The Effects of Neuroinflammation and Alzheimer's-like pathology on Gamma Frequency Oscillations in Mice

Abstract

Gamma frequency (20-80 Hz) oscillations are rhythmic neuronal network activities which are required in many cognitive processes, including working memory and selective attention. Neuroinflammation is increasingly seen as an important factor in neurological and psychiatric disorders, however there is a significant lack of understanding as to how neuroinflammation impacts gamma frequency oscillations in health and disease. Therefore, using ex vivo brain slice electrophysiology the effect of neuroinflammation on gamma frequency oscillations was investigated in healthy mice and in APP/PS1 mice (an Alzheimer's disease mouse model). In young healthy mice exposure to various inflammatory mediators (TNFα, IL-1β, CCL2, CXCL10, and PGE-2) inhibited the normal growth in oscillation amplitude over time and in some instances reduced the oscillation frequency. Therefore, in healthy brain slices, gamma frequency oscillations are sensitive to increased neuroinflammation. In brain slices from APP/PS1 mice, which carry two mutations that cause early onset Alzheimer's disease, gamma frequency oscillations were found to be disrupted, specifically at 9-11 months the oscillation amplitude and frequency were reduced relative to oscillations in brain slices from age- matched wildtype littermate controls. Additionally, the incidence and length of epileptiform events were increased in APP/PS1 mice. These changes were likely due to the deficits observed in the neuronal networks responsible for generating gamma frequency oscillations in APP/PS1 mice. Specifically, in APP/PS1 mice parvalbumin and somatostatin inhibitory interneurons were shown to be reduced using immunohistochemistry and GABAergic transmission was shown to be altered by pharmacological manipulation. Given the immune system’s active role in Alzheimer's disease and the observed sensitivity of gamma frequency oscillations to neuroinflammation, immune effects on gamma frequency oscillations was investigated in control and APP/PS1 mice. The extent to which the observed alterations in APP/PS1 mice were due to increased neuroinflammation were investigated by treating mice with a microglial proliferation inhibitor, GW2580, which acts via CSF1R inhibition. This treatment was found to reduce microglial proliferation and reactivity, as intended. Gamma frequency oscillations were partially restored in APP/PS1 mice following GW2580 treatment – the amplitude reduction was reversed, however the incidence of epileptiform activity was increased. This demonstrated that the observed alterations in gamma frequency oscillations in APP/PS1 mice were, to at least some extent, caused by increased neuroinflammation in APP/PS1 mice. Further investigation of neuroinflammatory influences on gamma frequency oscillations in APP/PS1 mice, showed that the inhibitory effect of TNFα and IL-1β on the normal growth of gamma frequency oscillations, previously observed in normal, healthy mice, was not observed in APP/PS1 mice. Although this finding is not intuitive, the data are suggestive of further gamma frequency oscillation disruption in these mice. Overall, microglia and pro-inflammatory cytokines and chemokines were shown to cause significant alterations to gamma frequency oscillations in brain slices from both healthy control mice and APP/PS1 mice. Delineating the disruptive and beneficial aspects of these alterations will be essential if anti-inflammatory approaches to Alzheimer's disease move towards clinical trials. Therefore, further investigation of neuroinflammatory influences on gamma frequency oscillations is required.