Protecting connections from synapse elimination

Abstract

A recent paper by Cong et al. [1] provides exciting evidence that neurons contain proteins that protect synapses from complement-mediated synapse elimination. SRPX2 binds C1q and blocks microglial synapse engulfment. The findings point at SRPX2, and potentially other related sushi domain proteins, as possible targets for therapies for neurodevelopmental and neurodegenerative disorders.

Length:

The complement system is a critical part of the immune system, where it enhances the removal of infectious agents and damaged cells [2]. In recent years, a large body of literature has shown that complement also plays important roles in the brain, including the removal of connections between neurons during the process of synapse elimination [3]. Synapse elimination is essential for brain development and has been heavily studied in regions such as the lateral geniculate nucleus (LGN), that require activity-dependent refinement for proper development. Typically, neurons overproduce synapses in the early stages of brain development and only relevant connections remain after the pruning process, which is in part mediated by complement signaling. In the absence of complement signaling through components including C1q and C3, brain areas show aberrant elevations in synapses and dendritic spines [4] due to a lack of microglial engulfment of synapses [5].

Despite the prevalent role that complement-mediated synapse elimination plays in neural development, there are many mysteries that surround this process. One of the central questions that has dogged the field is how the typically widespread activation of complement can direct microglial synapse engulfment with the exquisite specificity that is so essential for proper neural development. Although several complement inhibitors have been found in the brain, a recent study by Cong et al. 2020 [1] identifies, for the first time, a complement inhibitor expressed on neurons (SRPX2, an orphan sushi domain protein) that protects synapses against complement-mediated synapse elimination. This study is a technical tour-de-force that not only provides data supporting a direct interaction and an epistatic relationship between complement and SRPX2, but also shows the functional effects of loss of SRPX2, molecularly, cell biologically, and electrophysiologically in two brain regions and at multiple types of synapses.

The study by Cong et al. [1] builds upon the observation that SRPX2 contains domains found in complement inhibitors and previous work showing that SRPX2 promotes synapse density [6]. In the recent paper, the authors used a newly generated knock-in mouse to identify tagged SRPX2 interactors. C1q was identified in this targeted screen, and its interaction with SRPX2 was confirmed via in vitro experiments. In vivo, SRPX2 colocalized primarily in neurons with C1q punctae, and to a lesser extent with the presynaptic proteins, VGLUT1/2. Importantly, SRPX2-null mice showed an increase in complement signaling, as measured through C3 expression and C2 cleavage. Functionally, the loss of SRPX2 reduced inputs into dLGN neurons and increased early synapse pruning in this brain area, an effect which is reversed and, more importantly, occluded by the depletion of C3. Consistent with the reduction in synapse elimination, the number of synaptic engulfments by neighboring microglia were significantly increased by the loss of SRPX2 and occluded by loss of C3, thus indicating its role in synapse elimination through the complement signaling pathway. The use of the C3-null animals provides the primary evidence for the necessity of complement signaling for SRPX2 to affect synapse elimination.

The in vivo effects of SRPX2 exhibited remarkable specificity. They were not only brain region-specific, but also temporally-specific and cortical layer- and synapse-type-specific. In the LGN, SRPX2 protected retinogeniculate synapses from C3-mediated synapse elimination only during a transient early postnatal time window. In contrast, in the somatosensory cortex, SRPX2 inhibited complement and microglial activation in layer 4, but not in layer 2/3. Moreover, SRPX2 deletion caused dendrites from layer 2/3 and layer 5 neurons passing through layer 4 to have reduced dendritic spine density. In contrast to the effects in LGN, these changes in cortical dendritic spines lasted well into adulthood. Finally, SRPX2 protected only a subset of specific types of synapses in layer 4 of somatosensory cortex. In layer 4, complement-mediated synapse elimination affected thalamocortical (VGlut2/PSD-95) and inhibitory (VGAT/gephyrin) synapses, but not cortico-cortical synapses (vGlut1/PSD-95). Yet, SRPX2 only protects the thalamocortical synapses since VGAT/gephyrin colocalization was elevated in the C3-null mouse line but not the SPRX2-null mice, indicating that SRPX2 may function selectively on VGLUT2-containing synapses.

Altogether, these insights not only indicate an important and novel role for SPRX2 in complement-dependent synapse elimination, but also raise many important questions for future work. First, are there additional regulatory proteins expressed in the nervous system that will play similar roles, but on distinct sets of synapses in specific brain regions? Second, are there additional inhibitors or activators expressed in the brain to titrate synapse elimination in response to activity and development? Third, what regulates expression of these complement inhibitors and could those regulators be targets to rescue disease-related synaptic dysregulation? Finally, how is SRPX2 secreted and how does it remain sufficiently local to protect specific types of synapses? Especially perplexing in this regard is how SRPX2 could selectively protect a subset among intermingled synapses.

Perhaps most exciting, these data suggest that complement inhibitors like SRPX2 may prove to be therapeutic agents for brain disorders. Complement signaling has been implicated in neurodegenerative diseases such as Alzheimer’s disease [7] and neurodevelopmental disorders such as schizophrenia [8] among others. The effectiveness of the general approach of using complement inhibitors to prevent synapse elimination in animal models of disease has been recently demonstrated using several approaches to increase expression of the complement inhibitor Crry, which protected visual function in a mouse model of multiple sclerosis [9] and inhibited neuroinflammation and removal of stressed neurons after stroke in mice [10]. It remains to be examined whether similar approaches using SRPX2 and/or other complement inhibitors as therapeutic agents may offer paths for treatment (or prevention) of neurological and psychiatric diseases.