Complement proteins may be helpful when it comes to rallying clearance of hazardous debris from the brain, but scientists believe the immune activators can also inflict irreparable damage to neurons and their synapses. A paper published August 20 in Cell Reports provides support for these ideas. Researchers led by Jesse Hanson of Genentech in South San Francisco reported that in an amyloidosis model, knocking out the complement protein C3 spared dendritic spines but stoked the development of plaques and dystrophic neurites. In a tauopathy model, nixing C3 saved synapses, assuaged neurodegeneration, and even rescued some behavioral abnormalities in male mice. Complement protein levels were also higher in the cerebrospinal fluid of AD patients, and correlated with CSF tau. The findings add to previous evidence suggesting that when chronically activated, the complement system does the brain more harm than good.

  • Knocking out C3 rescued neurons and their synapses from demise in TauP301S mice.
  • Deleting C3 in PS2APP led to higher plaque load, but saved dendritic spines.
  • In people with AD, complement proteins rose in the CSF in concordance with tau.

“Overall, I think this is an excellent paper with many rigorous measures that, in the end, continue to suggest that lowering of C3 (or a downstream complement fragment/receptor) may be protective against neurodegeneration,” commented Cynthia Lemere of Brigham and Women’s Hospital, Boston. “Future therapies targeting these pathways, including cell-specific targeting, are gaining interest in many areas of neurodegeneration.”

The complement system has long been appreciated as a double-edged sword in neurodegenerative disease. On one hand, complement proteins bind to and facilitate the clearance of neurotoxic debris, including Aβ plaques (Jiang et al., 1994; Jun 2008 news; Jun 2017 news). On the other, they instigate inflammatory responses that wreak havoc on neurons and their synapses in models of amyloidosis and other neurodegenerative diseases (Apr 2016 news; Jul 2019 conference news). Last year, the Genentech researchers extended complement’s toll to tauopathy, reporting that in the Tau-P301S mouse model, the complement protein C1q lingered in postsynapses, where it flagged down microglia to engulf synapses (Jul 2018 conference newsDejanovic et al., 2018). 

Ditching C3 Spares Neurons. C3 crowds synapses in the AD brain (top). Knocking it out protects synapses in a mouse model of amyloidosis (center), and spares neurons in a model of tauopathy (bottom). [Courtesy of Wu et al., Cell Reports, 2019.]

For their current study, co-first authors Tiffany Wu and Borislav Dejanovic and colleagues further dissected the role of complement in the TauP301S model, and also examined how complement affects neuropathology and degeneration in the PS2APP model of amyloidosis. They started by assessing expression of complement genes in cells. In the TauP301S model, the researchers sorted astrocytes and microglia from the hippocampi of 6-month-old mice, an age when tau pathology is apparent but neurodegeneration has not yet begun. They found robust expression of classical complement genes in astrocytes, and to a lesser degree, microglia, compared  with wild-type. In PS2APP mice, the researchers sequenced RNA from sorted cells extracted from the forebrains of mice at different ages, ranging from 7 to 13 months. At all ages they tested, this complement surge in astrocytes was less robust than in the TauP301S mice, and not detectable in microglia.

Immunohistochemistry told a similar story: C1q and C3 staining lit up the brains of Tau-P301S mice, while PS2APP posted a mild uptick only in C3 (see image below).

Complements of TauP301S. C1q and C3 staining barely rise above wild-type levels in PS2APP mice (left panels), while both complement proteins are elevated in TauP301S mice (right panels). [Courtesy of Wu et al., Cell Reports, 2019.]

The researchers next knocked out C3 in each model. In PS2APP mice, a dearth of C3 nearly doubled the number of plaques and Aβ-laden dystrophic neurites. Despite this, dendritic spines near plaques were spared, while they fell by half in PS2APP controls. Though PS2APP mice had more astrocytes and microglia than did wild-type animals, C3 deficiency did not affect this gliosis.

In male Tau-P301S mice, C3 deficiency reduced neuronal death and hippocampal and cortical volume loss, which happens between 6 and 9 months of age. It also rescued synaptic deficits, and ameliorated hyperactivity. Because 9-month-old female mice had no synaptic or behavioral deficits and subtler neurodegeneration than their male counterparts, knocking out C3 had no discernable effect. Tau pathology, neuroinflammation, and neurodegeneration affect female P301S mice at an older age than males, Hanson said, but the researchers have yet to test if C3 deficiency protects them.

Does the complement cascade switch on in the AD brain? And how does that relate to tau pathology? To address these questions, the researchers measured C3 protein in postsynaptic densities isolated from the superior frontal gyri of people with AD. While people who died in the earlier stages of the disease—based on cognitive scores— had C3 levels similar to controls’, those with later-stage, pathologically confirmed AD had substantially more. Brain lysates from those with confirmed AD also had higher levels of phospho-tau than did control lysates.

