Lipopolysaccharide-induced neuroinflammation induces presynaptic disruption through a direct action on brain tissue involving microglia-derived interleukin 1 beta.

backgroundSystemic inflammation has been linked to synapse loss and cognitive decline in human patients and animal models. A role for microglial release of pro-inflammatory cytokines has been proposed based on in vivo and primary culture studies. However, mechanisms are hard to study in vivo as specific microglial ablation is challenging and the extracellular fluid cannot be sampled without invasive methods. Primary cultures have different limitations as the intricate multicellular architecture in the brain is not fully reproduced. It is essential to confirm proposed brain-specific mechanisms of inflammatory synapse loss directly in brain tissue. Organotypic hippocampal slice cultures (OHSCs) retain much of the in vivo neuronal architecture, synaptic connections and diversity of cell types whilst providing convenient access to manipulate and sample the culture medium and observe cellular reactions.

methodsOHSCs were generated from P6-P9 C57BL/6 mice. Inflammation was induced via addition of lipopolysaccharide (LPS), and cultures were analysed for changes in synaptic proteins, gene expression and protein secretion. Microglia were selectively depleted using clodronate, and the effect of IL1β was assessed using a specific neutralising monoclonal antibody.

resultsLPS treatment induced loss of the presynaptic protein synaptophysin without altering PSD95 or Aβ protein levels. Depletion of microglia prior to LPS application prevented the loss of synaptophysin, whilst microglia depletion after the inflammatory insult was partially effective, although less so than pre-emptive treatment, indicating a time-critical window in which microglia can induce synaptic damage. IL1β protein and mRNA were increased after LPS addition, with these effects also prevented by microglia depletion. Direct application of IL1β to OHSCs resulted in synaptophysin loss whilst pre-treatment with IL1β neutralising antibody prior to LPS addition prevented a significant loss of synaptophysin but may also impact basal synaptic levels.

conclusionsThe loss of synaptophysin in this system confirms LPS can act directly within brain tissue to disrupt synapses, and we show that microglia are the relevant cellular target when all major CNS cell types are present. By overcoming limitations of primary culture and in vivo work, our study strengthens the evidence for a key role of microglia-derived IL1β in synaptic dysfunction after inflammatory insult.