Interleukin 1 receptor antagonist knockout mice show enhanced microglial activation and neuronal damage induced by intracerebroventricular infusion of human beta-amyloid

Abstract

Background: Interleukin 1 (IL-1) is a key mediator of immune responses in health and disease. Although classically the function of IL-1 has been studied in the systemic immune system, research in the past decade has revealed analogous roles in the CNS where the cytokine can contribute to the neuroinflammation and neuropathology seen in a number of neurodegenerative diseases. In Alzheimer's disease (AD), for example, pre-clinical and clinical studies have implicated IL-1 in the progression of a pathologic, glia-mediated pro-inflammatory state in the CNS. The glia-driven neuroinflammation can lead to neuronal damage, which, in turn, stimulates further glia activation, potentially propagating a detrimental cycle that contributes to progression of pathology. A prediction of this neuroinflammation hypothesis is that increased IL-1 signaling in vivo would correlate with increased severity of AD-relevant neuroinflammation and neuronal damage.

Methods: To test the hypothesis that increased IL-1 signaling predisposes animals to beta-amyloid (Abeta)-induced damage, we used IL-1 receptor antagonist Knock-Out (IL1raKO) and wild-type (WT) littermate mice in a model that involves intracerebroventricular infusion of human oligomeric Abeta1-42. This model mimics many features of AD, including robust neuroinflammation, Abeta plaques, synaptic damage and neuronal loss in the hippocampus. IL1raKO and WT mice were infused with Abeta for 28 days, sacrificed at 42 days, and hippocampal endpoints analyzed.

Results: IL1raKO mice showed increased vulnerability to Abeta-induced neuropathology relative to their WT counterparts. Specifically, IL1raKO mice exhibited increased mortality, enhanced microglial activation and neuroinflammation, and more pronounced loss of synaptic markers. Interestingly, Abeta-induced astrocyte responses were not significantly different between WT and IL1raKO mice, suggesting that enhanced IL-1 signaling predominately affects microglia.

Conclusion: Our data are consistent with the neuroinflammation hypothesis whereby increased IL-1 signaling in AD enhances glia activation and leads to an augmented neuroinflammatory process that increases the severity of neuropathologic sequelae.

Figure 1

Increased mortality in IL1raKO mice during Aβ infusion. Alzet pumps containing Aβ1–42 or vehicle were surgically implanted in IL1raKO and WT littermate mice (n = 10–12 mice per Aβ-infused group; n = 5 mice per vehicle-infused group), and post-operative survival was monitored for 42 days. Kaplan-Meier survival curves show that WT mice infused with vehicle or Aβ, and IL1raKO mice infused with vehicle experienced no mortality during the time period. In contrast, Aβ-infused IL1raKO mice experienced a 50% mortality rate (6 of the 12 animals died before 42 days). This mortality was significantly different from the other experimental and control groups (error bars = SEM; p < 0.05).

Figure 2

Microglia activation following Aβ infusion. WT and IL1raKO mice infused with Aβ or vehicle for 28 days were sacrificed on day 42 (n = 5–10 mice/group survived for analysis). Brains were bisected and the right side of the brain was processed for immunohistochemistry while the left hippocampus was dissected and used for biochemical analysis. A) Levels of the pro-inflammatory cytokine IL-1β were significantly higher in IL1raKO mice infused with Aβ compared to WT mice infused with Aβ. B) TNFα levels also showed a stronger upregulation in Aβ-infused IL1raKO mice compared to Aβ-infused WT mice. C) F4/80 immunostaining for activated microglia also revealed a significant increase in IL1raKO mice infused with Aβ versus WT mice infused with Aβ. Representative photomicrographs of F4/80-positive microglia in the hippocampus of a D) WT mouse infused with Aβ, and E) IL1raKO mouse infused with Aβ. Arrowheads point to microglia cell bodies. Bar in D-E = 50 μm (error bars = SEM; * Significantly different, p < 0.01; ***Significantly different, p < 0.001).

Figure 3

Astrocyte activation following Aβ infusion. WT and IL1raKO mice were infused with Aβ or vehicle, and brains prepared as in Figure 2. A) Levels of the pro-inflammatory astrocyte-derived cytokine S100B showed a similar degree of upregulation in Aβ-infused IL1raKO and WT mice. B) Numbers of GFAP-positive astrocytes were increased to a similar degree in both WT and IL1raKO mice infused with Aβ. (error bars = SEM; p > 0.05 between Aβ-infused IL1raKO and WT mice).

Figure 4

Loss of synaptic markers following Aβ infusion. WT and IL1raKO mice were infused with Aβ or vehicle, and brains prepared as in Figure 2. A) Aβ-infused mice had reduced hippocampal PSD-95 levels compared to vehicle-infused mice, and there was a significantly larger decrease in IL1raKO mice infused with Aβ versus their WT counterparts (error bars = SEM; * p < 0.05). B) The presynaptic marker synaptophysin was reduced in Aβ-infused mice compared to the vehicle-infused mice. In addition, the reduction in synaptophysin in Aβ-infused mice was greater in IL1raKO mice compared to WT mice, although the difference did not quite reach statistical significance (error bars = SEM; p = 0.07).