Microglia in Alzheimer's disease

Abstract

Proliferation and activation of microglia in the brain, concentrated around amyloid plaques, is a prominent feature of Alzheimer's disease (AD). Human genetics data point to a key role for microglia in the pathogenesis of AD. The majority of risk genes for AD are highly expressed (and many are selectively expressed) by microglia in the brain. There is mounting evidence that microglia protect against the incidence of AD, as impaired microglial activities and altered microglial responses to β-amyloid are associated with increased AD risk. On the other hand, there is also abundant evidence that activated microglia can be harmful to neurons. Microglia can mediate synapse loss by engulfment of synapses, likely via a complement-dependent mechanism; they can also exacerbate tau pathology and secrete inflammatory factors that can injure neurons directly or via activation of neurotoxic astrocytes. Gene expression profiles indicate multiple states of microglial activation in neurodegenerative disease settings, which might explain the disparate roles of microglia in the development and progression of AD pathology.

Figure 1.

Expression and function of AD risk genes in microglia. (A) These heat maps depict relative expression levels of GWAS-identified AD risk genes among CNS cell types purified from human (Zhang et al., 2016) or mouse (Srinivasan et al., 2016) brain tissues and analyzed by RNA sequencing. Each column within a cell type represents one sample of those cells purified from a different brain. From human dataset GSE73721, samples derived from “normal” cortex are plotted, ranging in age from 8 to 63 yr. From mouse dataset GSE75431, samples from cortex of 13-mo PS2APP β-amyloid model or age-matched nontransgenic littermates are plotted. Z-score represents the number of standard deviations by which a sample’s expression level for a gene differs from the mean expression level for that gene across all samples. For human genes lacking clear mouse orthologues, suitable mouse homologues were selected. (B) This simplified schematic depicts how selected proteins encoded by AD risk genes (red, bold font) participate in pathways for microglial uptake and cellular activation. Lipoproteins containing apoE or apoJ may convey Aβ to microglia for uptake and degradation or may bind to TREM2 and stimulate ITAM-mediated cellular activation leading to chemotaxis, phagocytosis, survival, and transcription.

Figure 2.

Depiction of microglial cellular activities related to β-amyloid pathology. The left side illustrates protective microglial activities that limit disease progression. Microglia may clear Aβ peptides via macropinocytosis of soluble Aβ (1; Mandrekar et al., 2009), uptake of lipoprotein-associated Aβ (2), or phagocytosis of fibrillar Aβ aggregates (3). Microglia also help corral larger deposits of Aβ in plaques (4), minimizing damage to the adjacent neuropil. The right side illustrates disease states when microglial containment mechanisms are defective or outstripped. Aβ fibrils on the outskirts of plaque act as substrate for additional Aβ fibrillization and a reservoir of toxic Aβ species that induce neuritic dystrophy (5). Microglia can secrete factors that activate astrocytes (6) and participate in amyloid-dependent synapse loss (7). See also Fig. 3.

Figure 3.

Summary of studies manipulating the complement system or depleting microglia in mouse models of AD. (A) Simplified schematic of the complement pathway illustrating selected proteins. The complement system can be initiated by the classical, lectin, or alternative pathways. Central to complement activation is the cleavage of C3. Effects downstream of C3 cleavage include (1) phagocytosis after recognition of C3b opsonized material by complement receptors, including CR3 (inset); (2) inflammatory signaling by C3a and C5a fragment activation of C3aR and C5aR; and (3) lysis via formation of the C5b-C9 membrane attack complex. In the brain, microglia (yellow) mediate phagocytosis and respond to inflammatory signaling, and are also the cell type that produces C1q. Complement proteins demonstrated to play a role in synapse removal during developmental refinement of retinal ganglion cell projections to the lateral geniculate nucleus using knockout mice are indicated with a heavy border. Proteins that have been studied using knockout mice or inhibitors in the context of AD model mice are highlighted in red. (B) Table indicating the amyloid mouse models that have been tested, manipulations that were tested (complement protein knockout or inhibition or microglia depletion), the resulting impact on synapse or neuronal loss (“protection” indicates rescue of amyloid model deficits, whereas “loss” indicates the manipulation causes deficits), and effects on amyloid load. Blue fonts indicate phenotypes that suggest a beneficial effect of reducing complement activation or microglial cell numbers, and red fonts indicate phenotypes that suggest undesirable effects of reducing complement activation. *Note that in this study, the authors claimed not to deplete microglia but to block microglial proliferation. (C) Similar table as in panel B, except showing models of tauopathy and impacts on tau pathology. *Note that CD59 is an inhibitor of complement pathway activity, so the synapse/neuron loss seen with CD59 knockout is consistent with a beneficial effect of reducing complement activation. Reference (Ref) 1, Fonseca et al., 2004; Ref 2, Hong et al., 2016; Ref 3, Fonseca et al., 2017; Ref 4, Shi et al., 2015; Ref 5, Shi et al., 2017a; Ref 6, Maier et al., 2008; Ref 7, Wyss-Coray et al., 2002; Ref 8, Czirr et al., 2017; Ref 9, Fonseca et al., 2009; Ref 10, Olmos-Alonso et al., 2016; Ref 11, Spangenberg et al., 2016; Ref 12, Britschgi et al., 2012; and Ref 13, Asai et al., 2015. FB (FD, FH, FI), complement factor B (D, H, I); KO, knockout; MASP, MBL-associated serine protease; MBL, mannose-binding lectin.