CORDIS - EU research results

Molecular signals for synaptic pruning by microglia

Final Report Summary - SYNSIGNAL (Molecular signals for synaptic pruning by microglia)

Development of mammalian nervous system is associated with generation of excess neuronal connections that is followed by their removal. In the primate brain 70% of connections are lost within the first six months of life. Aberrant synaptic pruning may lead to abnormal synaptic densities and dysfunctional connectivity in the brain, thus causing neurodevelopmental and neuropsychiatric diseases, such as autism spectrum disorders or schizophrenia. So what is the mechanism that drives this massive synapse loss, known as synaptic pruning? Recently the role of synapse elimination has been assigned to brain immune cells, called microglia. Microglia are the guardians of the brain that sense any damage or misbalance and restore homeostasis by removing any unnecessary cells, their connections or debris. This process is called phagocytosis, and it is driven by specific molecules that attract microglia to the required site of the brain, control the recognition of phagocytic target and promote the engulfment. There is accumulating evidence on attractant molecules and microglial receptors that are required for phagocytic elimination of synapses. However, no neuronal molecule has been identified that allows discrimination between strong synapses that need to be maintained and weak synapses that need to be removed. Thus the objective of this project was to evaluate how two types of molecules – “eat-me” signal phosphatidylserine and “don’t-eat-me” signal sialic acid – control synaptic pruning in developing brain.

To facilitate the investigation of developmental elimination of synapses we have first developed an ex vivo system of organotypic hippocampal slice culture. We evaluated different features of slice culture and found that multiple characteristics of developing slices resemble those of in vivo. Both synaptogenesis and the morphology of microglial cells were comparable in slices and in developing hippocampus. Furthermore, in the slices prepared from mouse lines with impaired synaptic pruning, we observed similar deficits in microglial phagocytic activity. Having confirmed that organotypic slice culture can be used to study microglia-synapse interactions, we gained access to an invaluable tool to investigate cellular and molecular mechanisms of microglia-mediated synapse elimination in such detail that was not possible in vivo.

We employed organotypic cultures to investigate “eat-me” and “don’t-eat-me” signals using highly specific click chemistry technology to label sialic acid and phosphatidylserine and to observe their dynamics during synaptic pruning. Click chemistry-based metabolic labeling of newly synthesized sialic acid revealed that de novo sialyation is targeted to presynaptic and postsynaptic structures, indicating the role of sialic acid in the maintenance of synapses. In addition, de novo sialyation and sialic acid turnover was altered in organotypic slices from mouse line with impaired synaptic pruning. To assess the role of sialyation in synaptic pruning in vivo, we evaluated synapses in developing hippocampus of mice that lack sialidase – an enzyme required to eliminate sialic acid residues. We observed both presynaptic and postsynaptic changes in mice lacking sialidase compared to wild-type controls. Thus we have demonstrated that sialic acid is involved in molecular regulation of developmental synaptic pruning.

To explore the role of phosphatidylserine in synapse elimination we developed a new chemical tool which enabled us to label live slices and observe phosphatidylserine dynamics without interfering with its signaling. We observed increased phosphatidylserine exposure on synaptic structures compared to shafts, suggesting its role in synaptic turnover. Furthermore, we evaluated synapses in hippocampus from mice lacking scramblase – an enzyme that exposes phosphatidylserine on the surface of the cells. We found that synaptic density and morphology was affected in scramblase deficient mice compared to wild-type controls, indicating that exposure of phosphatidylserine is essential for elimination of unnecessary synapses.

Altogether our findings provided first evidence that molecular communication between neurons and synapses requires neuronal presentation of “eat-me” and “don’t-eat-me” signals. Identification of sialic acid and phosphatidylserine as signals mediating the selection of synapses to be either eliminated or maintained gives us a way to better understand and manipulate synaptic pruning. The study performed during this project shed the light onto cellular and molecular mechanisms by which microglia communicate to neurons to refine circuits and to contribute to synaptic plasticity. It helps us to understand how different types of cells in the brain interact with each other and how this communication can go wrong. Since aberrant of impaired synaptic pruning may be involved in the pathology of various neurodevelopmental illnesses, such as autism or schizophrenia, uncovering the mechanisms that mediate circuit formation and refinement are crucial for understanding these disorders and will help to develop new approaches to their treatment.