obscure the link between inflammation and impaired memory. Therefore, we used a transgenic mouse model with a dormant human IL-1β excisional activation transgene to direct overexpression of IL-1β with temporal and regional control. Two weeks of hippocampal IL-1β overexpression impaired long-term contextual and spatial memory in both male and female mice, while hippocampal-independent and short-term memory remained intact. Human IL-1β overexpression activated glia, elevated murine IL-1β protein and PGE₂ levels, and increased pro-inflammatory cytokine and chemokine mRNAs specifically within the hippocampus, while having no detectable effect on inflammatory mRNAs in the liver. Sustained neuroinflammation also reduced basal and conditioning-induced levels of the plasticity-related gene Arc.

Introduction

Neuroinflammatory conditions such as traumatic brain injury, aging, Alzheimer’s disease, and Down syndrome are often associated with cognitive dysfunction, including memory deficits (Casadesus et al., 2007, Griffin et al., 1989, Montine et al., 1999). Much research has focused on anti-inflammatory treatments to try and prevent these memory deficits with mixed results (ADAPT Research Group et al., 2007, Hurley et al., 2002, Rogers et al., 1993). It is clear that acute inflammation, caused by injection of lipopolysaccharide or interleukin-1β (IL-1β) itself, impairs memory (Gibertini et al., 1995, Hein et al., 2007, Oitzl et al., 1993, Pugh et al., 1998, Pugh et al., 2001, Thomson and Sutherland, 2005). However, a lack of experimental models has made the study of the effects of chronic neuroinflammation on memory difficult, especially when trying to parse the potential influences of resulting sickness or neurodegeneration on memory. Therefore, we utilized a recently developed mouse model, which can maintain sustained IL-1β overexpression with temporal and regional control, to study the effects of prolonged IL-1β overexpression and ensuing neuroinflammation on learning and memory processes.

This mouse model retains a dormant human IL-1β excisional activation transgene (IL-1β^XAT). When activated by microinjection of a virus expressing Cre, astrocytes local to the injection site express human IL-1β (Shaftel et al., 2007b). Human IL-1β binds to the murine IL-1 receptor and signals downstream pathways to induce inflammation. Research in this model has shown that a single microinjection of virus expressing Cre into the hippocampus causes sustained IL-1β overexpression for up to nearly a year after injection (Shaftel et al., 2007b). Within 2 weeks of transgene activation, IL-1β overexpression within the hippocampus leads to glial activation, elevated cytokines, elevated chemokines, leukocyte infiltration, and increased vascular permeability (Moore et al., 2009, Shaftel et al., 2007a, Shaftel et al., 2007b). However, after 2 weeks or 2 months of IL-1β overexpression, no overt loss of neurons or neuronal integrity is apparent within the hippocampus as measured by apoptotic stains and neuronal, synaptophysin, and acetylcholinesterase labeling (Moore et al., 2009, Shaftel et al., 2007a). Therefore, this model allows us to study the role of chronic IL-1β driven neuroinflammation on learning and memory processes.

Initial behavioral studies in this mouse model revealed spatial learning deficits in the Morris water maze 2 weeks following transgene activation (Moore et al., 2009). Here we further characterize these deficits by testing hippocampal-dependent and -independent, as well as short- and long-term, fear memory and potential gender differences in fear conditioning and the Morris water maze. We also further characterize the central and peripheral inflammatory responses to sustained hippocampal IL-1β overexpression and find changes in one plasticity-related gene, activity-regulated cytoskeleton-associated protein (Arc; also Arg 3.1).