Next, the authors took stock of intact C3, as well as its cleaved and activated forms (C3b, iC3b, and C3c) in CSF samples from cognitively normal people and from those with AD. Intact and processed C3 were 1.5- and 2.5-fold higher, respectively, in patients than in controls. In the CSF of both AD and controls, intact and processed C3 levels, as well as their ratios, correlated with total tau, but not with Aβ42. Hanson said they did not measure phospho-tau in the CSF samples. The researchers obtained similar results from a second cohort of AD cases and controls.

In all, the findings implicate the complement cascade in a deadly blow to neurons in response to tau pathology. Hanson said that tau pathology likely stresses neurons, which may lead C1q proteins to bind synapses. Production of other complement components, including C3, by activated glial cells, primarily astrocytes, then fully unleashes the cascade. Hanson noted that the data suggest that the complement cascade is most activated in stages of AD marked by substantial tau pathology. This suggests that targeting complement could have a therapeutic effect at later stages of the disease than an amyloid-targeted therapy would, he said.

Lars Ittner of Macquarie University in Sydney agreed that the complement system could be a therapeutic target, however, he cautioned that blocking the cascade outside of the brain could have detrimental effects, especially in an aged population that is already vulnerable to infections.

Ittner also recently reported an uptick in complement gene expression in a different strain of Tau-P301S mice (Ke et al., 2019). He found complement round neurons, but did not observe activated, complement-producing astrocytes as Hanson did. These mice, both male and female, have scant neurodegeneration, which may be due to the way the transgene is expressed, Ittner speculated. He also suggested that the lower degree of neuronal damage in these mice could explain why he did not observe activated astrocytes, as Hanson did.

Andrea Tenner of the University of California, Irvine, said the study was well done, but did not address which pathways downstream of C3 in the complement cascade are responsible for neuronal loss. Specifically, the findings leave open the possibility that the C5a component of the complement system—which ramps up harmful neuroinflammation—delivered the ultimate death blow to neurons and synapses. Tenner previously reported that in multiple AD mouse models, inhibiting the C5a receptor on microglia lessened Aβ and tau pathology, and also rescued neuronal loss and behavior abnormalities (Fonseca et al., 2009;  Ager et al., 2010Hernandez et al., 2017).

“Overall this paper is an important replication and extension of work in mouse models and human brain tissue supporting a role for the complement system in neurodegeneration related both to amyloid and tau pathology,” wrote Tara Spires-Jones, Makis Tzioras, Caitlin Davies, and Declan King of the University of Edinburgh in a joint comment to Alzforum. “Moving forward, it will be important to further unpick how and when glia are contributing to degeneration. In particular we think it is essential to understand whether microglia (and possibly also astrocytes) phagocytose live healthy synapses, sick synapses, or synaptic debris and whether human glia phagocytose synapses in the same manner as those in mice.”—Jessica Shugart


No Available Comments

Make a Comment

To make a comment you must login or register.


News Citations

  1. Complement: AD Friend or Foe? New Work Tips Balance to Former
  2. Sans Complement: Amyloid Grows, Synapses and Memory Stay
  3. Paper Alert: Microglia Mediate Synaptic Loss in Early Alzheimer’s Disease
  4. Do Microglia Finish Off Stressed Neurons Before their Time?
  5. Synaptic Tau Clangs the Dinner Bell for Hungry Microglia

Research Models Citations

  1. Tau P301S (Line PS19)
  2. PS2APP

Paper Citations

  1. . beta-Amyloid activates complement by binding to a specific region of the collagen-like domain of the C1q A chain. J Immunol. 1994 May 15;152(10):5050-9. PubMed.
  2. . Changes in the Synaptic Proteome in Tauopathy and Rescue of Tau-Induced Synapse Loss by C1q Antibodies. Neuron. 2018 Dec 19;100(6):1322-1336.e7. Epub 2018 Nov 1 PubMed.
  3. . CNS cell type-specific gene profiling of P301S tau transgenic mice identifies genes dysregulated by progressive tau accumulation. J Biol Chem. 2019 Sep 20;294(38):14149-14162. Epub 2019 Jul 31 PubMed.
  4. . Treatment with a C5aR antagonist decreases pathology and enhances behavioral performance in murine models of Alzheimer's disease. J Immunol. 2009 Jul 15;183(2):1375-83. Epub 2009 Jun 26 PubMed.
  5. . Microglial C5aR (CD88) expression correlates with amyloid-beta deposition in murine models of Alzheimer's disease. J Neurochem. 2010 Apr 1;113(2):389-401. PubMed.
  6. . Prevention of C5aR1 signaling delays microglial inflammatory polarization, favors clearance pathways and suppresses cognitive loss. Mol Neurodegener. 2017 Sep 18;12(1):66. PubMed.

Further Reading


  1. . Cell-specific deletion of C1qa identifies microglia as the dominant source of C1q in mouse brain. J Neuroinflammation. 2017 Mar 6;14(1):48. PubMed.

Primary Papers

  1. . Complement C3 Is Activated in Human AD Brain and Is Required for Neurodegeneration in Mouse Models of Amyloidosis and Tauopathy. Cell Rep. 2019 Aug 20;28(8):2111-2123.e6. PubMed.