Acute and Chronic Cellular Effects of Systemic Inflammation on the Central Nervous System

Abstract

Sepsis-associated encephalopathy represents an acute cerebral dysfunction caused by sepsis via systemic inflammation. Notably, cognitive functions of elderly patients who developed and survived sepsis show dynamic changes lasting for several years after the onset of sepsis. However, preclinical studies reported mixed effects of systemic inflammation on neuronal damage, likely due to the use of juvenile mice and inconsistent use of bacterial strains of lipopolysaccharide (LPS). Thus, we intraperitoneally injected LPS either from <em>Salmonella </em>or from <em>E. coli </em>into 14- to 16-month-old mice and compared acute and chronic neuroinflammation and cerebral damage at 7 and 63 days post-injection (dpi). We found that LPS, especially those derived from E. coli, induced neuroinflammation in old mice for more than a week. While not finding changes in hippocampal neuron density and global synaptic protein levels, we detected that both LPS serotypes led to a delayed reduction of excitatory, but not inhibitory, synaptic puncta in the CA3 subfield of the hippocampus at 63 dpi. We postulated that this local synapse loss in CA3 might increment the inhibitory tone of the Schaffer collaterals and a gradual decline of the neuronal activity in the CA1 apical dendrites. Furthermore, both C3-coated synapses and synaptic C3 were reduced in CA3. Because our analysis did not reveal the enhanced synaptic pruning by microglia in CA3 at 7 dpi and 63 dpi, these results may indicate that our analysis of synaptic pruning was too late to detect the changes in synaptic pruning.<br /> Triggering receptor expressed on myeloid cells 2 (TREM2) is a microglial receptor required for synaptic pruning without relying on the complement pathway, and its loss-of-function mutation increases the risk of sporadic Alzheimer’s disease (AD). Synapse loss via synaptic pruning has been documented in the early stage of the animal model of amyloid pathology. However, it remains unclear whether the absence of functional TREM2 proteins can modify the synapse pathology in APP/PS1 mice. To this end, using a T66M loss-of-function mutation of the <em>Trem2 </em>gene (associated with the frontotemporal dementia-like syndrome), we found that the T66M mutation reduced the Aβ plaque load but incremented the oligomer levels in the cortex of 12-month-old APP/PS1 mice. These results indicated that <em>Trem2</em>-mutated microglia failed to compact the Aβ plaques, allowing for spreading the potentially synaptotoxic oligomeric Aβ in the cortex. Yet, our data also demonstrated that the synaptic puncta density was higher if APP/PS1 mice harboured this <em>Trem2 </em>mutation. This highlighted the possibility that synaptic pruning by microglia was impeded during the disease progression.