Neurology Central

Journal watch: Amyloid, microglia and synapse loss in Alzheimer’s – a complement-ary pathology

A novel study has indicated that beta-amyloid (Aβ) erroneously activates the complement cascade, inducing engulfment of synapses by microglia. Understanding this mechanism raises the possibility of novel therapeutic targets for treating early Alzheimer’s disease (AD).

Soyon Hong and Beth Stevens, both of Boston Children’s Hospital (MA, USA), have recently demonstrated the mechanism of synapse loss in early AD mouse models, prior to the accumulation of amyloid plaques. The study, published in Science, demonstrates that the complement cascade is pivotal in Aβ-induced microglia engulfment of synapses in the hippocampus. Targeting the complement system for therapeutic intervention may therefore prevent further synapse loss and thus halt the progressive cognitive impairment characteristic of AD.

Synapse loss and cognitive decline in neurodegeneration are strongly correlated, with the extent and prognosis of this decline being determined by the rapidity and localization of the synapse loss. In AD for example, synapse loss is primarily constrained to the medial temporal lobe manifesting the characteristic memory, navigation and judgement deficits. However, synapse loss is not limited to AD and is evident in a host of neurodegenerative diseases including prion disease, frontotemporal dementia, vascular and Lewy body dementias.

Stevens’ group adopted a novel, and superficially counterintuitive, approach to studying neurodegeneration by investigating neurodevelopment: “Understanding a normal developmental process (physiological pruning during maturation) has deeply provided us with novel insight into how to protect synapses in Alzheimer’s and potentially a host of other diseases”.

In the maturing brain, neuronal pruning is induced by superfluous neurons by expressing gylcoproteins such as C1q and C3. These proteins are essential components of the immune system, responsible for triggering and activating the complement pathways initiating the phagocytosis and engulfment of neurons en masse.

The complement system is comprised of many pathways to activation of the membrane attack complex but the two of interest to Stevens and colleagues were the ‘classical’ and  ‘alternative’ pathway initiated by the glycoprotein C1q. Typically the classical complement pathway is initiated in a specific immune response requiring antigen–antibody complexes for activation, whereas the alternative pathway can be activated by hydrolysis, pathogens, or damaged cells.

The authors investigated whether the initiating proteins of the complement system are erroneously activated by soluble Aβ oligomers and how inhibition of these proteins ceases the microglial-mediated pathological pruning of synapses.

Synapse loss was investigated in two transgenic mouse models both overexpressing human amyloid precursor protein (APP), in the form of J20 and APP/PS1 models, which typically express an age-dependent increase in Aβ evident throughout the hippocampus.

Crucially, Aβ plaque deposition is only apparent in the hippocampus from 3 months in APP/PS1 mice and 5 months in J20 mice. The authors demonstrated that by 3 months the J20 mouse exhibited heightened synaptic localization of C1q and both transgenic strains exhibited significant synapse loss and increased levels of C1q immunoreactivity in the hippocampus prior to plaque deposition.

The observation that C1q tags synapses for microglial-mediated elimination prior to plaque deposition supports the growing evidence that AD pathology accumulates earlier than available diagnostic measures can detect.

Stevens and colleagues further investigated the finding byfocusing on whether this immune reaction was robust or a confound limited to APP transgenic strains. Interestingly, injection of oligomeric Aβ into the lateral ventricles of wild-type mice increased synapse loss and C1q deposition in the hippocampus, this effect was inhibited in mice receiving an injection of anti-C1q antibodies.

Furthermore, synapse loss was not observed in C1q knockout mice injected with oligomeric Aβ. This double dissociation indicates that oligomeric Aβ is sufficient to induce C1q deposition and mediate its microglial-induced synaptotoxicity in the hippocampus.

The role of C1q in brain development is to initiate the ‘alternative’ complement cascade via activation of C3, which, unlike other complement pathways, does not require pathogen-binding antibodies. C3 has many subtypes generated by proteolytic cleavage of C3; these subtypes are optimized for their immunological functions including phagocytosis, chemotaxis, degranulation and opsonization.

The authors demonstrated that injecting oligomeric Aβ induced synaptic deposition of C3. Furthermore, the offspring bred from crossing APP/PS1 mice with C3 knockouts rescued this synapse loss.

Mice also exhibited augmented synaptic engulfment in response to oligomeric Aβ injections, this effect was rescued in C3 receptor knockout mice. Cumulatively, these results suggest that the presence of Aβ induces the classical complement cascade resulting in phagocytosis of synapses by microglia.

This landmark study highlights that the diagnostic timeline of AD must advance in order that clinical trials can be initiated when there is function left to salvage. The work of Hong and colleagues challenges the notion that the pathological function of microglia and the complement system are a secondary event facilitated by plaque-induced neuroinflammation.

Complement and microglia are evidently involved much earlier in the disease process than previously thought. This research necessitates the need for biomarkers measuring synaptotoxicity and gives optimism for ANX-005, a human form of the C1q antibody, currently being developed with therapeutic resonance in a host of neurodegenerative diseases.

Source: Hong, S. et al. Complement and microglia mediate early synapse loss in Alzheimer mouse models. Science DOI: 10.1126/science.aad8373 (2016); Boston Children’s Hospital press release